A Cryptand-Type Aluminum Tris(salophen) Complex: Synthesis, Characterization, and Cell Imaging Application

Abstract

Metal salen/salophen complexes have been used as fluorescent probes for cell imaging with various metal centers. Herein we synthesized cryptand-type aluminum salophen complexes LAl₃ and the corresponding mononuclear compound LAl. X-ray crystal diffraction verifies the cryptand-type structure of LAl₃ with C_3h symmetry. Both LAl₃ and LAl show moderate green fluorescence with quantum yields of 0.17 and 0.05, respectively. The hydrophilic and cationic nature of these aluminum salophen complexes renders them enhanced cellular uptake. Both complexes are internalized into cells by energy-dependent pathways and they distribute in lysosomal organelles.

1. Introduction

Metal salen/salophen complexes have been widely studied in catalysis [1,2,3,4] and supramolecular chemistry [5,6,7], and in recent years they have also attracted a great deal of attention for their biological applications as fluorescent probes and prodrugs [8,9,10,11,12,13,14,15,16]. In our group, a series of Zn salen/salophen complexes have been reported as fluorescent probes for cell imaging even with super-resolution, with multiple functions such as intracellular reactive oxygen species, monitoring the lipid droplets and autophagy processes [17,18,19,20,21,22,23]. However, Zn salen complexes tend to aggregate through the axial intermolecular interactions between the Zn center and the phenolic oxygen of another molecule, especially in non-coordinating solvents [24,25,26,27,28,29]. The aggregation not only weakens fluorescence intensity, but also influences the cellular uptake pathway and subcellular localization [30]. To minimize the effect of the intermolecular Zn···O interactions, several approaches have been reported, such as adding an axial ligand, introducing positive charges in the ligand, and increasing the number of sterically hindered substituents [31]. This also includes a series of trinuclear Zn salophen complexes with cryptand-type structures, which could reduce the intermolecular Zn···O interactions [32].

However, the trinuclear Zn salophen complexes suffer from little intracellular fluorescence signal, probably due to low cellular uptake arising from the hydrophobicity and neutral form and thus could not be directly used as cell imaging probes. To enhance the cellular uptake, we envisioned to use the main group element Al (3+) cation for (1) form the cation complexes; (2) closed shell structure similar to Zn²⁺; and (3) Lewis acid reactivity. These features render Al salen/salophen complexes better cell imaging probes than Zn complexes [33,34]. Previously, we demonstrated that monomeric [AlSalen]⁺Cl⁻ displayed higher sensitivity and stronger intracellular fluorescence in monitoring intracellular microvisicosity than Zn analogue. In this work, we synthesized the trinuclear aluminum salophen complex LAl₃ and the corresponding mononuclear complex LAl and performed live cell imaging experiments to reveal the ability of LAl₃ as a fluorescent probe, which highlights the importance of choice of metal ion in design of metal probes.

2. Results and Discussion

2.1. Synthesis and Characterization

Mononuclear salophen ligand and Al salophen complexes were synthesized according to the reported methods [35,36,37,38,39,40]. Salicylaldehyde and o-phenylenediamine were refluxed in ethanol for 16 h to give the orange salophen ligand. The reaction of salophen ligand with aluminum chloride in acetonitrile afforded the target mononuclear complex LAl. More importantly, we synthesized the trinuclear aluminum salophen complex LAl₃ to study the potential synergistic effect (Figure 1). Trinuclear cryptand-type salophen ligand was prepared according to the procedure reported by T. Nabeshima [41], and LAl₃ was obtained by stirring ligand and aluminum chloride in refluxing acetonitrile for 24 h. The axial chloride was replaced by fluoride to get LAl₃-F, which was used for crystal growth.

The three compounds were characterized by ¹H NMR, ESI-MS, and IR spectroscopies. ¹H-NMR spectra in DMSO-d6 show good resolution of the proton signals. Aromatic and imine protons in LAl₃ display nearly same chemical shifts with the mononuclear LAl, revealing the use of cryptand-like tri-salophen ligand has little effect to the conjugated system in the salophen moieties. The disappearance of an aromatic peak at δ = 7.68 and the emergence of a single peak at δ = 8.27 corresponding to the methenyl protons are strong proofs supporting the cryptand-type structure of LAl₃. IR spectra show characteristic C=N vibration (1625, 1620 cm⁻¹) of metal salophen complexes. The ESI mass spectra of the mononuclear and trinuclear Al salophen complexes show main peaks corresponding to m/z peak at 341.0866 and 1191.1954, respectively. The accurate exact mass and the successful assignment of ¹H NMR signals ensure that the cryptand structure of LAl₃ is right.

