A novel dual-capability naphthalimide-based fluorescent probe for Fe3+ ion detection and lysosomal tracking in living cells

We design and synthesize a novel 1,8-naphthalimide-based fluorescent probe MNP that features the dual capabilities of tracking lysosomes in living HeLa cells and sensitively detecting Fe3+ ions in aqueous solution. The MNP is obtained by modifying the morpholine group with a lysosomal targeting function and the piperazine group with an Fe3+ ion recognition function on the 1,8-naphthalimide matrix. In the presence of Fe3+ ions, the MNP acts as a recognition ligand to coordinate with the central Fe3+ ion, and the protonated [MNPH]+ is eventually generated, in which significant fluorescence enhancements are observed due to the intramolecular photo-induced electron transfer (PET) process being blocked. The limit of detection of Fe3+ ions is as low as 65.2 nM. A cell imaging experiment shows that the MNP has low cytotoxicity and excellent lysosomal targeting ability. Therefore, the MNP offers a promising tool for lysosomal tracking and relevant life process research.


Introduction
In recent years, subcellular-organelle targeting strategies have evolved signicantly and redened the future development of multifunctional nano drugs and the clinical transformation of precision medicine. [1][2][3] As eukaryotic organelles, lysosomes contain many hydrolases and secretory proteins that are active in the pH range of acidic solution (3.5-5.5). Lysosomes are the terminal degradation chambers of living cells, and are involved in many physiological processes such as metabolism, apoptosis, intracellular transport, and immunologic defense. Lysosomal dysfunction can lead to various diseases, especially cancerrelated diseases. [4][5][6] Therefore, an effective lysosomal tracking strategy for cancer cells is critical for the prevention and treatment of tumor invasion and metastasis, and will be helpful in guiding the diagnosis and treatment of lysosome-related diseases. 7,8 Compared with other methods, the uorescent probe method is more ideal for subcellular tracking due to its simplicity, rapid response, good biocompatibility, and high sensitivity. 9,10 Therefore, it has become an important focus of research to develop novel uorescent probes that can reversibly monitor lysosomal images to track lysosomes. Recently, a number of uorescent probes have been reported for lysosomal tracking in living cells. However, there are few reports of dual-capability probe that can track lysosomes and detect the analyte in lysosomes.
Iron is one of the most important trace elements in human body, and it plays a pivotal role in many physiological reactions. There are two main forms of iron in human body, iron storage compounds and iron-containing biologic molecules. Fe 3+ ions, in particular, play a key part in oxygen transport, oxygen metabolism, transfer of electrons, and many catalytic reactions. [11][12][13][14] The deciency or overload of iron may lead to biological dysfunction and disturb the cellular homeostasis in vivo, which will in turn cause anaemia, diabetes, Alzheimer's disease, liver injury, heart and renal failure, other conditions. [15][16][17][18][19][20] Besides, iron ions can also cause environmental pollutions, such as water pollution, which is harmful to human health. As stipulated by the U.S. Environmental Protection Agency, the maximum acceptable content of iron in drinking water should be 5.357 mmol L À1 . 21 Therefore, the sensitive and rapid determination of Fe 3+ ions is essential for the protection of physiological and natural environments. Recent years, a large number of uorescent probes have been designed and synthesized for the detection of Fe 3+ ions. 22 However, Fe 3+ ions tend to form insoluble Fe(OH) 3 under aqueous conditions at neutral pH. In lysosomes (pH 3.5-5.5), new iron-containing species such as [Fe(OH) 2 ] + and [Fe(OH)] 2+ can be formed with the release of protons. At present, there are few Fe 3+ ions uorescent probes that can detect these ions directly, especially when uorescence enhancement (turn-on) response is required. 23 1,8-Naphthalimide and its derivatives play a key role in the uorescent dye eld. They have been extensively applied in biochemistry, polymers, optical storage, uorescent sensors, and biological medicine due to its good thermal and oxidation stability, high electron affinity, large Stokes shis, good photostability, and high uorescence quantum yields. [24][25][26] As the "simplest" molecule, 1,8-naphthalimide has better water solubility and an easier functionalization process than other molecules. Moreover, 1,8-naphthalimide is a typical photo-induced electron transfer (PET) dye for fabricating uorescent probes, which is considered an important strategy for the design of uorescence sensors. These probes are usually built in the format of "uorophore-spacer-receptors". 27 Based on these characteristics, it can offer ideal uorophores for fabricating uorescent sensors. However, few naphthalimide derivativebased sensors are currently available for applications in subcellular imaging.
In this work, we designed and synthesized a new type of dualcapability uorescent probe (MNP) based on 1,8-naphthalimide. The probe comes in three parts: (1) a 1,8-naphthalimide group acting as the uorophore; (2) a morpholine group acting as the lysosomal targeted functional group; (3) a Nmethyl piperazine group acting as the Fe 3+ ions recognition group. It has been proved that the MNP can not only obtain lysosomal targeted images of living cells, but also detect Fe 3+ ions in aqueous solution with high sensitivity. Induced by Fe 3+ ions, the MNP eventually transforms into protonated [MNPH] + in aqueous solution, which leads to the formation of "turn-on" green uorescence due to the blocked PET process (Scheme 1). Besides, the applicability of the MNP in bioimaging was examined by confocal uorescence microscopy. This work provides an innovative idea for the design of dual-capability probes.

