Targeting and Imaging of Mitochondria Using Near-Infrared Cyanine Dye and Its Application to Multicolor Imaging.

Herein, we report water-soluble mitochondria-selective molecules that consist of a target-specific moiety conjugated with a near-infrared (NIR) imaging agent through variable spacer length. The presented NIR fluorescent cyanine-5 (Cy-5) chromophore exhibits excellent photostability, narrow NIR absorption and emission bands, high molar extinction coefficient, high fluorescence quantum yield, and long fluorescence lifetime. The biological compatibility and negligible cytotoxicity further make the dye an attractive choice for biological applications. Confocal fluorescence microscopic studies in the fixed human lung carcinoma cell line (A549) stained with the targeting NIR Cy-5 dyes (Cy-5a and Cy-5b) at 700 nM concentration show their cellular uptake and localization, which is compared with the nontargeting Cy-5c. Mitochondrial target specificity is demonstrated by colocalization experiments using the mitochondrion-tracking probe, MitoTracker Red and lysosome-tracking probe, LysoTracker Green. Multicolor imaging of cellular organelles in A549 cells is achieved in combination with suitable target-specific dyes with distinct excitation and emission, such as green emitting FM 1-43FX to selectively image the plasma membrane, blue-fluorescent DAPI to stain the nucleus, and the synthesized NIR Cy-5 to image the mitochondria. Higher accumulation of the dye inside the cancer cell mitochondria compared to the noncancerous cell is also demonstrated.

Introduction

Selective targeting and staining of particular cellular organelles such as nucleus, mitochondria, Golgi apparatus, endoplasmic reticulum, and so forth is an emerging field of current research.[1] However, it is exigent to work on cellular environment due to the intricacy in biological systems. Among the different cellular organelles, mitochondria are widespread targets for all cancer cells.[2] Mitochondria play a fundamental role in programmed cell death such as apoptosis.[3] Mitochondrial targeting, incorporation, and imaging are the critical steps for new drug design and development processes. Inner mitochondrial membrane (IMM) of normal healthy cells maintains a strong negative membrane potential [(ΔΨm)normal −150 to −180 mV].[4] Amendment in ΔΨm is an imperative feature of cancer [(ΔΨm)cancer ≈ −220 mV].[4b] This high negative ΔΨm of mitochondria is unique and is not present in any other cellular organelle (plasma membrane potential, ΔΨp −30 to −60 mV), which offers a design scope for selective targeting and tracking of mitochondria. Fluorescent probes are handy for monitoring a variety of biological processes in cells using confocal microscopy and flow cytometry. However, near-infrared (NIR) fluorescent dyes (650–900 nm) have a significant advantage over the visible fluorescent dyes.[5] Cellular or tissue components have negligible absorption and minimal autofluorescence in the NIR region; hence, highly sensitive, less scattered, and precise detection is possible when exogenous NIR dyes are incorporated. However, it is challenging to construct an organelle-selective organic NIR imaging agent because most NIR dyes are restricted within their biological applications due to photobleaching, propensity to aggregate in the buffer solution, low quantum yield, fast fluorescence decay, and insufficient stability in biological systems. Moreover, less target selectivity and broad absorption/emission band of most NIR probes limit its application to multicolor imaging of cells. Multicolor imaging and tracking of distinct cellular organelles inside the same cell is crucial for understanding complex biological processes.[6] Our strategy is to exploit the extraordinary biophysical membrane property of mitochondria to target IMM, as cationic molecules should attract and accumulate preferentially within the negatively charged mitochondrial matrix.[4] In addition to the cationic character, adequate lipophilicity is also necessary to accomplish a significant uptake in mitochondria.[7] Methodical explorations of the target-specific cationic moiety conjugated to an NIR cyanine-5 (Cy-5) chromophore through a relatively flexible linker with variable length have been conducted for selective targeting and staining of mitochondria. In this article, the synthesized NIR fluorescent molecules (Cy-5a and Cy-5b) exhibit excellent photophysical properties, exceptional photostability, narrow NIR absorption and emission band, high molar extinction coefficient, long fluorescence lifetime with high fluorescence quantum yield, bright NIR fluorescence with negligible background signal, good water solubility, outstanding biocompatibility, and low cytotoxicity. We envision that—(1) targeting functionality containing NIR molecule (Cy-5a and Cy-5b) should lead to a more selective, greater, and faster uptake by mitochondria than nontargeting Cy-5c molecule. Therefore, there is a need to inspect their mitochondrial uptake; (2) because of narrow NIR absorption and emission band of mitochondria-targeting Cy-5, acquisition of multicolor imaging of cellular organelles in the same cell could be possible using a combination of suitable dyes with distinct excitation and emission characteristics; (3) because of more hyperpolarized ΔΨm of cancer cells compared to normal healthy cells, the dye should accumulate more in the cancer cell mitochondria than noncancerous cell mitochondria; (4) cancer cell mitochondrial incorporation of the triply positive charged Cy-5 dye is expected to depolarize mitochondrial membrane potential. Therefore, its effect on malignant mitochondrial targeting, incorporation, staining, membrane potential depolarization, and multicolor imaging application is investigated using human carcinoma cell lines.

Results and Discussion

The heterocyclic 2,3,3-trimethylindolenine molecule is synthesized by Fischer indole synthesis and the N of the indolenine residue is covalently conjugated with the mitochondria-targeting triphenylphosphonium (TPP+) moiety using alkylating agents 2 and 3 (Scheme ).[8] The synthesis of Cy-5a [two −(CH2)4– spacer] and Cy-5b [two −(CH2)6– spacer] involves the condensation of heterocyclic compounds containing an activated methyl group (4 and 5) with malonaldehydebis(phenylimine)monohydrochloride (1), which is prepared from the commercially available 1,1,3,3-tetramethoxypropane (Scheme ). Cy-5a and Cy-5b (dark blue powder) are characterized by 1D (1H, 13C, and 31P) NMR, 2D (1H–1H DQF COSY) NMR, and high-resolution ESI-MS (Figures S2–S7). Both the symmetric molecules Cy-5a and Cy-5b consist of two P atoms and exhibit one 31P NMR peak at ∼24 ppm (Figures S3a and S6a). A control dye lacking the mitochondria-targeting TPP+ moiety (Cy-5c) is also synthesized and characterized (Scheme , Figures S8 and S9).

