New rhodamine B-based chromo-fluorogenic probes for highly selective detection of aluminium(iii) ions and their application in living cell imaging

Two rhodamine B-based fluorescent probes, BOS1 and BOS2, were designed and synthesized with good yields via the condensation reactions between the o-diaminobenzene modified rhodamine core structure (RBO) and salicylaldehyde derivatives. Both the probes exhibited remarkable absorbance-on and fluorescence-on responses to Al3+ over other metal ions in ethanol–water (1 : 9, v/v) medium via the rhodamine ring-opening approach, which can be used for “naked-eye” Al3+ detection over a broad pH range (5–9). The fluorescence intensities of the probes were linear with the Al3+ ion concentration, resulting in a low limit of detection of 1.839 μM (BOS1) and 1.374 μM (BOS2) for Al3+. In addition, the MTT assays and cell imaging experiments of Al3+ in SGC-7901 living cells demonstrated that the probes had negligible cytotoxicity, and were cell permeable and suitable for sensing Al3+ in biological systems.


Introduction
As the most abundant metal element in the earth's crust, aluminium has been widely used in our daily life, such as in food additives, clinical drugs, kitchen utensils, packing materials, water treatment, etc. [1][2][3][4][5] However, the heavy and inappropriate use of aluminium products in recent years has created adverse effects on life and environmental systems due to its toxicity. [6][7][8] In particular, the excessive intake of dissolved Al 3+ easily causes accumulation in the human body, which may cause several disorders, including Alzheimer's disease, osteomalacia, Parkinson's disease, and breast cancer. [9][10][11][12] Therefore, the fast detection and quantitative analysis of Al 3+ in the human body are crucially important for health warnings.
Because of the insufficient spectroscopic characteristics of Al 3+ , the reported Al 3+ -detection techniques, such as 27 Al NMR technology, atomic absorption spectrometry (AAS), 13 inductively coupled plasma atomic emission spectrometry (ICP-AES), 14 mass spectrometry and electrochemical methods, 15 are usually complex, time-consuming and costly. Fluorescence analysis, as a type of highly selective and sensitive detection method, not only possesses the features of easy operation, low cost, low detection limit and rapid response, but also can be used for real-time living organism detection, has attracted signicant interest of researchers. [16][17][18][19][20] Apparently, the reasonable design and preparation of effective biocompatible Al 3+ uorescent chemosensors are urgent problems need to be solved currently. [21][22][23][24] Considering the weak coordination ability and strong tendency to hydrolysis of Al 3+ ions, small-molecular uorescent probes with bright chromophore and multidentate coordination sites are highly promising. Recent studies reveal that the rhodamine-based probes are regarded as ideal candidates for the "off-on" uorescent sensors owing to the excellent photophysical properties such as remarkable photostability, high uorescence quantum yields, special response mechanism, strong anti-interference ability and long emission wavelength. [25][26][27][28] On one hand, with the introduction of target metal ions, the rhodamine framework can display "turn-on" uorescence signals through structural change between the spirocyclic and open-cycle forms. 29 On the other hand, the free-rotating benzoic acid group can be accessibly substituted by the functional groups with multidentate chelation sites to chelate target ions for uorescence sensor without changing the spirolactam forms. 30 Particularly, the Schiff base rhodamine probes have received lots of attention due to their high sensitivity and selectivity, rapid response time and ease of synthesis, 31,32 where the imines functional groups can be generated by the introduction of different carbonyl or amines to adjust the spectral response range of the rhodamine matrix, and at the same time to form suitable "ONO", "ONN" or "ONNO" sites to rapidly anchor the target metal ions. 33 Moreover, the imines also provide the opportunity of C]N isomerization for probes to generate richer uorescence sensing pathway. 34 Herein, two new rhodamine B-based sensors, BOS1 and BOS2 were designed and prepared by a three-step synthesis with simple raw materials (Scheme 1). The two chemosensors show highly selective and sensitive uorescence response to Al 3+ with the ring-opening sensor mechanisms described above, which can be used for naked-eye detection with rapid switch-on uorescence and remarkable color changes. Furthermore, these advanced characteristics endow the two probes highly promising for biological imaging applications.