Characterization
The 1 H and 13 C NMR spectra were obtained with a Bruker DRX-400 spectrometer and CDCl 3 was used as the solvent. The UV-vis spectra were acquired on a Varian Cary-5000 spectrometer. The corrected steady-state excitation and emission spectra were obtained on an F-7000 spectrometer. Using the time-correlated single photon counting technique in 4096 channels, uorescence decay histograms were obtained on an Edinburgh Instruments FLS-920 spectrometer equipped with a supercontinuum white laser (400-700 nm). Histograms of the instrument response functions (using a LUDOX scatterer) and sample decays were recorded until they typically reached 1.0 Â 10 4 counts in the peak channel. The quantum yield (QY) of MNP was calculated on an Edinburgh Instruments FLS-920 through a comparative method in which coumarin 6 in methanol was used as the reference. The QY of the MNP was calculated by this formula: where F is the QY, I is the measured integrated emission intensity, h is the refractive index of the solvent, and A is the absorption of the reference (r) and as-prepared MNP sample (s). Mass spectra were detected in E.I. mode. High-resolution mass spectra (HR-MS) were obtained on a Bruker micro-TOF-Q II mass spectrometer. Melting points were obtained with an X-4 precise micro melting point cryoscope. To obtain the effects of pH, pH measurements were made with a pH-10C digital pH meter. Cell images were obtained on an FV1000-IX81 confocal laser scanning microscope (Olympus, Japan).

Synthesis of MNP
As shown in Scheme 2, the intermediate product (MN) was synthesized via a method similar to that in the literature. 28 Briey, 4-bromo-1,8-naphthalene anhydride (1.39 g, 5 mmol) and 2-morpholin-4-ylethanamine (0.67 g, 5 mmol) were added into 30 mL of ethanol. Then the mixture was stirred and reuxed for 4 h. The compound MNP (probe) was synthesized by the following method, which has never been reported before.

Detection of Fe 3+ ions
The detection of Fe 3+ ions was performed at room temperature in HEPES buffer. The aqueous solutions of Fe 3+ ions and other metal ions with different concentrations were freshly prepared before use, and some of them were stored in acidic conditions. To evaluate the sensitivity towards Fe 3+ ions, different concentrations of Fe 3+ ions were added into the HEPES buffer containing 10 mM MNP, and equilibrated for 5 min before spectral measurements. The uorescence spectra were recorded by F-7000 spectrometer with an excitation wavelength of 405 nm. The river water sample was obtained from Hunhe River of Shenyang, Liaoning Province, China. The sample was rst centrifuged at 8000 rpm for 10 min to remove main impurities and then ltered with 0.22 mm membrane. The river water samples with various concentrations of Fe 3+ ions were added to the MNP sensing system and then the uorescence spectra were collected.