To further confirm the structure of LAl₃, we grew a single crystal (CCDC 1814298) suitable for X-ray structure analysis via the diffusion of N-hexane to a methanol/toluene mixed solution of LAl₃-F. The detailed parameters are listed in Table S1. As shown in Figure 2, each Al³⁺ shows octahedral coordination geometry, with a fluorinion and a methanol molecule as the axial ligand. Al³⁺ only 0.109 Å above the N₂O₂ plane and the angles between each two N₂O₂ planes are near 60°, indicating the C₃ symmetry of the molecule. Given that there is a symmetry plane perpendicular to the C₃ axial, the LAl₃ molecule belongs to C_3h point group. Each salophen moiety shows great distortion, with the angle between two phenyl planes being near 58.57°, larger than that in the crystal of mononuclear LAl reported by Darensbourg et al. [35]. The octahedral coordination geometry as well as the distortion of salophen moieties inhibit the aggregation of the molecules.

2.2. UV–Vis and Fluorescent Spectra

The absorption spectra of two Al salophen complexes in DMSO are shown in the Figure 3. The two main absorption bands at near 300 and 400 nm can be assigned to salophen-centered π–π* transitions. Different from the previously reported Zn salen complexes [17,21,31], no intramolecular charge-transfer transition bands between 500–600 nm is observed for there is no “D–π–A” structure in the salophen ligands. The absorption peaks of LAl₃ red-shifts ca. 20 nm compared to LAl. Besides, the extinction coefficient of LAl₃ is obviously larger than LAl because of the increasing number of the salophen chromophores in one molecule. (

Table 1. Photophysical properties of Al salophen complexes ¹.

Compound	λabs (log ε)/nm (log M−1cm−1) 2	λem/nm 3	τ/ns	Φ 4
LAl	308 (4.12), 337 (4.06), 387 (4.07)	500	0.58	0.054
LAl3	315 (4.76), 341 (4.67), 408 (4.66)	523	2.90	0.17

¹ All data was determined in DMSO. ² Wavelength of absorption peaks and the corresponding extinction coefficients. ³ Wavelength of fluorescence emission peaks. ⁴ Fluorescence quantum yield with the quinine sulfate (0.1 N H₂SO₄, Φ = 0.54) as standard, the uncertainty is ±10%.

)

Emission spectra were recorded to investigate the excited state properties of two Al salophen complexes. (Figure 3) LAl and LAl₃ show similar single emission bands with emission maxima at 500 and 523 nm, respectively. The large Stokes shift (ca. 115 nm) of LAl₃ is consistent with the previous report of cryptand-type Zn salophen complexes. Fluorescence lifetime of the two complexes were monitored at their emission maxima, and fluorescence quantum yields (±10%) were determined as 0.054 and 0.17 for LAl and LAl₃ in DMSO, using quinine sulfate (Φ_f = 0.54 in 0.1 N H₂SO₄) as standard. The longer lifetime and higher fluorescence quantum yield of LAl₃ can be attributed to its high structure rigidity. The fluorescence quantum yields of Al salophen complexes were only slightly larger than corresponding Zn complexes reported by us previously, indicating that similar closed-shell metal centers show little effect to the fluorescence quantum yield of salophen complexes.

2.3. Lipophilicity

Molecular lipophilicity is an important parameter affecting cellular uptake pathway and subcellular localization. To evaluate lipophilicity of the two complexes, we measured the N-octanol–water partition coefficient (log P,

Table 2. Octanol–water partition coefficients (log P) of Al salophen complexes.

Compound	Log P
LAl	0.10
LAl3	0.18

) according to Leo’s method [42]. Compared to LAl, trinuclear LAl₃ showed larger log P. This increase of the hydrophobicity might be due to the cage-like structure of the cryptand-type ligand, exposing the hydrophobic groups and protecting the hydrophilic metal centers inside the cage. Log P of Al complexes are significantly smaller compared to corresponding Zn complexes (log P > 1.0), indicating Al salophen complexes are much more hydrophilic, which might be due to the cationic nature of the Al salophen complexes.