Materials and methods
All chemicals were of analytical-reagent grade, and they were commercially available from commercial sources and used without further purication. The SGC-7901 living cells (human gastric carcinoma cells) were obtained from Xi'an Jiaotong University Health Science Center. The twice-distilled water was used throughout the experiment. The solid powders of probes BOS1/BOS2 were dissolved in ethanol solution in concentration of 1 mM as stock solutions. And then took out quanticational BOS1/BOS2 in different testing systems. Fluorescence spectra were carried on a HITACHI F-4500 uorescence spectrophotometer. UV-vis spectra were performed on a Shimadzu UV-1700 spectrophotometer. The elemental analyses of C, H, and N were performed on a Vario EL III elemental analyzer. IR spectra were recorded on a Bruker Tensor 27 spectrometer. NMR spectra were obtained on a Varian INOVA-400 MHz spectrometer (at 100 MHz for 13 C NMR and 400 MHz for 1 H NMR). A Bruker micro TOF-Q II ESI-TOF LC/MS/MS Spectroscopy was used to perform mass spectra. Melting point tests were taken on an XT-4 micromelting apparatus and uncorrected. Results of cytotoxicity were analyzed with the So max pro soware (version 2.2.1) in Spectra max190-Molecular Devices. The living cells imaging were performed on an Olympus FV1000 confocal microscopy with l ex ¼ 400 nm.

Synthetic procedures
Synthesis of RBO. Phosphorus oxychloride (1 mL, 10.09 mmol) was added to a dry 1,2-dichloroethane solution (20 mL) of rhodamine B (1.59 g, 3 mmol), and the resulting mixture was reuxed for 6 h. Aer cooling to room temperature and removal of the solvent in vacuo, the crude product of rhodamine B chloride oxide was obtained without any purication. Then the residue was directly reacted with an acetonitrile solution (30 mL) of o-phenylenediamine (0.324 g, 3 mmol) at room temperature for 30 min, followed by addition of 2 mL triethylamine as acidcapturer, the mixture was stirred for 8 h. Aer removal of the solvent under reduced pressure, the crude product was puried by silica gel column chromatography to give RBO in a yield of 89.3%. HRMS (ESI) calcd for (C 34

Preparation of the test solution
The 10 mM stock solution of probes BOS1/BOS2 were prepared in ethanol-water (1 : 9, v/v, Tris-HCl, pH ¼ 7.2). The solutions of various testing cation species were prepared from Ca( Before spectroscopic measurements, the corresponding solutions of probes were freshly prepared by diluting the high concentration stock solutions. All the measurements were made according to the procedures as follows. Placing 1 mL of the probe solution and an appropriate aliquot of each metal stock into a 10 mL glass tube, and diluting the solution to 10 mL with ethanol-water (1 : 9, v/v, Tris-HCl, pH ¼ 7.2) solution.

Cytotoxicity assays
The MTT assays were performed to evaluate the toxicity of BOS1, BOS2 and Al 3+ by SGC-7901 living cells. 35 90% conuent cells were chosen, digested by 1 mL 0.25% of trypsin, and transferred in 96-well plates. The cells were treated and incubated at 37 C under 5% CO 2 in culture medium (DMEM (Dulbecco's Modied Eagle Medium) + 10% FBS (Fetal Bovine Serum)) and maintained 24 h. Different concentrations of BOS1, BOS2 and Al 3+ were added to the 96-well plates, respectively. Another 24 hours incubation was taken at the same condition. Following this, the medium was removed and washed three times with phosphate buffered saline (PBS). Then the medium was replaced with mixed liquor of MTT (5 mg mL À1 ) and culture medium, and incubated for an additional 4 h. Aer that, the MTT was removed and washed three times with PBS. Subsequently, 150 mL DMSO was carefully added to each well and ultrasonic oscillation for 10 minutes. All the experiments were conducted in triplicate. The cell viability (%) was calculated according to the equation: cell viability (%) ¼ [OD 490 (sample)/OD 490 (control)] Â 100%, where OD 490 (sample) represents the optical density of the wells treated with various concentration of probes or metal ions and OD 490 (control) represents that of the wells treated with ethanol.