Scheme 1

Synthesis of Cy-5a, Cy-5b, and Cy-5c Dyes

The synthesized Cy-5a and Cy-5b are small molecules with excellent solubility in aqueous solutions (5.65 × 10–5 mol L–1 for Cy-5a and 8.15 × 10–6 mol L–1 for Cy-5b in H2O, Figures S11e, S13c) due to triply positive charge. Moreover, they dissolve well in organic solvents. The absorption and emission features of Cy-5a and Cy-5b are examined in various solvents [H2O, DMF, DMSO, CH3OH, CH3CN, CHCl3] including phosphate-buffered saline (PBS) and 2-[4-(2-hydroxyethyl)piperazin-1-yl]ethanesulfonic acid (HEPES) buffer solutions (Figures , S10–S13). The absorption spectra exhibited an intense peak (λmax) at 660 nm in CHCl3 due to π → π* transitions with large molar extinction coefficient of ca. 1.90 × 105 M–1 cm–1 for Cy-5a as well as Cy-5b (Table S1). For both the probes, the fluorescence maximum (λem) is observed at ca. 685 nm in CHCl3 (λex = 660) with ca. 25 nm Stokes shift (Table S1). Cy-5a and Cy-5b exhibit modest solvatochromism in the absorption and emission with ∼20 nm bathochromic shifts being observed by decreasing the solvent polarity from H2O to CHCl3, which is probably due to combined effects of dielectric constant and refractive indices of the solvents (Table S1, Figures S10, S12). The absorption and emission spectra, shapes, and λex/λem of Cy-5a and Cy-5b in PBS and HEPES buffer (pH 7.4, 37 °C) are comparable to those observed in H2O (Figure S12e,f). Cy-5a and Cy-5b exhibit narrow absorption and emission bands, which is extremely important for the multicolor imaging process, where a combination of various dyes is frequently applied. Additionally, no evidence of aggregation in aqueous buffer and organic solvents is observed for Cy-5a and Cy-5b (Figures S11 and S13a), which is a crucial issue for most NIR cyanine molecules that absorb in this range. The lack of aggregation is due to the presence of two cationic TPP+ residues, which provide electrostatic repulsion between molecules. Fluorescence quantum yields, Φf = 0.30 and Φf = 0.24 are observed in DMSO for Cy-5a and Cy-5b, respectively, which are higher than that of NIR dyes zinc phthalocyanine (Φf = 0.20 in DMSO), oxazine 1 (Φf = 0.141 in EtOH), cryptocyanine (Φf = 0.012 in EtOH), IR-125 (Φf = 0.132 in EtOH), and comparable with other Cy-5 dyes (Table S2). Fluorescence lifetimes (τ) of Cy-5a and Cy-5b are determined in various solvents by the time-correlated single photon counting (TCSPC) technique. Values of τ ≈ 1.948 ns for Cy-5a and τ ≈ 1.769 ns for Cy-5b in DMSO are found (Figures e, S14 and Table S3). Cy-5a and Cy-5b dyes also exhibit modest solvatochromism in their fluorescence lifetime, and τ is longer in CHCl3 than in H2O (Table S3).

Figure 1

(a) Design and chemical structure of mitochondria-targeting NIR Cy-5a and Cy-5b molecules. (b) Absorption spectra of Cy-5a (4 μM) in various solvents. (c) Normalized absorption (solid line) and emission spectra (dashed line) of Cy-5a in CHCl3. (d) Normalized absorption (solid line) and emission spectra (dashed line) of Cy-5b in different solvents. (e) TCSPC plots of Cy-5b in diverse solvents after excitation with delta diode laser at 650 nm.

The lipophilicity of Cy-5a and Cy-5b is determined by octanol/PBS partition coefficient measurements using Poctanol/PBS = [C]octanol layer/[C]PBS layer.[8] Cy-5b with two −(CH2)6– linkers is found to be more lipophilic (P = 7.12, log P = +0.85) than Cy-5a with two −(CH2)4– linkers (P = 2.88, log P = +0.46) (Figure S15). To scrutinize biocompatibility and cellular localization of the dyes, biological experiments are designed and performed.

UV/vis experiment shows that Cy-5a and Cy-5b are highly stable in PBS (pH 7.4, 37 °C) over 24 h (Figure S16). Moreover, these dyes are stable over a wide range of pH (Figure S16d,e).The probable interferences of a range of biological analytes are examined. The fluorescence experiment of Cy-5b is performed in the presence of different crucial metal ions (Na+, K+, Ca2+, Mg2+, Cu2+, Zn2+, and Fe3+ as their chloride salts) and redox substances related with oxidative stress, including glutathione under physiological conditions (PBS, pH 7.4, 37 °C) (Figure S17). No obvious changes are observed in the fluorescence spectra of Cy-5b in the presence of biological interferents. As a proof of concept for transportation of the Cy-5 dye from aqueous buffer to lipid membrane, we prepared large unilamellar vesicles (LUVs) made of 1,2-dioleoyl-sn-glycero-3-phosphocholine (DOPC)/cholesterol (80:20) in HEPES buffer at pH 7.4 using the extrusion technique (hydrodynamic diameter = 168 nm, PDI = 0.115) (Figure S18a). λem of Cy-5b is observed at 661 nm in the HEPES buffer at pH 7.4, and when this buffer solution of the dye is added to the LUVs, a 14 nm red-shift in λem is observed which indicates transportation of the dye from the buffer to the lipid membrane (Figure S18b).

