Article pubs.acs.org/ac
Spatially Dependent Fluorescent Probe for Detecting Different Situations of Mitochondrial Membrane Potential Conveniently and Efficiently Xuechen Li,† Minggang Tian,† Ge Zhang,† Ruoyao Zhang,† Ruiqing Feng,† Lifang Guo,† Xiaoqiang Yu,*,† Ning Zhao,*,‡ and Xiuquan He*,§
Anal. Chem. 2017.89:3335-3344. Downloaded from pubs.acs.org by IOWA STATE UNIV on 01/09/19. For personal use only.
†
Center of Bio and Micro/Nano Functional Materials, State Key Laboratory of Crystal Materials, Shandong University, Jinan 250100, Shandong, People’s Republic of China ‡ Shandong Key Laboratory for Adhesive Materials, Advanced Materials Institute, Shandong Academy of Sciences, Jinan 250014, Shandong, People’s Republic of China § Department of Anatomy, Shandong University School of Medicine, Jinan 250012, Shandong, People’s Republic of China S Supporting Information *
ABSTRACT: The feedback from mitochondrial membrane potential (MMP) in different situations (normal, decreasing, and vanishing) can reflect different cellular status, which can be applied in biomedical research and diagnosis of the related diseases. Thus, the efficient and convenient detection for MMP in different situations is particularly important, yet the operations of current fluorescent probes are complex. In order to address this concern, we presented herein a spatially dependent fluorescent probe composed of organic cationic salt. The experimental results from normal and immortalized cells showed that it could accumulate in mitochondria selectively when MMP was normal. Also, it would move into the nucleus from mitochondria gradually with the decrease of MMP, and finally it targeted the nucleus exclusively when MMP vanished. According to the cell morphology, there is a straightforward spatial boundary between the nucleus and cytoplasm where mitochondria locate; thus, the three situations of MMP can be point-to-point indicated just by fluorescence images of the probe: that all probes accumulate in mitochondria corresponds to normal MMP; that probes locate both in the mitochondria and nucleus corresponds to decreasing MMP; that probes only target the nucleus corresponds to vanishing MMP. It is worth noting that counterstaining results with S-11348 indicated that the spatially dependent probe could be applied to distinguishing dead from viable cells in the same cell population. Compared with the commercial Cellstain-Double staining kit containing calcein-AM and propidium iodide (PI), this probe can address this concern by itself and shorten the testing time, which brings enormous convenience for relevant researches.
M
quenching (ACQ) character;3,8,11 they can intensively accumulate in mitochondria with normal MMP and emit weak fluorescence due to the quenching effect from higher concentration. When MMP decreases, parts of probes release to the cytosol and the others still remain in the mitochondria, while these probes at low concentrations emit stronger fluorescence. When MMP vanishes, all the probes release to the cytosol and keep low concentration, and fluorescence intensity is similar to one in a decreasing situation. Therefore, it is difficult to distinguish MMP decreasing from vanishing just by the fluorescence image. (2) Fluorescent dichromatic probes, such as JC-1;6,7,11 it can emit red fluorescence in aggregations or green fluorescence in monomer when MMP is high or low, respectively. So it can be used to judge the variation trend of MMP by comparing the changes of JC-1 fluorescent color.
itochondrial membrane potential (MMP) is a key parameter representing mitochondrial functions, and its decrease can imply the disruption of mitochondrial electron transport chain, which results in cellular dysfunction or even death.1,2 So, the feedback from different situations of MMP, such as normal, decreasing, and vanishing, is practically useful in biomedical research and diagnosis of the related diseases.3−6 Generally, MMP of healthy cells is normal; MMP decreasing indicates the occurrence of dysfunctions, for example, the decrease of sperm MMP indicates its activity reduce;7 MMP vanishing shows that cells are in disease status, for instance, MMP vanished in some parasites killed by artemisinin.8 Thus, developing a convenient method to efficiently reveal different situations of MMP is of great importance in biology and medicine. Recently, some fluorescent probes to detect the MMP have been studied extensively.9−11 To date, there are mainly two kinds of probes: (1) fluorescence intensity probes, such as rhodamine 123, TMRE, and TMRM, with aggregation-caused © 2017 American Chemical Society
Received: September 29, 2016 Accepted: February 14, 2017 Published: February 14, 2017 3335
DOI: 10.1021/acs.analchem.6b03842 Anal. Chem. 2017, 89, 3335−3344
Article
