|Home | About | Journals | Submit | Contact Us | Français|
An ideal fluorescent dye for staining cell organelles should have multiple properties including specificity, stability, biocompatibility and a large Stokes shift. Tuneable photophysical properties enable 1,8-naphthalimide to be an excellent fluorophore in biomedical applications. Many naphthalimide derivatives have been developed into drugs, sensors and other dyes. In this study, a series of 1,8-naphthalimide derivatives targeting live cell mitochondria were synthesized. Among these probes, Mt-4 was characterized as the best one with highly specific mitochondrial localization, low cytotoxicity and large Stokes shift. More importantly, Mt-4 stood out as a potential mitochondrial dye for living cell experiments with induced mitochondrial stress from treatments since Mt-4 has enhanced fluorescence while undergoing mitochondrial stress.
A series of 1,8-naphthalimide derivatives targeting live cell mitochondria were synthesized. Among these probes, Mt-4 is the best one with highly specific mitochondrial localization, low cytotoxicity and large Stokes shift. It is a potential mitochondrial dye for living cell experiments with induced mitochondrial stress from treatments.
Mitochondria play a central role in energy metabolism through oxidative phosphorylation (OXPHOS) and lipid oxidation within eukaryotic cells. They are also involved in many cell signaling pathways including apoptosis. The morphology and abundance of mitochondria varies highly with cell type, proliferative state and other metabolic requirements. Mutation in mitochondrial proteins encoded by mitochondrial DNA or nuclear DNA will cause defects in mitochondrial function, which will result in serious diseases and developmental retardation. Bound by a double membrane, mitochondria are divided into two distinct parts: the intermembrane space and the mitochondrial matrix. In a healthy cell, a relatively constant membrane potential of around ca. −180 mV across the mitochondrial inner membrane is maintained by oxidative phosphorylation and ion channels. Based on these physical characteristics of mitochondria, many fluorescent dyes and probes have been developed to specifically stain this organelle. One of the widely used carriers for such probes is triphenyl phosphonium (TPP), a lipophilic cation with an excellent ability to pass through the mitochondrial lipid bilayer membrane.
1,8-naphthalimide derivatives are strong fluorescent compounds with a large Stokes shift. They are attractive fluorophores for the development of bioimaging probes with low auto-fluorescence and low light scattering from biological environments. Our lab has developed a series of fluorescent biosensors, including 1,8-naphthalimide derivatives as intracellular and extracellular pH sensors. In this study, a series of naphthalimide derivatives (Mt-dyes) were synthesized and functionalized with different mitochondria targeting groups, which are TPP, pyridine and 4-dimethyaminopyridine (DMAP), respectively (Scheme 1).
Mt-dyes have low solubility in an aqueous condition (PBS buffer) but are completely soluble in a polar solvent, such as DMSO. To characterize the photophysical properties of Mt-dyes, the compounds were dissolved in DMSO and diluted with PBS buffer by 1:500. The typical absorption and emission spectra were collected from compound Mt-4. As shown in Figure 1, an absorption maximum peak was seen at 408 nm, while an emission maximum was detected at 525 nm. There is a large Stokes shift of 117 nm with a very narrow overlap between the absorption and the emission peak, which is one of the highly desired characteristics for fluorescence imaging.
Cytotoxicity of these mitochondrial dyes derived from 1,8-naphthalimide was assessed by MTT (3-[4,5–dimethylthiazol-2-yl]-2,5 diphenyltetrazoliumbromide) assay, which determines mitochondrial activity. The HeLa cell line was used as a typical cell line to test toxicity. Cells were incubated with 1.0 µM mitochondrial dyes for 3, 6, 12 and 24 hours, respectively. Compared to negative control (DMSO solvent only) and other derivatives, Mt-1 and Mt-5, both of which have six carbon chains and TPP as the targeting group, showed strong cytotoxicity in 12 hour and longer incubation, while the Mt-4 did not show much toxicity to cells (Figure 2). The only difference between Mt-4 and Mt-5 is the length of carbon chain connecting the fluorophore and mitochondrial targeting group TPP. The mechanism of the toxicity will be studied further in future research.