2.4. Cell Imaging

In order to demonstrate different cellular behaviors of two Al salophen complexes, the cytotoxicity of the two compounds was evaluated by CCK-8 assay after 24 h incubation of the complexes with HeLa cells firstly (Figure 4). The result showed that in a concentration of 5 μM, the high cell viability up to 95% indicated low cytotoxicity. We chose this concentration for the subsequent imaging experiments.

Intracellular fluorescence of the two complexes in HeLa cells was studied by laser scanning confocal microscopy. As shown in Figure 5, incubation with 5 μM LAl or LAl₃ for 12 h led to green punctuate luminescence recorded in a perinuclear pattern. The corresponding Zn complexes showed almost invisible intracellular fluorescence under the same incubation condition, reflecting the relatively higher cellular uptake levels of Al salophens than that of Zn salophens. We ascribed this difference to the positive central charge of Al salophens, which allows Al salophens to traverse the negatively charged cellular membrane more easily than Zn salophens, driven by the membrane potential.

Then we examined the subcellular localization of LAl and LAl₃ by co-incubating the complex with a commercial lysosome tracker LysoTracker Red. As shown in Figure 5, the perinuclear punctuate green fluorescence of LAl and LAl₃ overlapped well with the red fluorescence of the lysosome tracker, with Pearson’s co-localization coefficients being ca. 0.85 and 0.96, respectively. This result indicated that Al salophen complexes were mainly distributed in lysosomal organelles following internalization.

To determine whether the cellular uptake mechanisms of LAl and LAl₃ is energy dependent or independent, we studied the effects of low temperature and ATP depletion in HeLa cells. Incubating cells under low temperature (4 °C) or in the presence of metabolic inhibitors blocks cellular uptake processes requiring energy supply. 2-deoxy-d-glucose and sodium azide, known as inhibitors of oxidative phosphorylation and glycolytic pathway respectively, having been widely used in cellular uptake mechanism studies [43,44,45]. As shown in Figure 6, no obvious intracellular luminescence of LAl or LAl₃ displayed after incubation under 4 °C, revealing that their uptake was greatly inhibited. When cells were pre-incubated with 10 mM sodium azide or 6 mM 2-deoxy-d-glucose to deplete the cellular ATP level, a smaller but still significant inhibition on the uptake of Al complexes was observed. The above temperature and ATP dependence of the cell uptake suggests that internalization of Al salophen complexes is energy-dependent. Little difference between LAl and LAl₃ is observed.

3. Materials and Methods

3.1. General Experimental Information

All solvents and chemicals were purchased from Alfa Aesar (Haverhill, MA, USA) and J&K (Beijing, China) and used without further purification, unless specifically mentioned. Cellular imaging trackers were purchased from Invitrogen (Life Technologies, Carlsbad, CA, USA). The ¹H NMR spectroscopic measurements were carried out using a Bruker-400 NMR spectrometer (Bruker, Billerica, MA, USA), at 400 MHz. Tetramethysilane (TMS) is used as the internal reference. The ¹⁹F NMR spectroscopic measurements were carried out using a Varian-400 NMR (Varian, Palo Alto, CA, USA). Electrospray ionization (ESI) mass spectra were performed on a Fourier transform ion cyclotron resonance mass spectrometer (FT-ICR, Bruker, Billerica, MA, USA). FTIR spectra were taken on a Nicolet iN10 MX Fourier transform infrared spectrometer (ThermoFisher, Carlsbad, CA, USA). Elemental analysis is performed on an Elementar Vario EL CUBE (Elementar, Langenselbold, Germany). The steady-state absorption spectra were attained on an Agilent 8453 UV–vis spectrophotometer (Agilent, Santa Clara, CA, USA) in 1 cm path length quartz cells. Single-photon luminescence spectra were recorded using fluorescence lifetime and steady state spectrophotometer (Edinburgh Instrument FLS920, Livingston, UK). Quantum yields of one photon emission of all the synthesized compounds were measured relative to the fluorescence of quinine sulfate in 0.1 N H₂SO₄. Confocal fluorescent images of living cells were performed using Nikon A1R-si laser scanning confocal microscope (Nikon, Tokyo, Japan), equipped with the laser of 405 nm.