Analytical Chemistry However, the staining concentration needs to be adjusted repeatedly before every experiment: it is difficult to form Jaggregates when using too low staining concentrations, and high concentrations may produce block precipitates due to its very poor water solubility, which brings great inconvenience.7,11 Thus, a fluorescent probe which can easily and efficiently show normal, decreasing, and vanishing status of MMP is lacking, yet valuable. As mentioned above, rhodamine 123 releases to the cytosol when MMP decreases. Because the boundary between the mitochondria and cytosol is ambiguous, one cannot judge whether some probes have released to the cytosol or not just by fluorescence images. Inspired by this phenomenon and considering there is the straightforward spatial boundary between the nucleus and cytosol, we conceive such a probe with characters as follows: it can accumulate in mitochondria selectively when MMP is normal; also, it will move into the nucleus from the mitochondria gradually with the decrease of MMP and finally target the nucleus exclusively when MMP vanishes. If so, the three situations of MMP will have three point-to-point fluorescence images: that all probes accumulate in mitochondria corresponds to normal MMP; that probes locate both in the mitochondria and nucleus corresponds to decreasing MMP; that probes only target the nucleus corresponds to vanishing MMP. In order to attain the objective, it should be rational to design such a probe based on organic cationic compounds (OCCs), because some OCCs can accumulate in mitochondria with normal MMP.12−14 On the other hand, some other OCCs just locate in the nucleus due to its static interaction with DNA as well as the inserted layer effect with DNA grooves.15,16 Thus, we believe that, by balancing the interaction of OCC between MMP and DNA to finely control its targeting ability, a spatially dependent probe can be obtained which meets the requirements above, but it is still challenging. Previously, our group has reported two OCCs probes: CAI17 and 9E-BMVC.18 More importantly, the two molecules shared the same fluorophore, although they exclusively label the mitochondria (CAI) and nucleus (9E-BMVC), respectively. From Figure 1, the n-hexyl endows CAI with appropriate hydrophobic ionic radius (HIR),13 so CAI can promptly penetrate the mitochondrial inner membrane and accumulate in mitochondria preferentially. Compared with CAI, 9E-BMVC has an additional cationic pyridine salt to increase the static interaction with DNA and its enlarged conjugation plane enhances the inset effect with DNA grooves. In addition, the methyl in 9E-BMVC reduces its HIR, which increases its difficulty in entering the mitochondrial inner membrane. Most importantly, the calculation results of a DNA titration experiment showed that the intrinsic binding constant (k) to DNA of 9E-BMVC was 6.46 × 106 M−1 while that of CAI was almost close to zero (Figure S1). Inspired by this, a new OCC probe LAD with spatially dependent character was designed and synthesized (Figure 1b) by balancing the structure of CAI and 9E-BMVC. LAD remained the appropriate HIR using butyl and added a planarizable pyridine group which increased the inserting effect with DNA grooves. In addition, the experiment results showed that the k value of LAD was 8.30 × 105 M−1, which was appropriately less than that of 9E-BMVC but larger than that of CAI. A series of experimental results demonstrated that LAD accumulated in mitochondria when MMP was normal, immigrated to the nucleus gradually with MMP decreasing, and located in the nucleus completely when MMP vanished.
Figure 1. Structures of CAI, LAD, LAD-1, and 9E-BMVC (a); synthetic routine of LAD and LAD-1 (b).
Moreover, LAD has been successfully applied to distinguishing dead cells without MMP from viable ones with normal MMP in the same cell population, which has great significance in biology and medicine. Also, the mechanism of LAD was discussed in detail and provided a theoretical guidance for the development of this type of probe. Recently, Moritomo et al. reported a probe (BP6) with similar properties.19 However, the staining time is up to 12−16 h due to the very low membrane permeability of BP6, which greatly limits its application in convenient and efficient detection of MMP. Comparatively, the incubating time of LAD is only 15 min. Moreover, as a spatially dependent fluorescent probe detecting different situations of MMP, the action mechanism of LAD has been explained in detail. Thus, LAD-1, a homologous compound of LAD, has been synthesized, which is also a spatially dependent fluorescent probe for detecting MMP. LAD-1 needs longer time to enter cells in comparison with LAD, mainly due to the weaker lipotropy of the methyl in LAD-1 than the butyl in LAD. Very importantly, according to our experimental results in living HeLa cells, the sensitivity of LAD based on the spatially dependent mechanism to MMP changes is higher than that of JC-1 on fluorescence colors, indicating that a spatially dependent fluorescent probe has huge potential in detecting MMP. Yet, relative researches have just started.