Cell viability studies using 3-(4,5-dimethyl-2-thiazolyl)-2,5-diphenyltetrazolium bromide (MTT) assay shows that the probe Cy-5a and Cy-5b have negligible cytotoxicity against human noncancerous [peripheral blood mononuclear cells (PBMC), human embryonic kidney (HEK293), and lung fibroblast (WI38)] and cancerous [human pancreatic (MIA PaCa-2) and lung (A549)] cell lines at different concentrations over 24 h (Figure S19). Flow cytometric titration using Cy-5b at various concentrations for A549 cells indicates cellular uptake and internalization (Figure S20). Cellular imaging experiments are carried out using confocal microscopy at 0.7 μM concentration of the Cy-5 dye with short incubation time (10 min) on fixed A549 carcinoma cell line, resulting in bright NIR fluorescence and fast cellular internalization (Figures and S21). Both Cy-5a and Cy-5b dyes corroborate good photostability and allow prolonged confocal imaging without significant photobleaching. In order to confirm the mitochondrial target specificity, colocalization experiments involving Cy-5a and Cy-5b (λex/λem 650/675) are performed in the fixed A549 carcinoma cell line using a commercially available mitochondrion-tracking probe, MitoTracker Red (λex/λem 579/599) (Figure S22).[9] The fluorescence image produced using Cy-5b (log P = +0.85) exhibits high levels of colocalization with that of MitoTracker Red [Pearson’s correlation (PC) coefficient: 0.91 in A549 cells] (Figure ). For Cy-5a (log P = +0.46), PC coefficient: 0.86 in A549 cells is found (Figures , S21A). The observed high PC coefficient established the mitochondrial specificity of the two probes. Cy-5a and Cy-5b are compared with a control dye Cy-5c (λex/λem 650/671 in DMSO, Figure S13D), lacking the mitochondria-targeting TPP+ moiety and exhibit weaker and nonselective staining in confocal microscopy (PC coefficient: 0.56 in A549 cells) (Figure S21B). In order to confirm the mitochondrial selectivity of Cy-5a and Cy-5b over lysosome, colocalization experiments are performed using a commercially available green-fluorescent lysosome-tracking probe, LysoTracker Green DND-26 (λex/λem 504/511), which exhibits poorer overlap in confocal microscopy (PC coefficient: 0.25 and 0.24 in A549 cells for Cy-5a and Cy-5b, respectively) (Figure S23). This suggests that tethering to target-specific lipophilic TPP+ cation with NIR Cy-5 chromophore is an efficient strategy for selective mitochondrial targeting and staining.

Figure 2

Confocal microscopic images of (a) Cy-5a and (b) Cy-5b colocalized with MitoTracker Red (MTR) in fixed A549 carcinoma cells. DAPI, MTR, and Cy-5 dyes are recorded using laser excitation wavelengths at 405, 559, and 635 nm, respectively. Colocalization scatter plots of the overlay pictures show PC: 0.86 for Cy-5a and PC: 0.91 for Cy-5b.

The TPP+ containing Cy-5 dyes can target mitochondria of both normal cells and cancer cells. The flow cytometric experiment demonstrates that the Cy-5b dye penetrates both the human lung cancer cells, A549, as well as the noncancerous WI38 cells (Figure S24). However, a 2-fold enhancement in the fluorescence intensity of Cy-5b is found inside the A549 cancer cells compared to WI38 noncancerous cells (Figure S24). Flow cytometric analysis corroborates our hypothesis that the dye is more attracted and accumulated in the mitochondria of cancer cells compared to normal healthy cells due to higher potential gradient of the former.[4b]

The tricationic Cy-5b is expected to depolarize mitochondrial membrane potential after internalization. To decipher our hypothesis of mitochondrial targeting and depolarization by tricationic Cy-5b dye, mitochondrial membrane potential changes are examined using a JC-1 dye.[10] A considerable depolarization in the mitochondrial membrane potential in A549 cells is observed after treatment with Cy-5b. At 0.5 and 0.7 μM concentration of Cy-5b, the percentage of cells with depolarized mitochondria increases to 21.7 and 26.2%, respectively, without noticeable leakage of Cy-5b dye from mitochondria (Figures , S25).

Figure 3

Determination of mitochondrial membrane potential depolarization in A549 cancer cells using JC-1 dye. After cellular staining with JC-1 dye (1 mg mL–1), the fluorescent signal is analyzed in the FITC and PE channels using a 494 nm laser. (a) JC-1 staining of control A549 cells without Cy-5 dye. (b) JC-1 staining of A549 cells after incubation with Cy-5b at 0.5 μM concentration. (c) JC-1 staining of A549 cells after incubation with Cy-5b at 0.7 μM concentration. The percentage of cells with depolarized mitochondria increases from 6.63% (control) to 21.7% (0.5 μM Cy-5b) to 26.2% (0.7 μM Cy-5b).

Multicolor imaging and tracking of cellular organelles in the same cell remains challenging because of the lack of suitable target-specific fluorescent probes with well-separated excitation and emission bands. Narrow NIR absorption and emission bands of mitochondria-targeting Cy-5 prompted us for the multicolor imaging of cellular organelles using a combination of suitable targeting dyes with distinct excitation and emission.[6] Three-color fluorescence A549 cell staining highlighting the plasma membrane (green color), nucleus (blue color), and mitochondria (red color) is achieved by using lipophilic green emitting styryl dye, FM 1-43FX to selectively stain the plasma membrane, blue-fluorescent DAPI to label the nucleus, and Cy-5b to selectively target and stain the mitochondria of A549 cells (Figures , S26).

Figure 4

Multicolor confocal microscopic image of A549 cell labeled with lipophilic green emitting FM 1-43FX (laser ex/em 473/590 nm) to stain the plasma membrane (green color), blue-fluorescent DAPI (laser ex/em 405/461 nm) to target nucleus (blue color), and NIR Cy-5b (laser ex/em 635/668 nm) to selectively stain mitochondria (red color). This multicolor imaging of cellular organelles in the same cell is acquired using appropriate filter sets in confocal microscopy.