3.2. Synthesis and Characterization

LAl was synthesized according to the reported methods. First, we synthesized the mononuclear salophen ligand. Salicylaldehyde (10 mmol, 10.4 mL) and o-phenylenediamine (5 mmol, 540.7 mg) were refluxing in 20 mL ethanol for 16 h, after cooling to the room temperature, the mixture was filtered to achieve the ligand (1.3 g, yield 85%). The ligand (158.2 mg, 0.5 mmol) and aluminum chloride (68 mg, 0.51 mmol) were refluxing in 20 mL acetonitrile solution for 16 h, then filtered and the precipitate was wash by CH₃CN to afford the aimed product (147 mg, yield: 78%). ¹H NMR (400 MHz, DMSO-d6): δ (ppm) 9.29 (2H, s), 8.22–8.07 (2H, m), 7.68 (2H, dd, J 7.8, 1.6), 7.61–7.45 (4H, m), 6.97 (2H, d, J 8.4), 6.85 (2H, t, J 7.4). HR MS (ESI⁺, DMSO, FT-ICR): m/z calcd. For C₂₀H₁₄AlN₂O₂ ([M–Cl]⁺) 341.0866, found 341.0865. FTIR (KBr pellete, cm⁻¹): 1625 (C=N), 1546 (Ar C=C), 1473 (Ar C=C), 1196 (C–O).

LAl₃: Cryptand-type salophen ligand was prepared according to the procedure reported by T. Nabeshima [41]. A mixture of 137.5 mg (0.14 mmol) of trinuclear salophen free base ligand and 58 mg (0.44 mmol) anhydrous aluminum chloride was dissolved in acetonitrile (20.0 mL) the solution was refluxed overnight. After cooling to the room temperature, the mixture was filtered and the solid was washed in turn by acetonitrile and diethyl ether, 1 mL each time, to remove the extra aluminum salts. After dried under reduced pressure, the product was obtained as yellow powder (110 mg, 68%). ¹H NMR (400 MHz, DMSO-d6): δ (ppm) 9.32 (6H, s), 8.27 (2H, s), 8.14 (6H, dd, J 5.9, 3.5), 7.52 (12H, m, J 6.7), 7.07 (6H, dd, J 7.0), 6.71 (6H, t). HR MS (ESI⁺, DMSO, FT-ICR): m/z calcd. For C₆₂H₃₈Al₃Cl₂N₆O₆ ([M–Cl + DMSO]⁺) 1191.1810, found 1191.1832. FTIR (KBr pellete, cm⁻¹): 1620 (C=N), 1551 (Ar C=C), 1442 (Ar C=C), 1393 (Ar C=C), 1199 (C–O). Anal. Calcd for C₆₂H₃₈N₆O₆Al₃Cl₃·33H₂O: C, 42.68; H, 6.01; N, 4.82. Found: C, 42.55; H, 4.60; N, 7.72. Anal. Calcd for C₆₂H₃₈Al₃Cl₃N₆O₆·CH₃CN·12H₂O: C, 54.61; H, 4.65; N, 6.97. Found: C, 54.72; H, 4.15; N, 6.94.

LAl₃-F: A mixture of 45.9 mg (0.040 mol) of LAl₃ and 4.6 mg (0.124 mol) NH₄F was dissolved in DMSO (10.0 mL), the solution stirred overnight at room temperature, then reduced pressure distillation. The precipitate was washed in turn by MeOH and diethyl ether, 1 mL each time, to remove the extra salts. After dried under reduced pressure, the product was obtained as yellow powder (30 mg, 68%). ¹H NMR (400 MHz, DMSO-d6): δ (ppm) 8.34 (6H, s), 7.34–7.17 (8 H, m), 7.05–6.77 (12 H, m), 6.30 (6H, dd, J 6.0), 6.12 (6H, t, J 7.6). HR MS (ESI⁺, DMSO, FT-ICR): m/z calcd. For C₆₂H₃₈Al₃Cl₂N₆O₆ ([M–F]⁺) 1081.2262, found 1081.2255.

3.3. Photophysical Properties

Quantum yields of emission of synthesized two compounds were measured with quinine sulfate (dissolved in 0.1 N H₂SO₄) as reference. The fluorescence measurements were performed in 1 cm quartz cells with 1 μM compound in DMSO on a fluorescence lifetime and steady state spectrophotometer (Edinburgh Instrument FLS920) equipped 450 W Xenon light, slits 2.5 × 2.5. The values of fluorescence quantum yield, Φ (sample), were calculated according to the equation

$${\mathsf{\Phi }}_{sample}={\mathsf{\Phi }}_{ref}\frac{O{D}_{ref}\bullet I_{sample}\bullet d_{sample}^2}{O{D}_{sample}\bullet I_{ref}\bullet d_{ref}^2}$$

• Φ_ref: The values of fluorescence quantum yield of the reference.
• I: integrated emission intensity.
• OD: optical density at the excitation wavelength.
• d: the refractive index of solvents. d_DMSO = 1.478, d_H2O = 1.333.