■
EXPERIMENTAL SECTION Synthesis. The details of the synthesis of LAD have been given in the Supporting Information. LAD was synthesized by the Knoevenagel condensation reaction between 1-butyl-4methylpyridin-1-ium iodide and (E)-9-(2-ethoxymethyl)-6-(2(pyridine-4-yl) vinyl)-9H-carbazole-3-carbaldehyde using piperidine as the catalyst. Apparatus and General Methods. The UV−vis−near-IR absorption spectra of dilute solutions were recorded on a Hitachi U-2910 spectrophotometer using a quartz cuvette 3336
DOI: 10.1021/acs.analchem.6b03842 Anal. Chem. 2017, 89, 3335−3344
Article
Analytical Chemistry
diluted in DMEM at a concentration of 2 μM was added into the wells of the treatment group, and DMSO (200 μL/well) diluted in DMEM at final concentration of 0.2% to the negative control group, respectively. The cells in one of the 96-well plates were incubated with 2 μM LAD for 2, 8, and 24 h at 37 °C under 5% CO2, and the cells in another 96-well plate were incubated with 2, 5, 10 μM LAD for 2 h at 37 °C under 5% CO2, then MTT (5 mg/mL in PBS) was added into each well. After 4 h of incubation, the culture medium in each well was removed and DMSO (200 μL) was added to dissolve the purple crystals. After 20 min of incubation, the optical density readings at 492 nm were taken using a plate reader. Each individual cytotoxicity experiment was repeated for three times. In addition, for the study of the cytotoxicity of LAD on the viability of cells in the mitochondrial level, the living SiHa cells were first incubated with LAD (2 μM, 15 min) and then were placed for a period of time (2, 5, and 10 h). Thereafter, these cells were stained with 0.2 μM MTR for 30 min, respectively. Then, these cells were imaged immediately, after rinsing with PBS twice. For the sensitivity assay, the living HeLa cells were incubated with 20 μg/mL of JC-1 first for 30 min, and then the cells were washed to remove the unbound probes. After rinsing with PBS twice, cells were imaged immediately and the fluorescence signals from the green channel (500−550) and the red channel (580−630 nm) were collected, respectively. Consequently, these cells were treated by H2O2 (0.3%, 10 μL) and the fluorescence signal from the two channels above at different incubation time of H2O2 (1, 5, 7, 9, 10 min) were collected again. Each experiment was repeated three times. And then, the fluorescence intensity ratio between the green channel and the red one at different incubation time of H2O2 was calculated as a parameter of the sensitivity of JC-1 to MMP. Fluorescence Imaging. Confocal fluorescence images were obtained with an LSM 780 confocal laser scanning microscope, an Olympus FV 300 confocal microscope, or an Olympus FV 1200 confocal microscope. The overlap coefficients and fluorescence intensity of the images were determined by the software with the LSM 780 confocal microscope.
having 1 cm path length. Single-photon excited fluorescence (SPEF) spectra were obtained on a Hitachi F-2700 spectrofluorimeter equipped with a 450 W Xe lamp. The ideal dead-cell stain SYTOX Blue nucleic acid stain (S-11348) and MitoTracker Red FM (MTR) were purchased from Molecular Probes. Viability of the cells was assayed using a Cell Proliferation kit I with the absorbance of 492 nm being detected using a PerkinElmer Victor plate reader. Measurement of Binding Parameters of Carbazole Derivatives to DNA. The intrinsic binding constants (k) were obtained by the one-photon fluorescence titration method. The excitation wavelengths of 458 nm for LAD and the fluorescence intensity monitored at 550 nm were analyzed with the Scatchard equation according to eq 1:20 r /Cf = kn − kr
(1)
where k is the binding constant, n is the number of dye sites of perphosphate, r is the ratio of the concentration of the binding dye to the concentration of DNA (in phosphate), and Cf is the concentration of free dye. The concentration of the binding compound was calculated using eq 2:21 Cb = Ct[(F − F0)/Fmax − F0)]
(2)
Ct is the total compound concentration, F is the observed fluorescence intensity at a given DNA concentration, F0 is the fluorescence intensity in the absence of DNA, and Fmax is the fluorescence intensity of the totally binding compound. Cell Culture and Staining. All cells were grown in a 5% CO2 incubator at 37 °C. SiHa and HeLa cells were grown in HDMEM supplemented with 10% fetal bovine serum (FBS) and 1% penicillin and streptomycin. Human umbilical vein endothelial (HUVEC) cells were grown in M199 medium supplemented with 10% FBS and 2 ng/mL FGF-2. For living cell staining experiments, cultured cells were stained with 2 μM LAD in culture medium for 15 min at 37 °C and then imaged with fluorescence microscopy. For fixed cell staining experiments, cultured cells were pretreated according to the following procedure: cells were first fixed