The mitochondrial localization of these compounds was characterized with HeLa cells (Figure 3 and S2 Figure). All of the Mt-1, Mt-4 and Mt-5 dyes, which have TPP as the mitochondrial targeting group, showed high colocalization with MitoTracker Red® FM (Life Technologies, MitoRed, the same below) with Pearson’s correlation coefficient (PCC) about 0.9. The other compounds (Mt-2, 3, 6 and 7) showed low colocalization with MitoRed, i.e. Pearson’s correlation is less than 0.7. The derivative without mitochondria targeting group, compound 5b, aggregated in cells and could not specifically stain mitochondria (S2 Figure). Considering the mitochondrial localization and cytotoxicity of these mitochondrial dyes, Mt-4 was selected as the best candidate for further study.
During the co-staining of Mt-4 and MitoRed, one interesting phenomena was observed. After Mt-4 (1.0 µM) was added into the medium with cells, which were first stained with MitoRed (0.2 µM), a quick decrease of fluorescence from MitoRed (Em650nm) was detected (Figure 4A, D and S3 Figure A). The green fluorescence from Mt-4 was saturated in five minutes. On the contrary, if cells were stained with Mt-4 first, the red fluorescence from MitoRed was saturated for about ten minutes after highly concentrated MitoRed (final concentration is 1.0 µM) was added, while only ~10% decrease of green fluorescence from pre-stained Mt-4 was observed (Figure 4B, D and S3 Figure B). As controls, cells were stained only using either Mt-4 (1.0 µM) or MitoRed (0.2 µM), no obvious fluorescence intensity decrease was observed during the same experimental time frame and similar experimental conditions (Figure 4C and S4 Figure).
In this study, HeLa cells were stained with either MitoRed or Mt-4 before cells were treated with 10 µM of CCCP. As shown in Figure 5, there is about a 25% signal loss from MitoRed in one hour of CCCP treatment. On the contrary, the fluorescence of Mt-4 increased about 40% in the first 40 minutes of treatment followed by slowly recovering with the longer incubation (Figure 5A). According to the microscopy images, except for the morphological changes of the mitochondria, no obvious mitochondrial dyes leakage was observed in HeLa cells stained by either Mt-4 or MitoRed (Figure 5B). The possible reason for the fast increase of Mt-4 fluorescence after CCCP treatment is the fragmentation of mitochondria caused by treatments, which may change the microenvironmental distribution of Mt-4.
With the need for increasing the sensitivity of biochemical assays, fluorescence is one of the most powerful tools which has been widely used in biomedical imaging and other related fields. An ideal fluorophore should have a large Stokes shift. Otherwise, it can cause self-quenching and false positives caused by excitation light and scattered light. Despite a cornucopia of options which are commercially available, researchers still have been spending much of their efforts towards the development of new dyes with specific excitation or emission requirements.
In this study, eight compounds were biologically characterized and only Mt-4 stood out as the best candidate with good mitochondrial localization (Figure 3) and biocompatible properties (Figure 2). The data collected from a co-staining assay (Figure 4 and S4 Figure) indicates that the fluorescence signal loss of the MtRed is not due to photobleaching during microscopy imaging. The possible “quenching effect” of MitoRed by Mt-4, which happens between mitochondrial dyes and some mitochondria targeting compounds, was also excluded by an in vitro assay. Various concentrations of the mixtures with Mt-4 and MitoRed were mixed in DMSO for the measurements of their emission properties. Results showed that the concentrations of MitoRed or Mt-4 did not affect the fluorescence intensities of the other dye, i.e. Mt-4 or MitoRed (S5 Figure). However, solvents with different polarities affect the fluorescent emission of Mt-4 and MitoRed in different patterns (S6 Figure). Therefore, the possible reason for the quick decay of fluorescence from MitoRed by Mt-4 is that the local microenvironmental change of MitoRed is caused by the fast mitochondrial accumulation of Mt-4, while accumulation of MitoRed could not affect the distribution of Mt-4 in the same way.
As mentioned above, the driving force for Mt-4 with a positive charge targeting mitochondria is the membrane potential of ca. −180 mV across the mitochondrial inner membrane. But this value and morphology of mitochondria are dynamically changed when cells are under stress. So the stability of fluorescence from dyes staining mitochondria is critical to the live cell imaging under treatments causing vibration of mitochondrial membrane potential (MMP). Carbonyl cyanide m-chlorophenylhydrazone (CCCP) is traditionally used as an OXPHOS uncoupling agent working as a proton transmembrane carrier which results in the quick reduction of mitochondrial inner membrane potential. It has been reported that the MitoRed lost its specificity to mitochondria when the cells were pre-treated with 10 µM of CCCP and the sensitivity and specificity to mitochondria through the observation of a decrease in fluorescence intensities. It was reported that CCCP leads to the fragmentation of mitochondria and inhibits the mitochondrial fusion, with the loss of membrane potential it induces. Similar to the decrease of MitoRed fluorescence during CCCP treatment, we think the increase of fluorescence from Mt-4 is the result of the microenvironmental changes of dyes induced by CCCP in a unknown mechanism.