Conclusions

In summary, we report some unique design concepts for the development of NIR Cy-5 molecules that have the ability to selectively target and stain mitochondria. Submicromolar concentration of the targeting dye is enough to localize and stain mitochondria within minutes. Narrow absorption and emission bands permit the acquisition in multicolor imaging application of the cells. Currently, further development of cellular organelle-selective targeting and multicolor imaging for targeted theranostic applications are under way in our lab. Moreover, in vivo targeting of NIR Cy-5 dye in tumors will be explored in future.

Experimental Section

Materials

5,5′,6,6′-Tetrachloro-1,1′,3,3′-tetraethyl-imidacarbocyanine iodide (JC-1, mitochondrion membrane potential sensitive probe), 1,1,3,3-tetramethoxy propane, HEPES, and HSA were obtained from Sigma-Aldrich. TLC silica gel 60 F254 was purchased from Merck. Zinc phthalocyanine was acquired from TCI chemicals. DOPC and cholesterol were purchased from Avanti Polar Lipids (USA) and stored in a freezer at −20 °C. NMR solvents were procured from Cambridge Isotope Laboratories, Inc. Commercially available solvents were used without further purification unless otherwise stated. HPLC grade solvents were purchased from Sigma-Aldrich for spectroscopic measurements. Human ductal pancreatic cancer cells (MIA PaCa-2), non-small lung adenocarcinoma cells (A549), noncancerous human embryonic kidney (HEK293) cell line, and lung fibroblast (WI38) cells were acquired from The National Centre for Cell Science, India. Human PBMC were isolated from healthy samples. Dulbecco’s modified eagle medium (DMEM), fetal bovine serum (FBS), trypsin EDTA mixture, and antibiotic antimycotic solution, MTT were obtained from Himedia. MitoTracker Red and LysoTracker Green DND-26, and FM 1-43FX were procured from Thermo Fisher Scientific. Ultrapure H2O (Millipore) was used for all experiments.

Synthesis, Purification, and Characterization of Cy-5a, Cy-5b, and Cy-5c Dyes

All syntheses were performed under N2 atmosphere using dry solvents. Analytical thin layer chromatography (TLC) was carried out on silica gel-coated aluminum sheets (TLC silica gel 60 F254) with appropriate solvents and the results were detected by naked eye or UV lamp and/or developed with I2. The compounds were purified via column chromatography using silica gel 100–200 mesh. The solvents for column chromatography were distilled prior to their use. 2,3,3-Trimethylindolenine was synthesized from a literature procedure by the Fischer indole synthesis.[11]

Malonaldehyde Bis(phenylimine)monohydrochloride (1)

1,1,3,3-Tetramethoxy propane (10.50 mL, 60 mmol) and dilute HCl (8.5 mL concentrated HCl and 171 mL H2O) were taken in a round bottom flask and heated to 50 °C while being continuously stirred. A solution of aniline (11.10 mL, 120 mmol) in dilute HCl (15 mL concentrated HCl and 200 mL H2O) was added dropwise to the solution of 1,1,3,3-tetramethoxy propane using an additional funnel over a period of 2 h. The reaction mixture’s color gradually changed from colorless to orange during the addition of aniline solution. After the addition was completed, the reaction mixture was stirred for an additional 4 h at 50 °C. The reaction mixture was then cooled to room temperature to get an orange precipitate. The orange precipitate was filtered and dried under vacuum. The product was used in the next step without any further purification [Rf = 0.50 (DCM/MeOH = 9:1)].

Yield: 10.60 g (68.26%).

1H NMR (500 MHz, DMSO-d6, 25 °C): δ 12.76 (2H, s), 8.98 (2H, d, J = 11.0 Hz), 7.45–7.42 (8H, m), 7.24–7.21 (2H, m), and 6.57 (1H, t, J = 11.5 Hz) ppm. 13C NMR (125 MHz, DMSO-d6, 25 °C): δ 158.4, 138.6, 129.7, 125.7, 117.3, and 98.6 ppm. HRMS (ESI +ve) m/z: observed for C15H15N2+ [M]+, 223.1224; [M]+ calcd, 223.1230.

(4-Bromobutyl)triphenylphosphonium Bromide (2)

1,4-Dibromobutane (2.7 mL, 23 mmol) was added to a stirred solution of PPh3 (3.01 g, 11.2 mmol) in toluene (25 mL) at 95 °C, and the mixture was refluxed for 20 h. The reaction mixture was then cooled to room temperature to get a white precipitate. The precipitate was filtered and washed several times with Et2O. The product was dried under vacuum to yield a white amorphous solid.

Yield: 5.056 g (91.94%).

1H NMR (500 MHz, DMSO-d6, 25 °C): δ 7.92–7.77 (15H, m), 3.70–3.61 (4H, m), 2.04–1.99 (2H, m), and 1.73–1.62 (2H, m) ppm. 13C NMR (125 MHz, DMSO-d6, 25 °C): δ 134.8, 134.7, 133.4, 130.2, 130.1, 118.8, 118.6, 118.1, 117.9, 33.5, 32.8, 32.4, 20.3, 20.2, and 18.6 ppm. HRMS (ESI +ve) m/z: observed for C22H23BrP+ [M]+, 397.1012; [M]+ calcd, 397.0716.

(6-Bromohexyl)triphenylphosphonium Bromide (3)

1,6-Dibromobutane (1.5 g, 3 mmol) was added to a stirred solution of PPh3 (0.52 g, 2.0 mmol) in toluene (35 mL) at 95 °C, and the mixture was refluxed for 20 h. The reaction mixture was thereafter cooled to room temperature, and the solvent was removed under reduced pressure to obtain a gummy solid. The gummy residue was washed with CH3CN/EtOAc (4:1) several times to get the desired product as an off-white solid.

Yield: 0.355 g (35%).

1H NMR (400 MHz, DMSO-d6, 25 °C): δ 7.90–7.74 (15H, m), 3.70–3.60 (2H, m), 3.46 (2H, t, J = 6.8 Hz), 1.73–1.69 (2H, m), and 1.41–1.30 (6H, m) ppm. 13C NMR (100 MHz, DMSO-d6, 25 °C): δ 134.73, 134.71, 133.43, 133.36, 118.70, 118.01, 28.66, 28.52, 21.35, 21.32, 20.39, and 19.99 ppm. HRMS (ESI +ve) m/z: observed for C24H27BrP+ [M]+, 425.1016; [M]+ calcd, 425.1028.