by 4% paraformaldehyde for 30 min and then with 2 μM LAD in culture medium for 15 min at 37 °C. For the investigating of the LAD staining effect, living cells were incubated with 2 μM LAD first for 15 min, then these cells were treated by H2O2 (0.3%,10 μL) for different time. Every time, the cells were washed to remove the unbound probes before adding H2O2. After rinsing with PBS twice, cells were imaged immediately. Second, the living SiHa cells were preloaded with 2 μM LAD for 15 min and then 0.2 μM carbonyl cyanide m-chlorophenylhydrazone (CCCP) was added to the culture medium. Every time, the cells were washed to remove the unbound probes before adding CCCP. After rinsing with PBS twice, cells were imaged immediately. After being treated for 20 min by CCCP, culture medium was sucked out to remove CCCP and the cells were imaged continuously until MMP recovered. To costaining experiments, living cells were incubated with LAD first for 15 min, then stained with 5 μM S-11348 for 30 min, or stained with 0.2 μM MTR for 30 min. Every time, the cells were washed to remove the unbound probes before staining with another probe. After rinsing with PBS twice, cells were imaged immediately. For cell viability assay, the study of the effect of LAD on the viability of cells was carried out using the methylthiazolyldiphenyl-tetrazolium bromide (MTT) assay. SiHa cells growing in log phase were seeded into 96-well plates (ca. 1 × 104 cells/ well) and allowed to adhere for 24 h. LAD (200 μL/well)
■
RESULTS AND DISCUSSION Photophysical Properties. The absorption (Figure S2a) and fluorescence spectra (Figure S2b) of LAD are shown in the Supporting Information. The structure of LAD was characterized by 1H NMR, 13C NMR, mass spectrometry (MS), and IR, as showed in the Supporting Information. Its maximum absorption and fluorescence wavelengths in various polarity solvents are in the range of 428−452 and 525−566 nm, respectively. Moreover, from Figure S1, the fluorescence intensity of LAD decreases with the increase of solvent polarity, while the maximum emission wavelength red-shifts, indicating an intramolecular charge-transfer (ICT) process of the probe. These properties are similar to that of the two previously reported probes CAI and 9E-BMVC.17,18 The ICT process can be reasonably attributed to the electron donor of the carbazole moiety and the electron acceptor of the pyridinium cation. In addition, the maxima wavelengths of absorption (λab) and emission (λem), the molar absorption coefficients (ε), and fluorescence quantum yields (Φ) in all solvents were obtained (Table S1). As shown in Table S1, the Φ of LAD is higher in 1,4-dioxane with lower polarity. Also, the absorption and fluorescence spectra of LAD-1 have been 3337
DOI: 10.1021/acs.analchem.6b03842 Anal. Chem. 2017, 89, 3335−3344
Article
Analytical Chemistry shown in Figure S3. From Figure S3, its maximum absorption and fluorescence wavelengths in EtOH are at 450 and 560 nm, respectively, which are similar to those of LAD. In addition, the fluorescence stability of LAD against different physicochemical conditions (ionic changes, temperature, pH, and viscosity) was tested, respectively (Figure S4). Because LAD was in a lipophilic environment when it accumulated in mitochondria or the groove of DNA, the mixed solvent of 1,4-dioxane and H2O (v/v = 1:1) was used when investigating the influence of ionic changes, temperature, and pH values on fluorescence of LAD, while glycerol (Gly) was used to study the viscosity effect. As shown in Figure S4a, various ions expressed in cytoplasm and the intracellular environment (Na+, K+, Ca2+, Mg2+, Mn2+, Ba2+, Br−, Cl−) were added to the solution of LAD, and the fluorescence intensity of LAD did not change obviously. At the same time, the influence of temperature (0−60 °C) was detected. From Figure S4b, the fluorescence intensity of LAD fluctuated within a small scope in a temperature range from 0 to 60 °C. Importantly, the fluorescence intensity of LAD was always stable at the physiological temperature range (30−40 °C). In addition, the influence of pH values in the physiological range (4.0−9.0) on the fluorescence intensity of LAD was also investigated. As shown in Figure S4c, the fluorescence intensity of LAD in the lipophilic environment (1,4-dioxane) was much higher than that in the mixed solvents with pH values of 4.0−9.0. Thus, the fluorescence intensity of LAD to detect MMP was not affected by the change of pH. Meanwhile, LAD showed enhanced emission in high-viscosity conditions (Figure S4d). This should be attributed to the limitation of intramolecular motions in such conditions. We think this property of LAD would benefit its application, because the mitochondria and nucleus can supply high-viscosity environments and enhance its fluorescence signals. Exclusively Staining Mitochondria with Normal MMP. Probe LAD was first used to stain vigorous cells to check its selectivity to mitochondria with normal MMP. As shown in Figure 2, when staining SiHa cells and HUVEC cells, LAD obviously displayed strong fluorescence in the cytosol. The emission in the cytosol sketched filament morphologies, which indicated that LAD may concentrate in mitochondria with