Cell culture and uorescence imaging
The SGC-7901 living cells (human gastric carcinoma cells) were cultured in DMEM replenished with 10% FBS. Before the experiments, cells were processed with probes BOS1/BOS2 (20 mM) for 1 h at 37 C in humidied air and 5% CO 2 , washed three times with PBS then imaged. Aer incubation with Al 3+ (20 mM) for another 1.5 h at 37 C, cells were washed three times with PBS to remove remaining metal ions and then imaged. Confocal uorescence imaging was carried out with an Olympus FV1000 laser scanning microscope with 80Â objective lens.

Spectroscopic properties
Both of the probes BOS1 and BOS2 exhibited highly selective and sensitive response to Al 3+ ion in ethanol-water (1 : 9, v/v). Although the different salicylaldehyde moieties had effects on the uorescence/absorption intensities and response time, no obvious differences were observed on the other spectral properties of the two probes. Therefore, only the property of BOS1 is described in detail. The optical spectra of BOS2 are given in the ESI. † The uorescence sensing ability of BOS1 toward metal ions was investigated in ethanol-water (1 : 9, v/v, Tris-HCl, pH ¼ 7.2) solution. As shown in Fig. 1a, the free BOS1 exhibited negligible uorescence emission due to its spirolactam form. Upon the further addition of different metal ions (Co 2+ , K + , Ca 2+ , Mg 2+ , Cd 2+ , Mn 2+ , Ni 2+ , Ba 2+ , Li + , Na + , Zn 2+ , Cu 2+ , Pb 2+ , Ag + , Cr 3+ , Pd 2+ , Hg 2+ , Sn 2+ , Fe 3+ , Al 3+ ), only the Al 3+ ion led to a remarkable luminescence enhancement at a maximum emission wavelength of 592 nm, while no obvious luminescent intensity changes were observed in the presence of other metal ions. These results indicated the high selectivity of the BOS1 probe for Al 3+ detection.
Fluorescence titration experiments were performed to investigate the interaction between BOS1 and Al 3+ (Fig. 1b). Since the stable and characteristic "spirolactam form" of rhodamine B group, free BOS1 shows colorless and no uorescence response in the visible region range from 480 to 660 nm. However, along with the gradual addition of Al 3+ , the uorescence emission intensity at 592 nm was signicantly enhanced with a color variation from colorless to orange (Fig. 2), suggesting that the xanthene moiety of rhodamine B was subjected to the delocalization interference, and BOS1 was a true "off-on" chemosensor for Al 3+ . Aer the addition of 1.0 equiv. of Al 3+ , the titration curve reached a steady plateau accompanied by more than 100-fold increase in the emission at 592 nm compared with that of free BOS1. Such signicant enhancement of uorescence clearly indicated that the spirolactam form of rhodamine B was unfolded or the rotation of the "C]N" group were inhibited owing to gradually adding Al 3+ to BOS1, and a highly delocalized p-conjugated system was ultimately formed. The association constant K, of BOS1 with Al 3+ was calculated according to the Benesi-Hildebrand equation: 36,37 where F max , F 0 and F x are the uorescence intensities of probe in the presence of Al 3+ at saturation, free probe, and any intermediate Al 3+ concentration, respectively. The binding constant value was found to be K ¼ 4.48 Â 10 À4 M (for BOS2: K ¼ 4.92 Â 10 À4 M, Fig. S3 †).
Moreover, the detection limit of BOS1 and BOS2 were determined from the result of titrating experiment. As shown in Fig. 3, according to the widely used method, 38,39 linear regression curves were tted based on the plots of (F min À F x )/(F min À F max ) vs. log[Al 3+ ], where the F x is the uorescence intensity at 592 nm at each concentration of Al 3+ added, F min and F max are respectively the minimum and maximum uorescence intensity at 592 nm, thus the intercepts of the lines at x-axis were taken as the detection limit of BOS1 (1.839 mM) and BOS2 (1.374 mM) (Fig. S4 †).