In summary, a series of naphthalimide derived mitochondrial dyes were synthesized and tested. Mt-4 has been identified as the best one among the compounds we prepared due to its large Stokes shifts, low toxicity, and unique mitochondrial staining properties. It is expected to be a useful imaging dye for mitochondria-related research either at the bulk cell level or single cell level.
All chemicals were purchased and used without further purification. MitoTracker® Red FM was ordered from Life Technologies (Carlsbad, CA). N-(2-hydroxyethyl)-4-bromine-1,8-naphthalimide (Figure 1, compound 1a, n=2) and N-(6-hydroxyhexyl)-4-bromine-1,8-naphthalimide (Figure 1, compound 1b, n=6), 9-(2-hydroxyethyl)-4-(N-piperidino)-1,8-naphthalimide (compound 2a, n=2) and 9-(6-hydroxyhexyl)-4-(N-piperidino)-1,8-naphthalimide (compound 2b, n=6) were prepared according to known procedure[8a, 19].
A Varian liquid-state NMR operated at 400 MHz was used for 1H and 13C NMR spectra measurements. An Applied Biosystems DE-STR MALDI-TOF Mass Spec was used to measure the mass spectra of the intermediates and the final product. A Shimadzu UV-3600 UV-Vis-NIR spectrophotometer (Shimadzu Scientific Instruments, Columbia, MD) was used for UV-Vis absorption spectra measurements. A Shimadzu RF-5301 spectrofluorophotometer was used for fluorescence measurements. Confocal microscope (Nikon, TE2000E, Melville, NY) was used for cell imaging.
HeLa cells (American Type Culture Collection, ATCC, Manassas, VA) were cultured in Eagle’s minimum essential medium (EMEM) supplemented with 10% fetal bovine serum, and incubated at 37 °C in 5% CO2 atmosphere. Cells were then seeded onto 96 well plates at 10,000 cells per well, and incubated for 1 day at 37 °C. For the colocalization assay, cells were pre-incubated with MitoTracker® Red FM for half hours. Mitochondrial dyes were dissolved into DMSO and added to medium to get the final concentration of dyes at 1.0 µM. Confocal images were taken five minutes after adding the mitochondrial dyes. For the kinetic imaging assay, confocal imaging was started immediately when the dyes were added.
For the cell toxicity assay, cells were incubated with the sensor for 6, 12 and 24 hours before the assay was performed with the cytotoxicity MTT kit from Promega (CellTiter 96® Non-Radioactive Cell Proliferation Assay) following the detailed procedure from the company.
Mitochondrial dyes were synthesized according to the procedure shown in Scheme 1. The crude products were purified by silica gel column chromatography. The structures of compounds were confirmed by 1H and 13C NMR spectroscopy and mass spectra. The details of the synthesis of each compound are given in the results part.
325 mg (1 mmol) of 9-(2-Hydroxyethyl)-4-(N-piperidino)-1,8-naphthalimide (compound 2a, n=2) was dissolved into 20 mL of ice-cold anhydrous tetrahydrofuran (THF) containing 303.6 mg of trimethylamine (Et3N). After dropwise addition of 10 mL of anhydrous THF containing 229 mg (2 mmol) of methanesulfonyl chloride, the reaction mixture was vigorously stirred overnight at room temperature. Solvent was removed under reduced pressure. The crude product was dissolved into 100 mL of dichloromethane and was washed with water three times. The organic layer was dried with sodium sulfate and the solvent was removed under vacuum. Without further purification, the product (compound 3a, n=2) was dissolved into 10 mL of anhydrous acetonitrile with 310 mg (3 mmol) sodium bromide (NaBr). The mixture was incubated at room temperature under vigorous stirring for 24 hours. 100 mL of dichloromethane was added to reaction mixture followed by washing two times with water and washing one time with brine. The organic phase was dried by sodium sulfate and the solvent was removed under reduced pressure. The product was purified by silica gel chromatography using 85:15 of hexane-ethyl acetate to get 272 mg of 2-(2-bromoethyl)-6-(piperidin-1-yl)-1H-benzo[de]isoquinoline-1,3(2H)-dione (compound 5a, n=2), yield 70%. 1H NMR (400 MHz, CDCl3, δ ppm): 1.73 (m, 2H), 1.86 (m, 4H), 3.24(t, 4H), 3.63 (t, 2H), 4.57 (t, 2H), 7.14 (d, 1H), 7.65 (t, 1H), 8.39 (d, 1H), 8.50 (d, 1H), 8.58 (d, 1H).