normal MMP. Furthermore, due to the ICT properties of LAD, it displayed rather weak fluorescence in high-polarity water and thus showed no background signals when staining the cells, and no fluorescence was observed in extracellular regions from the merged images. More importantly, according to the results in Figure 2, no fluorescence was emitted in the nucleus regions, which indicated LAD does not stain the nucleus in cells with normal MMP. In addition, as shown in Figure S5, LAD-1 obviously displayed strong fluorescence with filament morphologies in the cytosol, indicating that LAD-1 may concentrate in mitochondria with normal MMP. To strictly and accurately confirm the selectivity of LAD to mitochondria, conventional colocalization experiments have been performed with MTR,17 a commercially available mitochondria probe. As shown in Figure S6a, the absorption spectrum of the mixed solution of LAD and MTR was similar to the linear superposition (LAD plus MTR) of their absorption spectra, which indicated no chemical interaction between them. Very importantly, when illuminated by 405 nm (Figure S6b), only LAD could fluoresce in the mixed solution, while when excited by 561 nm (Figure S6c), only MTR fluoresced. These results demonstrated that LAD and MTR do not disturb with each other. Therefore, the green fluorescence in Figure 3 excited by 405 nm was solely from LAD, and the red signals excited with 561 nm came exclusively from MTR. According to the colocalization results, the colocalization coefficient of LAD and MTR was up to 0.87, and as shown in Figure 3d, the fluorescence from LAD and MTR showed similar distribution in cells. Therefore, LAD actually stained mitochondria with normal MMP in vigorous cells. Also, from Figure S7, there would not be optical interferences between the two dyes. Therefore, the green fluorescence in Figure S8 excited by 405 nm was solely from LAD-1, and the red signals excited with 543 nm came exclusively from MTR. According to colocalization results, the colocalization coefficient of LAD and MTR was up to 0.94. Thus, LAD-1 could also actually stain mitochondria with normal MMP in vigorous cells. High Selectively Targeting the Nucleus When MMP Vanishes. To verify the ability of LAD in high selectively staining of the nucleus in cells whose MMP have vanished, the SiHa cells have been fixed by paraformaldehyde and the strictly staining experiments of LAD were performed. According to Figure 4a, all nuclei had been stained in fixed cells, and the position and shape stained by LAD were the same as that of the nucleus in the DIC image, implying that LAD accurately labeled the nucleus in MMP vanishing cells. Most importantly, no fluorescence was observed in cytosol regions from the merged image of MMP vanishing cells. In addition, as we know, a long period of incubation of CCCP could also eliminate the MMP. So, the staining experiments of LAD in SiHa and HUVEC cells incubated by CCCP were performed. From Figure 4, parts b and c, similarly with the staining results in fixed cells, LAD also exclusively labeled the nucleus, which also indicated the high selectivity to the nucleus of LAD when MMP vanished. Also, from the Figure S9, LAD-1 could exclusively target the nucleus in the cells incubated for 30 min by CCCP, which showed the same targeting ability with LAD. To explain the mechanism that LAD and LAD-1 target nuclei in MMP vanished cells, the fluorescence DNA titration experiment18 was performed, and the calf thymus DNA (ctDNA) was used. As shown in Figure 5 and Figure S10, both LAD and LAD-1 have shown weak fluorescence in buffer solutions, and the fluorescence gradually enhanced with the
Figure 2. Fluorescence images, differential interference contrast (DIC) photos, and their merged pictures of SiHa cells (a) and HUVEC cells (b) stained by LAD (2 μM, 15 min). λex = 488 nm, λem > 565 nm. Bar = 20 μm. Confocal microscope: Olympus FV 300. 3338
DOI: 10.1021/acs.analchem.6b03842 Anal. Chem. 2017, 89, 3335−3344
Article
Analytical Chemistry
Figure 3. Confocal fluorescence images of SiHa cells stained with LAD (2 μM, 15 min) and MTR (0.2 μM, 20 min): (a) images of LAD, λex = 405 nm, λem = 450−700 nm; (b) images of MTR, λex = 561 nm, λem = 570−700 nm; (c) overlay images of panels a and b. (d) The relative emission intensity distributions of MTR and LAD in various pixel points along the red arrow in panel c. Bar = 10 μm. Confocal microscope: Zeiss LSM 780.