The outstanding selectivity to the target detective metal ion over other potentially competing species is crucial for the application of metal ions sensors. So the uorescence response of BOS1 (10 mM) towards Al 3+ ions and other various metal ions in EtOH/H 2 O ¼ 1/9 (v/v, Tris-HCl, pH ¼ 7.2) were investigated. As depicted in Fig. 4, no obvious changes of uorescence intensity could be detected when 1.0 equiv. metal ions (Co 2+ , K + , Ca 2+ , Mg 2+ , Cd 2+ , Mn 2+ , Ni 2+ , Ba 2+ , Li + , Na + , Zn 2+ , Cu 2+ , Pb 2+ , Ag + , Cr 3+ , Pd 2+ , Hg 2+ , Sn 2+ , Fe 3+ ) were added into the relevant solution. Conversely, about 100-fold enhancement of emission intensity at 592 nm appeared obviously in the presence of subsequent 1.0 equiv. Al 3+ ions, indicating that the recognition of Al 3+ ions by the probe BOS1 is not interfered by other coexisting metal ions. The above facts reveal that the BOS1 shows high selectivity, anti-interference and sensitivity toward Al 3+ ions, and could be potentially applied to detect Al 3+ ions in complex systems. The similar Al 3+ response performances were also observed for the probe BOS2 (Fig. S5 †).   The UV-vis absorption of the probes (10 mM) was also investigated in an ethanol-water (1 : 9, v/v, Tris-HCl, pH ¼ 7.2) solution. As can be seen in Fig. 5a and S6, † both the free BOS1 and BOS2 exhibited nearly no absorption bands in the visible region, which may be attributed to their closed spirolactam forms. As expected, a signicant enhancement of the absorption at 568 nm (BOS1) or at 559 nm (BOS2) was observed in the presence of Al 3+ ions, whereas tiny absorption changes occurred when nineteen other metal ions added, respectively. With the increasing concentration of Al 3+ ions in the range of 0-100 mM, the absorption band gradually enhanced, indicating the ringopening form of the rhodamine spirolactam of BOSs. The absorption intensity showed negligible changes with further increasing the concentration of Al 3+ (up to 12.5 mM), which suggested the saturated binding behaviours between Al 3+ and BOS probes (Fig. 5b and S7 †). The association constant for Al 3+ ions was calculated to be 4.39 Â 10 4 M À1 (BOS1) and 4.76 Â 10 4 M À1 (BOS2) from the absorption titration curves. Moreover, the dramatical color changes from colorless to peach-red associated with the reaction of probes with Al 3+ ions were visually detectable with good selectivity, which indicated that BOS1/BOS2 could successfully serve as "naked-eye" probes for Al 3+ detection (Fig. 2).
It is well known that the spirolactam ring of the rhodamine derivative is commonly open in acidic media and shows the uorescence of rhodamine. Therefore, the optimal pH conditions for the probes BOS1/BOS2 should be evaluated to affirm their stabilities for potential practical applications. The pH dependent uorescence responses of BOS1 and BOS2 in the presence and absence of Al 3+ were recorded in the pH range of 2-12 ( Fig. 6 and S9 †). For BOS1 system, the uorescence intensities of both BOS1 and the BOS1-Al 3+ species were strong enough when pH < 4, which could be due to the ring opening of rhodamine derivatives induced by strong protonation of the tertiary amine-N atom in acid conditions. No obvious emission of free BOS1 was observed when pH > 5, while the strong   Paper uorescence emissions aer the addition of Al 3+ within the pH range of 5-9 were detected, which revealed that the BOS1-Al 3+ complex was formed in this pH region and the BOS1 probe towards Al 3+ could work well in such approximate physiological conditions with a low background response. With further increasing the pH value, the emission intensities were quenched because of the decoordination of Al 3+ , leading to the formation of Al(OH) 3 and the reformation of spirolactam rings. Similar events were also found in the BOS2 system, the BOS2-Al 3+ species presented the strongest uorescent responses within an optimal pH range of 5-9. The suitable pH response range suggested that no buffer solutions were required for the detection of Al 3+ , and both BOS1 and BOS2 would provide the potential practical applications in environmental systems or living cells.
In addition, the response time is important to the application of naked-eye detection. So the time dependent uorescence responses of BOS1 and BOS2 in the presence of Al 3+ were carried out in a simulated in vivo environment (ethanol-water 1 : 9, v/v, Tris-HCl, pH ¼ 7.2) at room temperature. As the BOS1/BOS2 interacted with Al 3+ , the uorescence intensities of the analysis systems signicantly increased to the maximum value within approximately 30 s for BOS1 (Fig. 7) and 42 s for BOS2 (Fig. S10 †). These results show that BOS1 and BOS2 are reliable instantaneously responsive colorimetric sensor for Al 3+ .