(4-(3,3-Dimethyl-2-methyleneindolin-1-yl)butyl)triphenylphosphonium Iodide (4)

2,3,3-Trimethylindolenine (0.992 g, 6.23 mmol), 2[(4-bromobutyl)triphenylphosphonium bromide] (3.28 g, 6.85 mmol), and KI (1.14 g, 6.85 mmol) were taken in a 50 mL round bottom flask and dissolved in 25 mL CH3CN at room temperature. The reaction mixture was heated under reflux for 72 h and then cooled to room temperature. The resulting solution was decanted from the solid, and the solvent was removed under reduced pressure to get a reddish-brown precipitate. The precipitate was poured in water and extracted with EtOAc (3 × 25 mL). The organic layer was separated and dried over anhydrous Na2SO4, filtered, and concentrated under reduced pressure. The residue was washed with Et2O and dried under vacuum to get the pure compound 4 as a reddish-brown solid.

Yield: 2.63 g (70%).

1H NMR (400 MHz, CD3OD, 25 °C): δ 8.01–7.63 (19H, m), 4.64–4.59 (2H, m), 3.75–3.62 (2H, m), 3.32 (3H, s), 2.36–2.28 (2H, m), 1.99–1.89 (2H, m), and 1.59 (6H, s) ppm. 13C NMR (100 MHz, CD3OD, 25 °C): δ 196.1, 165.3, 141.8, 141.1, 134.9, 134.7, 133.6, 130.1, 129.1, 123.2, 118.7, 117.6, 115.5, 54.5, 28.4, 28.2, 21.5, 20.8, 19.9, and 19.8 ppm. HRMS (ESI +ve) m/z: observed for C33H35NP+ [M]+, 476.2805; [M]+ calcd, 476.2507.

(6-(3,3-Dimethyl-2-methyleneindolin-1-yl)hexyl)triphenylphosphonium Iodide (5)

2,3,3-Trimethylindolenine (0.992 g, 6.23 mmol), 3 [(6-bromohexyl)triphenylphosphonium bromide] (3.47 g, 6.85 mmol, 1.1 equiv), and KI (1.14 g, 6.85 mmol, 1.1 equiv) were taken in a 50 mL round bottom flask and dissolved in 25 mL CH3CN at room temperature. The reaction mixture was heated under reflux for 72 h and then cooled to room temperature. The resulting solution was decanted from the solid, and the solvent was removed under reduced pressure to get a reddish-brown residue. The residue was purified through column chromatography using DCM/MeOH (95:5) (Rf = 0.46) to acquire the desired compound 5 as a reddish-brown solid.

Yield: 1.7 g (43%).

1H NMR (400 MHz, CDCl3, 25 °C): δ 7.78–7.68 (15H, m), 7.07–7.03 (2H, t, J = 9.2 Hz), 6.69 (1H, t, J = 7.2 Hz), 6.48 (1H, d, J = 7.2 Hz), 3.76 (2H, br), 3.67–3.60 (2H, m), 3.43 (2H, t, J = 6.8 Hz), 1.78–1.52 (8H, m), and 1.32 (6H, s) ppm. 13C NMR (100 MHz, CDCl3, 25 °C): δ 161.6, 146.1, 137.2, 135.1, 133.8, 130.6, 127.6, 121.8, 118.6, 118.2, 117.8, 105.3, 73.2, 44.2, 42.1, 30.1, 29.3, 26.7, 25.8, 23.4, 22.9, and 22.4 ppm. HRMS (ESI +ve) m/z: observed for C35H39NP+ [M]+, 504.2821; [M]+ calcd, 504.2820.

1-Hexyl-2,3,3-trimethyl-3H-indol-1-ium Iodide (6)

2,3,3-Trimethylindolenine (0.50 g, 3.15 mmol), 1-bromo hexane (0.57 g, 3.47 mmol), and KI (0.58 g, 3.47 mmol) were taken in a round bottom flask and dissolved in 25 mL CH3CN at room temperature. The reaction mixture was heated under reflux for 72 h and cooled to room temperature. The resulting solution was decanted from the solid, and the solvent was removed under reduced pressure to get a reddish-brown residue. The residue was washed with hexane/EtOAc mixture (4:1) several times. The crude product was purified by column chromatography using DCM/MeOH (97:3) (Rf = 0.75) to get the desired compound 6 as a brown solid.

Yield: 0.70 g (60%).

1H NMR (400 MHz, CDCl3, 25 °C): δ 7.65–7.50 (4H, m), 4.65 (2H, t, J = 7.8 Hz), 3.11 (3H, s), 1.94–1.89 (2H, m), 1.65 (6H, s), 1.48–1.42 (2H, m), 1.36–1.29 (4H, m), and 0.87 (3H, t, J = 7.0 Hz) ppm. 13C NMR (100 MHz, CDCl3, 25 °C): δ 195.68, 141.75, 141.01, 130.24, 129.60, 123.49, 115.41, 54.76, 50.20, 31.23, 28.02, 26.53, 23.27, 22.41, 17.14, and 13.95 ppm. HRMS (ESI +ve) m/z: observed for C17H26N+ [M]+, 244.2056; [M]+ calcd, 244.2060.

Cy-5a dye

Compound 4 (0.56 g, 0.93 mmol), 1 (0.12 g, 0.465 mmol), and NaOAc (0.12 g, 1.4 mmol) were heated to 70 °C for 6 h in Ac2O (5 mL). The reaction mixture was allowed to cool to room temperature, and the solvent was removed under reduced pressure. Et2O was added to get a blue color residue, and this solution was kept overnight in the refrigerator. The blue solid was collected by filtration and washed with ether (3 × 5 mL). The crude product was purified by silica gel column chromatography using DCM/MeOH (95:5) (Rf = 0.34) to obtain the desired compound Cy-5a as a blue solid.