Cell culture
To obtain the cell permeability of the MNP, HeLa cells were cultured in Dulbecco's Modied Eagle Medium (DMEM) subjoined with 10% (v/v) fetal bovine serum (FBS). The cell lines were maintained under a humidied atmosphere of 5% CO 2 and at 37 C. HeLa cells were treated with 10 mM MNP in 1.0 mL of fresh culture medium for 30 min and then treated with 100 mM Fe 3+ ions for another 30 min, which were compared with those in the blank experiment. Before the cell imaging experiment, HeLa cells were washed three times with PBS buffer to remove free compounds. Confocal uorescence images of HeLa cells were captured on an Olympus FV1000-IX81 laser confocal microscope.
Moreover, a standard 3-(4,5-dimethylthiazol-2-yl)-2,5diphenyltetrazolium bromide (MTT) assay was performed to evaluate the cytotoxic effect of the MNP. HeLa cells were seeded in 96-well assay plates at a density of 10 4 cells per well (100 mL total volume/well) for 24 h. Then, various concentrations of MNP (10, 20, 40, 60, 80 and 100 mM) were added to the serumfree medium and incubated with HeLa cells for 24 h. The optical absorbance of the cells was detected through a microplate reader (German Berthold Mithras2LB943) at 450 nm. The assay was performed in ve sets for each concentration of MNP and the control experiment was conducted by measuring the growth culture medium without the MNP.

Synthesis and characterization
In this work, the probe (MNP) was designed using 1,8-naphthalimide as the uorophore, which is a typical PET-based uorescent dye. The MNP can be synthesized by the intermediate product MN and N-methyl piperazine in 2-methoxyethanol, with the -Br group in the MN substituted by the piperazine group (Scheme 2). In the MNP molecules, the 1,8naphthalimide group acted as the electron acceptor and the piperazine group as the electron donor, forming an electron transfer system within the molecules and inducing a PET process under the excitation of light. The active PET process also "turned off" the uorescent signal of the MNP, so that the MNP had very weak uorescence. A protonated species [MNPH] + with strong uorescence would be obtained when the MNP was treated with either excessive FeCl 3 aqueous solution or HCl (Scheme 1). It has been reported that the coordination compounds of FeCl 3 with various modied N-aryl or N-alkylpiperazine ligands. When Fe 3+ ions act as the central metal ions to coordinate with these ligands, the ligands may act as a bidentate ligand for coordination, which adopts a boat conguration with a stoichiometric ratio of 2 : 1. 29 The Job's plot analysis also showed that the binding mode of MNP ligands to Fe 3+ ions was 2 : 1 (Fig. S1 †). However, the rapidly formed intermediate species [(MNP) 2 FeCl 2 ] + cannot be extracted at laboratory scale from wet solvents used for uorescence sensing and bioimaging experiments (Fig. S3 †). This was because it rapidly decomposed to form a species [MNPH]Cl, which blocked the PET process in the system and signicantly enhanced uorescence (PET inactive) to achieve the uorescence sensing of Fe 3+ ions (Scheme 3). These new compounds were characterized by 1 H and 13 C NMR spectroscopies and HR-MS (Fig. S5-S10 †).
We prepared the stock solution of the free MNP in EtOH, and then investigated the UV-vis absorption and uorescence response to Fe 3+ ions at room temperature in 10 mM HEPES buffer (pH 7.4; EtOH : H 2 O ¼ 5%; v/v). Under this condition, the free MNP exhibited one main absorption band centred at $405 nm, which can be attributed to a p-p* transition. Aer the addition of Fe 3+ ions (0-10 eq. FeCl 3 ) into 10 mM MNP solution in the above-mentioned 10 mM HEPES buffer, the maximum absorption peak intensity increased gradually and exhibited a distinct blue-shi from 405 to 372 nm (Fig. 1a). Fig. 1b shows the uorescence changes of the MNP in the absence and presence of Fe 3+ ions in HEPES buffer (pH 7.4). When excited with an optimal excitation wavelength of 405 nm, the free MNP only showed weak uorescence emission (black line) at $510 nm due to the active PET process. However, when the concentration of the Fe 3+ ions increased from 0 to 100 mM, the intensity of uorescence emission increased signicantly at $510 nm, which can be attributed to the initial capture of Fe 3+ ions by the MNP molecules. This resulted in a lower electrondonating ability of piperazine-nitrogen, thus blocking the PET process. Besides, the protonated species [MNPH] + was nally generated in a very short time, which led to the recovery of strong uorescence emission (Scheme 3). Furthermore, the data analysis exhibited an excellent linear relationship (R ¼ 0.997) between the relative uorescence intensity (F À F 0 ) at $510 nm and the concentration of Fe 3+ ions (0-20 mM). Meanwhile, the limit of detection (LOD) was 65.2 nM, indicating that the MNP is a highly sensitive uorescent probe (Fig. 1c). The LOD was obtained by the 3s/k method, where s is the standard deviation of the blank sample and k is the slope of the calibration curve. 30 The quantum yield F s of the MNP was 0.01 AE 0.004, and in the presence of Fe 3+ ions (100 mM), it increased to 0.272 AE 0.008 due to the PET process.
Fluorescence decay traces of the MNP with Fe 3+ ions were recorded at 510 nm by the single-photon timing method. In the presence of Fe 3+ ions (1 eq. and 10 eq.), the uorescence decay can be tted to the double-exponential prole with lifetimes of $3.55 ns, $9.4 ns and 4.61 ns, 8.92 ns in HEPES buffer; the average uorescence lifetimes were 8.77 ns and 8.11 ns, respectively. As shown in Fig. 1d, aer the addition of Fe 3+ ions, the MNP uorescence decay became slower and the average life time became longer signicantly. These results can be attributed to the fact that the PET process was blocked by the reaction with Fe 3+ ions.