195 mg (0.5 mmol) of compound 5a (n=2) was dissolved into 10mL ethanol with 787 mg (3 mmol) of triphenylphosphine (TPP). The mixture was refluxed for 24 hours. The solvent was removed by vacuum and the product was purified by silica gel chromatography using 93:7 dichloromethane-methane as eluent to get 168 mg orange product, yield 52%. 1H NMR (400MHz, CDCl3, δ ppm): 1.72 (m, 2H), 1.86 (m, 4H), 3.24(t, 4H), 4.10 (m, 2H), 4.60 (t, 2H), 7.11 (d, 1H), 7.62–7.72 (m, 10H), 7.83–7.88(m, 6H), 8.35–8.37 (m, 2H), 8.43 (d, 1H). 13C NMR (100 MHz, CDCl3, δ ppm): 164.4, 163.8, 158.2, 135.4, 133.8, 133.7, 133.5, 131.7, 130.8, 130.7, 130.2, 126.2, 125.5, 122.3, 118.3, 117.5, 114.8, 114.7, 110.2, 54.7, 34.5, 26.3, 24.4, 22.3, 21.9) (S1 Fig). MS (MOLDI-TOF): m/z calcd C37H34N2O2P+ (M)+ 569.66, found 569.62.
381 mg (1 mmol) of 9-(6-Hydroxyhexyl)-4-(N-piperidino)-1,8-naphthalimide (Scheme 1, compound 2b, n=6) was dissolved into 20 mL of ice-cold anhydrous tetrahydrofuran (THF) containing 303.6 mg of trimethylamine (Et3N). After dropwise addition of 10 mL of anhydrous THF containing 229 mg (2 mmol) of methanesulfonyl chloride, the reaction mixture was vigorously stirred overnight at room temperature. Solvent was removed under reduced pressure. The crude product was dissolved into 100 mL of dichloromethane and was washed with water for three times. The organic phase was dried with sodium sulfate and the solvent was removed under vacuum. Without further purification, the product (compound 3, n=3) was dissolved into 30 mL of anhydrous acetonitrile with 1.32 g (5 mmol) of TPP. The mixture was refluxed for 24 hours. After removal of solvent by vacuum rotary evaporation, the product was purified by silica gel chromatography using 93:7 dichloromethane-methane as eluent to get 210 mg orange product, yield 29.1%. 1H NMR (CDCl3, δ ppm): 1.72 (m, 2H), 1.86 (m, 4H), 2.71 (s, 3H), 3.21(t, 4H), 3.77 (m, 2H), 4.06 (t, 4H), 7.13 (d, 1H), 7.23(d, 1H), 7.62–7.70 (m, 7H), 7.74–7.85(m, 9H), 8.35 (d, 1H), 8.43 (d, 1H), 8.49 (d, 1H). MS (MOLDI-TOF): m/z calcd C41H42N2O2P+ (M)+ 625.77, found 625.72.
By a procedure similar to the synthesis of Mt-1, 381 mg (1 mmol) of compound 2b with 6 carbons chain (n=6) and 367 mg (3 mmol, 3 times of starting compound 2b) of 4-dimethylaminopyridine (DMAP) were applied to the reaction to produce 260 mg pure product, yield 45%. 1H NMR (CDCl3, δ ppm): 1.72 (m, 2H), 1.86 (m, 4H), 2.71 (s, 3H), 3.20(s, 4H), 3.22(t, 6H), 4.08 (m, 2H), 4.22 (t, 2H), 6.97 (t, 2H), 7.12 (d, 1H), 7.23(d, 1H), 8.30–8.35(m, 3H), 8.43 (d, 1H), 8.50 (d, 1H). MS (MOLDI-TOF): m/z calcd C30H37N4O2+ (M)+ 485.65, found 485.60.