Yield: 0.186 g (32%).

1H NMR (500 MHz, CDCl3, 25 °C): δ 7.87–7.65 (34H, m), 7.58 (2H, d, J = 8.0 Hz), 7.38 (2H, t, J = 8.0 Hz), 7.27–7.22 (1H, m), 7.19–7.14 (2H, m), 6.48 (2H, d, J = 13.2 Hz), 4.23 (4H, t, J = 8.0 Hz), 3.82–3.74 (4H, m), 2.24–2.18 (4H, m), 2.04–1.98 (4H, m), and 1.62 (12H, s) ppm. 13C NMR (125 MHz, CDCl3, 25 °C): δ 172.39, 152.94, 141.97, 140.73, 135.24, 135.21, 134.15, 134.05, 130.76, 129.33, 127.41, 125.17, 121.84, 118.42, 117.57, 112.50, 104.32, 49.26, 44.15, 29.80, 28.19, 23.44, 22.93, and 20.66 ppm. 31P NMR (202 MHz, CDCl3, 25 °C): δ 24.78 ppm. HRMS (ESI +ve) m/z: observed for C69H71N2P23+ [M]3+, 329.8297; [M]3+ calcd, 329.8359.

Cy-5b dye

Compound 5 (0.652 g, 1.03 mmol), 1 (0.134 g, 0.52 mmol), and NaOAc (0.128 g, 1.56 mmol) were heated to 70 °C for 6 h in Ac2O (5 mL) under N2 atmosphere. The reaction mixture was then allowed to cool, and the solvent was removed under reduced pressure. Et2O was added to get a blue color residue, and this solution was kept overnight in the refrigerator. The blue solid was collected by filtration and washed with Et2O (3 × 5 mL). The crude product was purified by silica gel column chromatography using DCM/MeOH (92:8) (Rf = 0.30) to get the desired compound Cy-5b as a blue solid.

Yield: 0.199 g (30%).

1H NMR (500 MHz, CDCl3, 25 °C): δ 7.91 (2H, t, J = 13.0 Hz), 7.84–7.70 (30H, m), 7.38–7.35 (2H, m), 7.30 (4H, t, J = 7.5 Hz), 7.18 (2H, t, J = 7.5 Hz), 7.04 (1H, t, J = 12.5 Hz), 6.37 (2H, d, J = 14.0), 4.11 (4H, t, J = 7.0 Hz), 3.71–3.65 (4H, m), 1.81–1.74 (12H, m), 1.69 (12H, s), and 1.68–1.62 (4H, m) ppm. 13C NMR (125 MHz, CDCl3, 25 °C): δ 172.71, 153.19, 142.17, 141.20, 135.23, 133.96, 133.89, 130.83, 130.73, 129.05, 125.17, 122.11, 118.67, 117.99, 111.59, 104.00, 49.42, 44.51, 29.79, 28.33, 27.17, 26.09, 23.33, and 22.60 ppm. 31P NMR (202 MHz, CDCl3, 25 °C): δ 24.24 ppm. HRMS (ESI +ve) m/z: observed for C73H79N2P23+ [M]3+, 348.5202; [M]3+ calcd, 348.5234.

Cy-5c dye

Compound 6 (0.244 g, 0.66 mmol), 1 (0.086 g, 0.33 mmol), and NaOAc (0.081 g, 0.99 mmol) were heated to 70 °C for 6 h in Ac2O (5 mL) under N2 atmosphere. The reaction mixture was then allowed to cool, and the solvent was removed under reduced pressure. Et2O was added to get a blue color residue and this solution was kept overnight in the refrigerator. The residue was collected by filtration. The crude product was purified by column chromatography using DCM/MeOH (96:4) (Rf = 0.63) to obtain the pure compound Cy-5c as a blue solid.

Yield: 0.108 g (50%).

1H NMR (500 MHz, CDCl3, 25 °C): δ 8.24 (2H, t, J = 13.0 Hz), 7.37–7.33 (4H, m), 7.21 (2H, t, J = 7.5 Hz), 7.06 (2H, d, J = 8.0 Hz), 6.78 (1H, t, J = 13.0 Hz), 6.27 (2H, d, J = 13.5 Hz), 4.04 (4H, t, J = 7.5 Hz), 1.84–1.80 (4H, m), 1.79 (12H, s), 1.47–1.41 (4H, m), 1.37–1.25 (8H, m), and 0.88 (6H, t, J = 7.0 Hz). 13C NMR (125 MHz, CDCl3, 25 °C): δ 173.28, 154.08, 142.25, 141.67, 128.60, 126.35, 125.21, 122.49, 110.55, 103.78, 49.68, 44.68, 31.59, 28.30, 27.51, 26.74, 22.56, and 14.07 ppm. HRMS (ESI +ve) m/z: observed for C37H51N2+ [M]+, 523.4341; [M]+ calcd, 523.4047.

1H–1H DQF COSY

The full 1H NMR assignments including 1H–1H correlation of Cy-5a, Cy-5b, and Cy-5c dyes were performed using the 1H–1H DQF COSY experiment (Figures S4, S7, and S9b).

Methods

NMR Spectroscopy

1H, 13C, 31P NMR, and 1H–1H DQF COSY spectra were recorded on Bruker DPX400 MHz and Bruker DPX500 MHz spectrometers at 25 °C in suitable deuterated solvents.

HRMS (ESI)

High-resolution electrospray ionization mass spectrometry (HRMS-ESI) result was obtained using a Q-Tofmicro (Waters Corporation) mass spectrometer.

Absorption Spectroscopy

Absorption spectra were recorded in various solvents on a Shimadzu UV-1800 spectrometer. All measurements were carried out in a quartz cuvette with a path length of 1 cm. Stabilities of the Cy-5a and Cy-5b dyes were measured in PBS at pH 7.4, 37 °C, over 24 h using a Peltier temperature controlling unit attached with the UV/vis instrument.

Fluorescence Spectroscopy

Fluorescence experiment was performed on a HORIBA Jobin Yvon FluoroMax-4 spectrofluorometer in various solvents.