binding to DNA also resulted in the isolation of probes from water. The k values have been obtained as 8.30 × 105 and 2.47 × 106 M−1, respectively, by analyzing the fluorescence titration data with the Scatchard equation.20,22 As shown in Figure S1, the k value of DNA probe 9E-BMVC was 6.46 × 106 M−1 and the fluorescence intensity of mitochondria probe CAI barely increased with the addition of DNA, indicating no interaction between the CAI and DNA. In order to provide good insights, the targeting abilities of LAD, LAD-1, CAI, and 9E-BMVC are compared in detail in Table S2. Therefore, it is very necessary to possess appropriate affinity to DNA for a spatially dependent probe for detecting MMP. According to our experiments, the k values to DNA at the magnitude of 105 or 106 M−1 may be suitable. Locating Both in the Mitochondria and Nucleus When MMP Decreased. According to the results above, LAD located in mitochondria with normal MMP, and it targeted the nucleus when MMP vanished. From this, we may deduce that LAD should locate both in the mitochondria and nucleus when MMP decreased. In order to detect this inference, we investigated the staining effect of LAD in cells treated by two MMP reducing agents, CCCP and H2O2. In the first experiment, the living SiHa cells were loaded with LAD and the staining effect was recorded. Consequently, these cells were treated by H2O2, and influence of the treating time on the targeting site of LAD was investigated in detail (Figure 6 and Video S1). As shown in Figure 6, LAD was initially accumulated in mitochondria before adding H2O2. After this, with the extension of H2O2 treating time, fluorescence gradually appeared in the nucleus, indicating that LAD gradually moved to the nucleus. After 12 min, almost all fluorescence emitted
Figure 4. Fluorescence images, DIC photos, and their merged pictures of SiHa cells (a and b) and HUVEC cells (c) stained by LAD (2 μM, 15 min) after being fixed by paraformaldehyde (30 min) (a) or incubated with CCCP (30 min, b and c). λex = 488 nm, λem > 565 nm. Bar = 20 μm. Confocal microscope: Olympus FV 300.
addition of ctDNA solution. This turn-on behavior of the probe could also be interpreted by the ICT properties, because
Figure 5. Fluorescence titration of LAD with DNA (a) and the fitted curve of the graph according to Scatchard equation (b). λex = 458 nm and λem = 550 nm for LAD. [LAD] = 5 μM, [phosphate in DNA] = 0−20 mM for LAD. For LAD, the binding constant is 8.30 × 105 M−1, the corresponding site is 1.72 phosphates/dye, goodness-of-fit R2 = 0.979. 3339
DOI: 10.1021/acs.analchem.6b03842 Anal. Chem. 2017, 89, 3335−3344
Article
Analytical Chemistry
Figure 6. Confocal fluorescence images of SiHa cells stained by LAD (2 μM, 15 min) before (a, control) and after being incubated by H2O2 (0.3%, 10 μL) at 1, 2, 5, 8, 10, 12 min (b−g). (h) The merged image of panels a and g, whose color was replaced by green. λex = 488 nm, λem > 565 nm, bar = 50 μm. (i) The ratios of fluorescence intensity in the nucleus to that in whole cells with different incubation time of H2O2. Data are expressed as mean ± SD, control, experiment times n = 3. Confocal microscope: Olympus FV 300.
Figure 7. Confocal fluorescence images of HeLa cells stained by LAD (2 μM, 15 min) before (a) and after being incubated (b−f; 1, 3, 6, 8, 10 min) with H2O2 (0.3%, 10 μL). λex = 488 nm, λem > 565 nm, bar = 50 μm. (g) The ratios of fluorescence intensity in the nucleus to that in whole cells with different incubation time of H2O2. Data are expressed as mean ± SD, control, experiment times n = 3. Confocal microscope: Olympus FV 300.
Figure 8. Confocal fluorescence images of SiHa cells stained with LAD (2 μM) for 15 min (a) and then incubated with CCCP (20 μM) for different time (b−f, 5−28 min), and after removal of CCCP (g−j, 8−20 min). (k) The merged image of panels a and f, whose color was replaced by green. (l) The merged image of panel f and its DIC. λex = 473 nm, λem > 560 nm, bar = 50 μm. (m) The ratios of fluorescence intensity in the nucleus to that in whole cells when incubating and removing CCCP at different time. Data are expressed as mean ± SD, control, experiment times n = 3. Confocal microscope: Olympus FV 300.