3.4. Determination of the Octanol–Water Partition Coefficients (Log P)

Equal volume (200 mL) of N-octanol and water were thoroughly mixed by an oscillator and separated after 24 h. Two compounds (1 mg each) were then dissolved in 40 mL of the separated N-octanol and the solution was allowed to equilibrate for further 24 h. The extinction coefficient was then calculated and 40 mL of water (previously separated from the mixture) was added. The new octanol–water system was allowed to equilibrate for additional 24 h. After separating, both fractions were analyzed by UV–vis spectra. The log P values were calculated by

$$\log P=\log \frac{C_{octanol}}{C_{water}}$$

where C_octanol and C_water refer to the concentration of two compounds in the N-octanol and water, respectively.

3.5. Live Cell Imaging

All cells were incubated in complete medium (Dulbecco’s modified Eagle’s medium, supplemented with 10% fetal bovine serum (FBS) and 1% penicillin-streptomycin) at 37 °C in atmosphere containing 5% CO₂. For imaging, HeLa cells were grown in poly-d-lysine-coated dishes and incubated in 2 mL of complete medium for 24 h. Cells were washed with PBS, and stocked dyes (1 mM in DMSO) were added to obtain a final concentration of 5 μM. The treated cells were incubated for 12 h at 37 °C. A few minutes prior to confocal imaging cells were washed twice with PBS. A confocal laser scanning microscope (A1R-si, Nikon, Tokyo, Japan) was used to obtain images. Cells were imaged via the fluorescence mode with a 60× immersion lens with the following parameters: laser power 100%, pinhole 4.0 a.u., excitation wavelength 405 nm, detector slit 500–530 nm, resolution 1024 × 1024, and a scan speed 0.5 frames per second.

HeLa cells were placed onto 0.1 mM poly-d-lysine coated glasses in complete media and the cells were incubated for 24 h. A stock solution of salophen complexes in chromatographic grade, anhydrous DMSO was prepared as 1 mM. The solution was diluted to a final concentration of 5 μM by complete growth medium and coincubated with HeLa cells for 12 h. Stock solutions of LysoTracker Red was prepared as 1 mM, and the stock solution was diluted to the working concentrations in complete medium (1 μM). After incubating for half an hour, cells were washed with PBS buffer twice before confocal experiments. Differential interference contrast (DIC) and fluorescent images were processed and analyzed using Image J (version 1.8.0).

For the cellular uptake experiment, the cells were incubated with LAl or LAl₃ at 37 °C for 2 h, while cells at 4 °C were treated with precooled complete growth medium containing 5 μM LAl or LAl₃, respectively. For the ATP depletion groups, HeLa cells were preincubated with 10 mM sodium azide or 6 mM 2-deoxy-d-glucose for 1 h, then treated with 10 mM sodium azide or 6 mM 2-deoxy-d-glucose complete growth medium containing 5 μM LAl or LAl₃ for 2 h.

Hela cells were seeded in flat-bottomed 96-well plates, 10⁴ cells per well, with 200 μL complete culture media in the dark for 24 h. After being washed with PBS three times (200 μL × 3), the cells were incubated with 5 μM concentrations of the studied salophens for another 24 h in the dark while untreated-cells and wells containing no cells are set as the controls. HeLa cells were then washed with PBS three times (200 μL × 3). 10 μL Cell Counting Kit-8 (CCK-8) solution and 90 μL PBS were added per well. After 2 h, the absorbance at 450 nm was read by 96-well plate reader. The viability of Hela cells was calculated by

$$\mathrm{CV}=\frac{A_s-A_b}{A_c-A_b}\times 100\%$$

CV stands for the viability of cells, A_s, A_c, and A_b stand for the absorbance of cells containing 2-Glu, cell control (0 μΜ 2-Glu), and blank control (wells containing neither cells nor 2-Glu).

4. Conclusions

In summary, we reported the synthesis and characterization of the cryptand-type Al salophen complex LAl₃. The cryptand-type structure was confirmed by the X-ray crystal diffraction. LAl₃ showed green fluorescence emission similar to the mononuclear LAl, and can be internalized into HeLa cells by energy-dependent pathways. It could be used as lysosome localized fluorescent probes for live cell imaging, which was superior to their Zn analogues.