Selectivity to M 3+ metal cation
In addition to good sensitivity, good specicity was also required. The selectivity of the MNP to Lewis acids such as Fe 3+ , Cr 3+ and Al 3+ ions was evaluated by screening its uorescent response to various biological ions and toxic metal ions in HEPES buffer (pH 7.4; EtOH : H 2 O ¼ 5%; v/v). As shown in Fig. 2, under the same condition, the addition of trivalent cation such as Fe 3+ , Cr 3+ and Al 3+ ions resulted in a signicant uorescence enhancement, and no obvious changes in uorescent signal were observed by adding 10 eq. of various biological ions (Ca 2+ , Co 2+ , K + , Mn 2+ , Na + , Ni 2+ , Fe 2+ , Mg 2+ and Cu 2+ ) and toxic metal ions (Hg 2+ , Pd 2+ and Ag + ). The obtained results demonstrated that the probe MNP has high selectivity and sensitivity towards trivalent cation Lewis acids such as Fe 3+ , Cr 3+ and Al 3+ ions, especially aqueous Fe 3+ ions, which are essential for living organisms and are abundantly present on Earth. To demonstrate the application of the MNP in complicated environment, we further detected the concentration of Fe 3+ ions in tap water and river water samples. As shown in Tables S1 and S2 † recoveries of different known amounts of added Fe 3+ ions were obtained from 98.4% to 113.5% in tap water samples and 96.7% to 109.3% in river water samples. Therefore, the probe can be applied to the sensing of trivalent cations and the rapid detection of Fe 3+ ions with high sensitivity and selectivity under specic conditions.