from the nucleus in the cells. Generally, the longer time being treated by H2O2, the more MMP decreased, which would result in that more probes target the nucleus. Accordingly, LAD located both in the mitochondria and nucleus in Figure 6b−f when MMP decreased. At the same time, an important fact that the fluorescence intensity ratios between the nucleus and whole
cells increased from 0.1 to 0.8 with extension of H2O2 treating time supported this deduction above. (The calculation method about the ratio is instructed in detail at Supporting Information, Figure S11.) Also, in order to more visually show the influence of H2O2 treating on targeting, Figure 6h was given by merging parts a and g of Figure 6, whose fluorescent color was replaced 3340
DOI: 10.1021/acs.analchem.6b03842 Anal. Chem. 2017, 89, 3335−3344
Article
Analytical Chemistry by green. From Figure 6h, the position emerged by green color nicely padded the nucleus regions in Figure 6a, which indicated again that LAD can stain both in the mitochondria and nucleus when MMP decreased. The same results have been also obtained in HeLa cells (Figure 7 and Video S2). From Figure 7a, LAD exclusively stained mitochondria with normal MMP. When the H2O2 treating time was 1−10 min, LAD labeled the mitochondria and nucleus simultaneously (Figure 7b-f), and the corresponding fluorescence intensity ratios between the nucleus and whole cells increased from 0.1 to 0.8. This fact proved again that LAD located both in the mitochondria and nucleus when MMP decreased. When affected by various physiological and pathological conditions, the MMP is often fluctuant and its change is reversible sometimes. Thus, the reversibility of staining sites of LAD was also studied in detail in the process of MMP reversible change. For this purpose, CCCP was added to reduce MMP and then MMP could be recovered by washing out CCCP (Figure 8 and Video S3). As shown in Figure 8a, LAD was initially accumulated in mitochondria before adding CCCP. From Figure 8b−f, with the extension of CCCP treating time, fluorescence gradually appeared in the nucleus, indicating that LAD gradually moved to the nucleus. After treating the cells by CCCP for 20 min, almost all fluorescence emitted from the nucleus. Importantly, as shown in Figure 8g−j, after washing out CCCP, the MMP recovered and the fluorescence reappeared gradually in mitochondria. Moreover, after 18 min, almost all probes moved back to the mitochondria. Therefore, LAD can detect the process of not only MMP decreasing but also MMP recovering, which could be used to long-time observe MMP fluctuation. Also, in order to intuitively show the influence of CCCP treating on the targeting site of LAD, the Figure 8k was given by merging parts a and f of Figure 8, whose pseudocolor was replaced by green. From Figure 8k, the position emerged by green color nicely padded the nucleus regions in Figure 8a. Also, from Figure 8l, the position and shape stained by LAD were the same as that of the nucleus in its DIC image, implying that LAD accurately labeled the nucleus in cells incubated for 28 min by CCCP. Further, the ratio values of fluorescence intensity between the nucleus and the whole cells were also obtained. As shown in Figure 8m, the ratio gradually increased from 0.1 to 0.9 during 28 min after addition of CCCP and decreased from 0.9 to 0.1 during 20 min after washing out CCCP. This fact proved that LAD could be used to monitor the process of MMP reversible change. Most importantly, it demonstrated again that LAD could locate both in the mitochondria and nucleus when MMP decreased. Comparison of the Sensitivity to MMP between LAD and JC-1. The fluorescence images of HeLa cells stained by JC-1 were first obtained (Figure 9a). Previously, the sensitivity of LAD to the change of MMP has been detected in the study of the MMP decreasing process. Also, the fluorescence intensity ratio in the nucleus to that in whole cells from LAD was given again (Figure 9b, black line), which could serve as a parameter of the sensitivity of LAD to MMP. Comparatively, the sensitivity of JC-1 to MMP was also investigated in cells treated by H2O2 with the same experimental conditions to LAD in Figure 7. The fluorescence images from the green and red channels as well as their merged images have been shown in Figure 9a. According to Figure 9a, the red fluorescence signal was bright while the green fluorescence was very weak before the H2O2 treatment, indicating that the MMP was normal.
Figure 9. (a) Confocal fluorescence images of HeLa cells stained by JC-1 (20 μg/mL, 30 min) before (a, control) and after being incubated by H2O2 (0.3%, 10 μL) at 1, 5, 7, 9, 10 min. Green, λex = 473 nm, λem = 500−550 nm; red, λex = 543 nm, λem = 580−630 nm; bar = 50 μm. (b) The fluorescence intensity ratios in the nucleus to that in whole cells from LAD (black line, coming from Figure 7g) and the fluorescence intensity ratios between the green channel and the red one from JC-1 at different incubation time of H2O2. Data are expressed as mean ± SD, control, experiment times n = 3. Confocal microscope: Olympus FV 1200.