Time-Correlated Single Photon Counting Experiment

Fluorescence lifetimes of the probes were measured in various solvents using the TCSPC technique on a HORIBA DeltaFlex lifetime instrument (HORIBA Jobin Yvon IBH Ltd, Glasgow, Scotland, UK). A 650 nm delta diode laser excitation source (model: DD-650L, HORIBA Scientific) was used to measure fluorescence lifetimes of Cy-5a and Cy-5b. Data analysis and lifetime measurements were carried out using the HORIBA EzTime decay analysis software.

pH Meter

A Mettler Toledo pH meter was used to prepare PBS and HEPES buffer solutions at pH 7.4.

Calculation of Relative Quantum Yield of Cy-5a and Cy-5b Dyes

Fluorescence quantum yields (Φf) of Cy-5a and Cy-5b dyes were measured by the relative method.[12] Here, the integrated fluorescence intensities of the samples were compared with fluorescence intensities of a reference compound according to the following equationΦf(st) and Φf(x) are fluorescence quantum yields of the reference and sample compounds respectively; Ast and Ax are absorbance of reference and sample at the excitation wavelength, respectively. Fst and Fx are the integrated fluorescence areas under the corrected fluorescence spectra for the reference and sample compounds, respectively. ηst and ηx are the refractive indices of the solvent in which the reference and sample compounds are measured, respectively [here, both the reference and sample compounds were dissolved in DMSO; therefore, (ηx2/ηst2) = 1]. “st” stands for the standard and “x” refers to the unknown sample. Φf(st) of zinc pthalocyanine in DMSO is 0.20, which is used as the reference.

The relative quantum yields (Φf) of Cy-5a and Cy-5b were measured as 0.30 and 0.24, respectively, in DMSO.

Water Solubility of Cy-5a and Cy-5b Dyes

Cy-5a and Cy-5b dyes were separately dissolved in water with the maximum known amount, and the solubility was determined by the UV/vis spectroscopic method. With the known concentration of Cy-5a and Cy-5b, an absorbance versus concentration calibration curve was plotted for both dyes. Then, the absorbance of the saturated solution of both dyes with proper dilution was measured and using the absorbance versus concentration calibration curve, corresponding concentration of the dyes was determined. This gave the measure of solubility of the Cy-5a and Cy-5b dyes in water. Cy-5a and Cy-5b are small molecules and soluble in water because of triply positive charge.

Determination of Octanol/PBS Partition Coefficient Values of Cy-5a and Cy-5b Dyes

Octanol/PBS partition coefficient values of Cy-5a and Cy-5b were measured according to the literature procedure.[8] To measure the partition coefficient of Cy-5 dye, 1 mg of each Cy-5a and Cy-5b was separately dissolved in 2 mL octanol. Absorbance of the stock solution in octanol for Cy-5a and Cy-5b was determined separately by UV/vis spectroscopy using proper dilution. PBS (pH 7.4, 1 mL) and equal volume of stock solution in octanol were mixed and vortexed for 1 m, and kept at 37 °C for 1 h to achieve equilibrium. The mixture was centrifuged at 2500 rpm for 10 m to separate the organic and aqueous layers. After centrifugation, absorbance of the organic (Ao) layer was determined by UV/vis spectroscopy with appropriate dilution. The ratio of absorbance in organic layer (after mixing) to that of an aqueous layer (determined from difference in absorbance value of organic layer before and after mixing with PBS) gives the partition coefficient value and it is expressed as Ao/Aw.where (Ao)i and (Ao)f are initial and final absorbance in the octanol layer, respectively.

Partition coefficient (P) values of Cy-5a and Cy-5b were found to be 2.88 and 7.12, and log P values were 0.46 and 0.85, respectively.

Preparation of Lipid Film

DOPC (10 mg mL–1) and cholesterol (10 mg mL–1) were dissolved in CHCl3. Appropriate quantities were mixed to get a defined composition of DOPC/Ch = 80:20. The solvent (CHCl3) was removed by N2 stream for 20 min, and the residual solvent was removed under high vacuum for 2 h to form DOPC/Ch lipid film on the test tube walls. The lipid films were rehydrated with an appropriate amount of HEPES buffer solution (20 mM, 150 mM NaCl, pH 7.4).

Multilamellar Vesicles

After 1 h of incubation at T > tm, the hydrated lipid films were subjected to five freeze/thaw cycles (immerse the sample in liquid N2 followed by 40 °C water bath) to detach the lipid film from the wall of the test tube, while also dispersing the vesicles. The dispersed multilamellar vesicles (MLVs) were incubated at T > tm for 5 min and vortexed.

Large Unilamellar Vesicles

The MLV suspensions were extruded 31 times through a polycarbonate membrane (200 nm pore size, 19 mm diameter) using a LiposoFast extruder (Avestin, Ottawa, Canada) at T > tm to produce unilamellar vesicle suspension.

Dynamic Light Scattering

Dynamic light scattering experiments were performed using a Malvern instrument (UK) to examine the hydrodynamic diameters of the liposomes in HEPES buffer at pH 7.4. All measurements were conducted at a backscattering angle 173° and a temperature of 25 °C. Three runs were conducted per measurement and the average values were taken. A monodisperse population of the liposome with hydrodynamic diameter = 168 nm and PDI = 0.115 was observed in HEPES buffer.

Cell Culture

Human pancreatic cancer cells (MIA PaCa-2), lung carcinoma (A549) cells, noncancerous PBMC, human embryonic kidney (HEK293) cells, and lung fibroblast (WI38) cells were cultured in DMEM media (pH 7.4) supplemented with 10% FBS and antibiotic-antimycotic solution 100× (containing 10 000 units penicillin, 10 mg streptomycin, and 25 μg amphotericin B per mL in 0.9% normal saline). The cell lines were maintained at 37 °C in an air-jacketed 5% CO2 incubator and were routinely passaged.