Photostability and pH dependence
We investigated the photostability of the MNP in HEPES buffer (pH 7.4; EtOH : H 2 O ¼ 5%; v/v) and as shown in Fig. 3a, the results revealed that the MNP has excellent photostability and its uorescence emission intensity at $510 nm remained almost unchanged under continuous illumination with a UV lamp for 30 min at room temperature. In addition, we investigated the uorescence responses of the MNP to Fe 3+ ions in HEPES buffer (EtOH : H 2 O ¼ 5%; v/v) at various pH values.
Scheme 3 A plausible mechanism of the MNP for ultrasensitive sensing of Fe 3+ ions.
Furthermore, the uorescence intensity changes of MNP were independently tested in a wide range of pH (3)(4)(5)(6)(7)(8)(9)(10)(11)(12). With the increase of pH values, the MNP showed a quenching trend, the uorescence intensity was almost entirely quenched at pH > 8. Under acidic condition, the "turn-on" uorescent signal was observed, which is very suitable for uorescence imaging of  lysosomes. The uorescence intensity of the MNP in the presence of (10 eq.) Fe 3+ ions was almost constant within a pH range of 3-8, and the uorescence quenching of MNP + Fe 3+ started at approximately pH 8 and entirely quenched at pH > 11. Therefore, we could conclude that the MNP has the potential to be used as a sensitive "turn-on" uorescent probe for detecting Fe 3+ ions in the biologically relevant pH range in vitro (Fig. 3b).

Cellular imaging
For the applicability of the MNP in living cell imaging, its cytotoxicity is a signicant consideration. The cytotoxicity of the MNP was evaluated by the conventional MTT assay in living HeLa cells. Aer treatment with the MNP at a concentration of 100 mM at 37 C for 24 h, the cell viability of HeLa showed no signicant decrease (Fig. S4 †), which indicated that the MNP prepared in this study has good biocompatibility and low cytotoxicity in living cells. Consequently, it can be predicted that the MNP is effective in tracking lysosomes in living cells.
Based on the above results, we assessed the applicability of the MNP in the uorescence imaging of lysosomes in vitro. As shown in Fig. 4a-c, the imaging of normal HeLa cells under the confocal uorescence microscope has no uorescence (blank experiment). Unlike the blank experiment, HeLa cells were incubated with 10 mM MNP and 10 eq. Fe 3+ ions at 37 C for 30 min, and then the results were analyzed by confocal uorescence microscopy. A signicant unevenly distributed punctate green uorescence image was observed in the living HeLa   Fig. 4d-f). The experimental results revealed that the MNP has good membrane permeability and can be used for uorescence imaging in living cells.

cells (
To examine the suitability of the MNP for in situ imaging of lysosomes, co-localization tests were performed by co-staining HeLa cells with a commercial Lyso-Tracker Red, a living cell lysosomes tracker along with the MNP. Confocal uorescence images of the MNP with Lyso-Tracker Red were recorded in separate optical inspection windows, with minimum interference between each other. The MNP showed punctate uorescence extremely similar to Lyso-Tracker Red under the confocal uorescence microscope, and a large area of overlap appeared in the uorescence overlay image (Fig. 5a-d). The intensity curve analysis of the region of interest also showed that the peak location and peak intensity of the MNP are basically identical with that of the Lyso-Tracker Red (Fig. 4e). The above colocalization test results demonstrated that the MNP has excellent lysosomal targeted imaging capability and the potential to be applied for tracking lysosomes in living cells.

Conclusions
In summary, we designed a novel 1,8-naphthalimide-based uorescent probe MNP with dual capabilities for rapid and sensitive detection of Fe 3+ ions in vitro and for in situ imaging of lysosomes in living HeLa cells. When the piperazine group in MNP molecules are bound to Fe 3+ ions as recognition ligands, it can block the PET process and restore the strong uorescence emission, which offers high sensitivity to Fe 3+ ions in the aqueous medium, and the LOD is 65.2 nM. In the actual sample analysis, the good recoveries ranging from 96.7% to 113.5% also illustrated that the application of the MNP in biological and food sample analysis was anticipated to be promising. In addition, due to its excellent lysosomal targeted imaging ability conrmed by co-localization tests, the probe MNP can also be used to monitor the morphological changes of lysosomes in living cells in real time, which provides a new strategy for lysosome-related medical research and clinical diagnosis. Thus, we believe that this simple and cost-effective dual-capability probe strategy will nd wide applications in biochemical and medical elds.

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