With the extension of H2O2 treating time, the green fluorescence enhanced and red fluorescence waned gradually, which illustrated that the MMP decreased and JC-1 transformed from J-aggregates into monomers. Also, the fluorescence intensity ratio between the green channel and red one at different incubation time of H2O2 was calculated and plotted in Figure 9b (red line). As shown in Figure 9b, compared to JC1, LAD can sense the change of MMP more sensitively, indicating that a spatially dependent fluorescent probe has huge potential in detecting MMP. Application in Distinguishing Dead Cells from Viable Ones in the Same Cell Population. According to the sensing MMP character of LAD, we speculate that its fluorescence images can be used to distinguish dead cells from viable ones in the same cell population, which has great significance in biology and medicine. For example, it can help to screen anticancer drugs23 and evaluate cell viability.24 Also, it is useful in estimating whether a certain environmental condition, such as in bioreactors, is suitable for cell culture or not.25 In order to verify the speculation, counterstaining experiments of LAD with S-11348, a dye only staining dead cells, have been performed (Figure 10). As shown in Figure S12, upon binding DNA, S-11348 has an absorption band in the region of 380−470 nm with emission at 450−550 nm. From 3341
DOI: 10.1021/acs.analchem.6b03842 Anal. Chem. 2017, 89, 3335−3344
Article
Analytical Chemistry
Figure 10. Confocal fluorescence images in SiHa cells stained with 2 μM LAD and 5 μM S-11348 in sequence. Each incubation was 15 min. (a) LAD fluorescence image, λex = 473 nm, λem = 500−600 nm; (b) S-11348 fluorescence image, λex = 405 nm, λem = 410−460 nm; (c) DIC; (d) overlay images of panels a−c. Bar = 20 μm. Confocal microscope: Olympus FV 1200.
Figure 11. (a) MTT results of SiHa cell viabilities after incubation with LAD (2 μM) for 2, 8, and 24 h; (b) MTT results of SiHa cell viabilities after incubation with LAD for 2 h at different incubation concentrations (2, 5,10 μM). Data are expressed as mean ± SD (experiment times n = 3).
To date, the commercial fluorescent dye commonly used to distinguish dead and viable cells is the Cellstain-Double staining kit. In this kit, calcein-AM and propidium iodide (PI) can stain viable and dead cells, respectively.26 Thus, this method needs to incubate the two dyes in sequence. When imaging viable cells stained by calcein-AM, the 470 nm excitation and 535 nm emission wavelengths are used, and 540 nm excitation and 660 nm emission wavelengths can be used to image dead cells. Comparatively, a single probe LAD can accomplish the task of the Cellstain-Double staining kit using a single excitation wavelength and single detection range. In fluorescence photos, fluorescence in viable cells comes from only mitochondria in the cytosol but fluorescence in dead cells comes only from the nucleus. More importantly, using a single probe can greatly shorten the testing time and improve efficiency, which will bring enormous convenience in biology and pharmacology research. In practical application, another key consideration is toxicity of LAD itself to living cells. Thus, cytotoxicity of LAD was evaluated in living SiHa cells by MTT assays. The influence of LAD on SiHa and HeLa cells at different incubation time and concentrations were studied, and the cell viability assay data were quantified (Figure 11 and Figure S13). The results showed that both SiHa and HeLa cells showed >85% viability after 24 h of incubation with 2 μM LAD and 2 h of incubation
Figure S12, the absorption of LAD is about 400−500 nm and the emission is about 480−650 nm. Consequently, S-11348 displays rather low absorption at 473 nm, while LAD shows little fluorescence in the emission range of 410−460 nm. In consequence, red fluorescence excited by 473 nm should only come from LAD, while blue fluorescence selected from 410 to 460 nm should be attributed to S-11348 excited by 405 nm. Consequently, the relevant confocal microscope photos of a group of SiHa cells incubated with 2 μM LAD and 5 μM S11348 in sequence are shown in Figure 10. From Figure 10a, LAD stained the all cellular mitochondria except for two cells whose nuclei were stained. Importantly, the two exceptional cells in Figure 10a were just stained by S-11348 (Figure 10b). In addition, from the DIC picture (Figure 10c) the two exceptional cells have broken. More importantly, the two exceptional cells in Figure 10a and two cells stained by S-11348 in Figure 10b as well as the two broken cells in Figure 10c are fully overlapped in their merged image (Figure 10d), which indicated that the two exceptional cells in Figure 10a were dead and other cells should be intact and viable. The results confirmed that a single probe LAD can distinguish dead cells from viable ones in the same cell population conveniently and efficiently. The distinguishing principle is based on that LAD located in mitochondria with normal MMP and it targeted the nucleus when MMP vanished. 3342
DOI: 10.1021/acs.analchem.6b03842 Anal. Chem. 2017, 89, 3335−3344
Article
Analytical Chemistry with 10 μM LAD. Thus, the LAD of concentrations at a micromolar order of magnitude and short incubation time (such as