Proposed mechanism for the interactions between probes and Al 3+
According to the results of spectroscopic responses of BOS1/ BOS2 to Al 3+ , we speculated that the probable binding ways and interaction mechanisms between BOS1/BOS2 and Al 3+ were likely due to the chelation-induced ring opening of rhodamine spirolactam (Scheme 2), rather than other possible reactions, which were similar to those reported in previously literatures. [40][41][42][43][44] That is, the Al 3+ ions coordinated to the phenolic hydroxyl O, imino N, benzoylamide O and N atoms of the probes to form conjugated moieties, and the lactam rings of rhodamine were induced to be opened, exhibiting signicant uorescence enhancements.
Binding analysis was further performed to determine the ratio between probes and Al 3+ by using the method of continuous variations (Job's plots). As shown in Fig. 8, a maximum uorescence emission at 592 nm was observed when the molecular fraction of Al 3+ is close to 0.5, which revealed that the Al 3+ -chemodosimeter displayed 1 : 1 stoichiometry, and further proved the above-mentioned binding modes.
The binding reversibility of the probe BOS1 was also examined by the EDTA-adding experiments at room temperature. As illustrated in Fig. 9, the absorbance and uorescence intensities rapidly decreased when EDTA was added to the Al 3+ -BOS1 system. Meanwhile, the color of the solutions changed from peach-red to colorless. When Al 3+ ions were dropwise added into these systems again, the spectral signals were almost completely reproduced and the peach-red solutions appeared again. These reversible processes can be repeated several cycles without signicant uorescence changes (Fig. 9b, inset), indicating that the BOS probes are reversible uorescence sensors toward Al 3+ .

Cytotoxicity and uorescence imaging
The MTT assays were performed to explore the cytotoxic effects of BOS1, BOS2 and Al 3+ according to the reported method. 35 The relevant data expressed as mean AE standard deviation were listed in Table S1 † and the results were depicted in Fig. 10. The SGC-7901 living cells (human gastric carcinoma cells) viability remained 84.52%, 90.38% and 86.58% aer the treatment of 25 mM probes BOS1, BOS2 and Al 3+ , respectively, which indicated that all of them were low cytotoxic to cells and suitable for bioimaging.
Fluorescence imaging experiments were conducted in the living cells to further demonstrate the practical applicability of the probes in biological samples. 45 Fig. 11 presented the uorescence images of SGC-7901 cells recorded before and aer the addition of Al 3+ . Apparently, free BOS1 and BOS2 probes showed no detectable uorescence signals in living cells in the absence of Al 3+ (Fig. 11a). By contrast, bright uorescence signals were observed in living cells aer incubation with Al 3+ (Fig. 11c). Bright-eld transmission images of cells treated with probes and target Al 3+ ions showed that the cells were viable throughout the imaging experiments (Fig. 11b). The results suggested that probes BOS1/BOS2 possessed the capacity to readily penetrate the cell membrane and could be applied for in vitro imaging of Al 3+ in living cells, and potentially in vivo.

Conclusions
In summary, two novel rhodamine-based uorescent probes BOS1 and BOS2 were designed and synthesized. Upon binding with Al 3+ , dramatic uorescence and absorption enhancements were observed due to the formation of ring-opening of rhodamine species, showing distinct color changes and switch-on uorescence. The probes displayed high selectivity, low detection limit, and fast response to Al 3+ ions over other examined metal ions in ethanol-water system. Moreover, we have demonstrated their biological application by uorescence imaging intracellular Al 3+ in SGC-7901 living cells. We expect that the chemosensors would help to promote the studies of Al 3+ in complex biological systems.

Conflicts of interest
There are no conicts to declare.