By a procedure similar to the synthesis of Mt-2, 381 mg (1 mmol) of compound 2b with 6 carbons chain (n=6) and 237 mg (3 mmol, 3 times of starting compound 2b) of pyridine were applied to the reaction to produce 226 mg pure product, yield 40%. 1H NMR (CDCl3, δ ppm): 1.71 (m, 2H), 1.84 (m, 4H), 2.07 (t, 2H), 2.71 (s, 3H), 3.20(t, 4H), 3.22(s, 6H), 4.07 (t, 2H), 5.01 (t, 2H), 7.12 (d, 1H), 7.64(t, 1H), 8.10(t, 2H), 8.34 (d, 1H), 8.47–8.50 (m, 2H). MS (MOLDI-TOF): m/z calcd C28H32N3O2+ (M)+ 442.58, found 442.51.
By a procedure similar to the synthesis of Mt-4, 760 mg (2 mmol) of 9-(2-Hydroxyhexyl)-4-(N-piperidino)-1,8-naphthalimide (compound 2b, n=6) and 460 mg (4 mmol) of methanesulfonyl chloride were used to synthesize compound 5b (scheme 1, n=6), 2-(6-bromohexyl)-6-(piperidin-1-yl)-1H-benzo[de]isoquinoline-1,3(2H)-dione to produce 650 mg of pure product in yellow powder state, yield 74%. 1H NMR (CDCl3, δ ppm): 1.71 (m, 2H), 1.85 (m, 4H), 3.21(t, 4H), 3.83(t, 2H), 4.15 (t, 2H), 7.12 (d, 1H), 7.64(t, 1H), 8.38 (d, 1H), 8.50 (d, 1H), 8.58 (d, 1H).
221 mg (0.5 mmol) of compound 5b and 787 mg (3 mmol) of TPP were used to synthesize Mt-5 and got 230 mg of pure product, yield 65%. 1H NMR (CDCl3, δ ppm): 1.72 (m, 2H), 1.86 (m, 4H), 3.21(t, 4H), 3.77 (m, 2H), 4.06 (t, 4H), 7.13 (d, 1H), 7.23(d, 1H), 7.62–7.70 (m, 7H), 7.74–7.85(m, 9H), 8.35 (d, 1H), 8.43 (d, 1H), 8.49 (d, 1H). MS (MOLDI-TOF): m/z calcd C41H42N2O2P+ (M)+ 625.77, found 625.71.
By a procedure similar to the synthesis of Mt-5, 221 mg (0.5 mmol) of compound 5b and 183 mg (1.5 mmol) of DMAP were used to synthesize Mt-6 to produce 249 mg of pure product, yield 88%. 1H NMR (CDCl3, δ ppm): 1.72 (m, 2H), 1.86 (m, 4H), 3.20(s, 4H), 3.22(t, 6H), 4.08 (m, 2H), 4.22 (t, 2H), 6.97 (t, 2H), 7.12 (d, 1H), 7.23(d, 1H), 8.30–8.35(m, 3H), 8.43 (d, 1H), 8.50 (d, 1H). MS (MOLDI-TOF): m/z calcd C30H37N4O2+ (M)+ 485.65, found 485.60.
By a procedure similar to the synthesis of Mt-5, 221 mg (0.5 mmol) of compound 5b and 120 mg (1.5 mmol) of pyridine were used to synthesize Mt-7 to produce 175 mg of pure product, yield 67%. 1H NMR (CDCl3, δ ppm): 1.71 (m, 2H), 1.84 (m, 4H), 2.07 (t, 2H), 3.20(t, 4H), 3.22(s, 6H), 4.07 (t, 2H), 5.01 (t, 2H), 7.12 (d, 1H), 7.64(t, 1H), 8.10(t, 2H), 8.34 (d, 1H), 8.47–8.50 (m, 2H). MS (MOLDI-TOF): m/z calcd C28H32N3O2+ (M)+ 442.58, found 442.50.
This work was supported by the NIH National Human Genome Research Institute (NHGRI), Centers of Excellence in Genomic Science, grant number 5 P50 HG002360, and the NIH Common Fund LINCS program, grant number 5 U01 CA164250 (Professor Deirdre R. Meldrum, PI).