MTT Assay for Cell Viability

All cells were separately plated at a density of ∼100 cells in a 96-well plate separately using DMEM media with a 24 h incubation to allow appropriate cell growth. The cytotoxic effects of Cy-5a and Cy-5b on the various cancerous and noncancerous cell lines were determined by MTT assay. After the 24 h incubation, Cy-5 dye was treated at specified concentrations (0.25, 0.5, and 0.75 μM) for 24 h at 37 °C. The cells were then treated with 10 μL of MTT solution (5 mg mL–1 in PBS) for 4 h in darkness at 37 °C. Dark blue formazan crystals was formed which were dissolved in DMSO, and the absorbance (A) was measured at 575 nm using an ELISA plate reader. The results were expressed as the percentages of the viable cells by the following equation:

Flow Cytometric Analyses

Flow cytometric analyses were carried out by a BD FACSVerse flow cytometer (BD Biosciences, San Jose, CA), and the data were analyzed by BD Cell Quest software (BD Biosciences).

Cell Penetration Dose Determination

A549 cells (104 in number) were treated with different doses of Cy-5 dyes for 10 m in darkness. Then, the cells were washed with PBS and resuspended in 400 μL PBS. The samples were subjected to flow cytometric analysis using BD FACSVerse instrument. The overall fluorescence intensity of dye-positive cells was measured, and the optimal concentration was calculated from the titration.

Confocal Microscopy

Fluorescence confocal microscopic images were acquired with an Olympus Instrument (model IX81) equipped with a 60× and 100× oil plan apochromatic objective with numerical aperture = 1.49. Colocalization experiments involving Cy-5a and Cy-5b (λex/λem 650/675) were performed in the fixed A549 carcinoma cell line using a mitochondrion-tracking fluorescent probe MitoTracker Red (λex/λem 579/599).[13] The A549 cells were seeded on the cover slip and grown in DMEM media. Cells were fixed for 30 min with 4% paraformaldehyde and subjected to confocal microscopic studies. Fixed cells were incubated with 0.7 μM solution of Cy-5 dye at 37 °C for 10 m in darkness and then washed twice using 1× PBS. Cells were then incubated with 0.2 μM MitoTracker Red in darkness for 15 m, washed twice with 1× PBS, and incubated with DAPI for 15 m. After washing, following the same procedure, cells were mounted with antiquench agent n-propyl gallate for microscopic slide preparation and detected by a confocal fluorescence microscope. High-resolution images were analyzed by Olympus FV1000 software.

For DAPI: laser wavelength = 405 nm, excitation wave length = 405 nm, emission wavelength = 461 nm, BF range = 50 nm, BF position = 425 nm. For MitoTracker Red: laser wave length = 559 nm, excitation wave length = 559 nm, emission wavelength = 598 nm, BF range = 55 nm, BF position = 570 nm. For Cy-5 dye: laser wavelength = 635 nm, excitation wave length = 635 nm, emission wavelength = 668 nm, BA name = 655–755.

Mitochondrial Selectivity of Cy-5a and Cy-5b Over Lysosome

Colocalization experiments were performed using a green-fluorescent lysosome tracking probe, LysoTracker Green DND-26 (Figure S1). Fixed cells were incubated with 0.7 μM solution of Cy-5 dye at 37 °C for 10 m in darkness and then washed twice using 1× PBS. Cells were then incubated with 50 nM LysoTracker Green in darkness for 15 m, washed twice with 1× PBS, and incubated with DAPI for 15 m. After washing with the same method, cells were mounted with n-propyl gallate for microscopic slide preparation and detected by confocal fluorescence microscope. High-resolution images were analyzed by Olympus FV1000 software.

For LysoTracker Green DND-26: laser wave length = 473 nm, excitation wave length = 473 nm, emission wavelength = 519 nm, BF range = 100 nm, BF position = 485 nm.

Multicolor Confocal Imaging

Multicolor imaging of cellular organelles in the same cell was achieved using a combination of suitable targeting dyes with distinct excitation and emission bands. Lipophilic green emitting FM 1-43FX was used to selectively label the plasma membrane, blue-fluorescent DAPI to stain the nucleus, and Cy-5 to target and stain the mitochondria of A549 cells. Multicolor confocal imaging of cells was acquired by using appropriate filters for FM 1-43FX (laser ex/em 473/590 nm), DAPI (laser ex/em 405/461 nm), and Cy-5 dye (laser ex/em 635/668 nm). Fixed cells were incubated with 0.7 μM solution of Cy-5 dye at 37 °C for 10 m in darkness and then washed twice using 1× PBS. Cells were then incubated with DAPI for 15 m followed by incubation with 8 μM FM 1-43FX in darkness for 1 m. After washing, following the same process, cells were mounted with n-propyl gallate for microscopic slide preparation and detected by confocal fluorescence microscope. High-resolution images were analyzed by Olympus FV1000 software.

Determination of Mitochondrial Membrane Potential Using JC1 Dye

Alteration in mitochondrial membrane potential was determined with JC-1 dye (mitochondrion membrane potential assay kit) using a FACS flow cytometer, and the data were analyzed by Cell Quest software (BD Biosciences).[10] When the potential of mitochondria was high, JC-1 was entered from the cell membrane to the cytosol to the mitochondria and the J-aggregates of JC-1 were formed. However, in case of membrane depolarization, the J-aggregates leak out from the mitochondria to the cytosol as monomers. The λex for JC-1 was 494 nm and the λem was observed at 535 nm (green fluorescence for JC-1 monomers) and at 595 nm (orange-red fluorescence for JC-1 aggregates). Then, the fluorescence emission ratio of monomers to aggregates (535/595) was evaluated. The higher value of this ratio indicates higher membrane depolarization. A549 cells were stained with JC-1 dye for 15 min with 1 mg mL–1 JC-1 in the culture medium at 37 °C in darkness; then, mitochondrial membrane potential was determined using the FACSVerse flow cytometer and the data were analyzed by Cell Quest software (BD Biosciences). After cellular staining with JC-1 dye, the fluorescent signal was analyzed in the FITC and PE channels using a 494 nm laser.