Rational Design and Synthesis of Large Stokes Shift 2,6-Sulphur-Disubstituted BODIPYs for Cell Imaging

Abstract

Five new disubstituted 2,6-thioaryl-BODIPY dyes were synthesized via selective aromatic electrophilic substitution from commercially available thiophenols. The analysis of the photophysical properties via absorption and emission spectroscopy showed unusually large Stokes shifts for BODIPY fluorophores (70–100 nm), which makes them suitable probes for bioimaging. Selected compounds were evaluated for labelling primary immune cells as well as different cancer cell lines using confocal fluorescence microscopy.

1. Introduction

Compounds containing chalcogen elements—for instance, selenium, tellurium, and sulphur—have wide applications in many areas of science [1,2,3]. Importantly, they can act as redox centers and modulate the activity of biomolecules [4,5,6]. Chalcogen atoms can be also introduced in fluorescent structures to fine-tune their optical properties, including absorption/emission wavelengths and photodynamic capabilities [7,8]. In recent years, our research group and others have focused on the preparation of several classes of fluorescent compounds [9,10,11,12,13,14,15,16]. Antitumor compounds with fluorescent properties, optical probes activated by enzymes [17,18,19,20,21,22], fluorescent amino acids and synthetic peptides for imaging studies [23,24,25,26,27,28,29,30], and more recently, boron-dipyrromethene (BODIPY) derivatives [31,32,33,34,35,36,37,38,39,40] have been the targets of many investigations.

BODIPY dyes are versatile fluorescent scaffolds due to their remarkable photophysical properties. These include relatively large molar absorption coefficients, high fluorescence quantum yields, sharp emission bandwidths, and high photostability [41,42,43]. Moreover, the chemical versatility of the BODIPY core allows their spectroscopic and photophysical properties to be fine-tuned. This family of compounds has been extensively employed in many research fields, including chemosensors [44], photosensitizers [45,46,47,48,49], and OLEDs [50].

The functionalization of BODIPYs with chalcogen groups has multiple motivations, such as the enhancement of their physicochemical properties [51]. Among the BODIPY derivatives already reported in the literature [52], BODIPYs containing sulphur [53,54,55], selenium [56,57], and tellurium [58] atoms have been reported for the detection of hypochlorite or hypochlorous acid, among some applications.

Synthetic methods for the insertion of sulphur atoms into the basic architecture of BODIPYs have been described (Scheme 1); however, many of them require harsh reaction conditions [55]. Rezende and coworkers developed a methodology for the thiocyanation of 3-substituted and 3,5-disubstituted BODIPYs using oxone and ammonium thiocyanate [59]. Kim and coworkers prepared an alkylthioether BODIPY derivatives via an electrophilic substitution reaction in the presence of DMSO and POCl₃ [60]. Bröring’s group described other interesting examples using S₂Cl₂ to dimerize BODIPY moieties as sulfide analogues [61].

Upon compiling existing data on the photophysical properties of the above-mentioned BODIPYs, we observed that while having variable excitation and emission maxima and photoluminescence quantum yields (Φ_PL), all 2- and 2,6-thioether-substituted compounds exhibited large Stokes shifts > 50 nm [60,61] (Scheme 1B), partially due to electron delocalization, whereas the 3-, 3,5- and meso-thioether-substituted BODIPYs displayed smaller excitation–emission gaps of 10 to 30 nm (Scheme 1A) [53,54,55]. Because large Stokes shifts can improve signal-to-noise ratios or allow excitation of multiple dyes with the same laser source, they can be advantageous for bioimaging applications. Therefore, we designed new 2,6-sulphur-disubstituted BODIPYs based on thiophenol derivatives. Herein we present an optimized 2-step synthesis from readily available building blocks, their photophysical properties in different solvents, and potential use as probes for live-cell imaging (Scheme 1C).

2. Materials and Methods

2.1. Chemistry

All reagents and solvents used were of analytical grade. ¹H, ¹³C, and ¹⁹F spectra were recorded on a Varian VNMRS 300 MHz and Varian VNMRS 500 MHz spectrometer, using tetramethylsilane (TMS) or trifluoroacetic acid (TFA) as reference. Spectra were recorded typically at r.t. in CDCl₃. Chemical shifts are reported in ppm (δ) and coupling constants (J) are given in Hz. Column chromatographic purifications were carried out on silica gel SiliaFlash G60 70–230 mesh (SiliCycle) columns with the defined eluents. NMR spectra and HRMS data are presented in Supporting Information.

4,4-Difluoro-8-(4-methoxyphenyl)-1,3,5,7-tetramethyl-4-bora-3a,4a-diaza-S-indacene (1): 2,4-dimethyl-1H-pyrrole (1.03 mL, 10.0 mmol) and 4-methoxybenzaldehyde (0.61 mL, 5.0 mmol) were mixed with a pestle and mortar. Trifluoroacetic acid was added and the mixture was ground for 2 min. CHCl₃ (2.0 mL) was added and then p-chloranil (1.81 g, 7.4 mmol) was added to the mixture and ground for 2 min. Next, triethylamine (6.0 mL, 43.0 mmol) was added and the resulting paste was ground for 3 min. BF₃·OEt₂ (6.0 mL, 47.4 mmol) was then added dropwise and the mixture was ground for 2 min until a dark red paste formed. The reaction mixture was dissolved in CHCl₃ (200 mL) and washed with saturated Na₂CO₃ (3  ×  200 mL) and brine (2  ×  200 mL). The solvent was evaporated and the crude solid was purified by normal-phase chromatography with hexane:chloroform (9:1). The product was obtained as an orange solid (45% yield). ¹H NMR (500 MHz, CDCl₃) δ: 7.18–7.16 (d, J = 8.7 Hz, 2H), 7.02–7.00 (d, J = 8.7 Hz, 2H), 5.97 (s, 2H), 3.87 (s, 3H), 2.55 (s, 6H), 1.43 (s, 6H) ppm. ¹⁹F NMR (470 MHz, CDCl₃) δ: −143.5 (q, J_BF = 32.9 Hz) ppm. ¹³C NMR (75 MHz, CDCl₃) δ: 160.27, 155.39, 143.29, 141.99, 131.98, 127.1, 121.2, 114.6, 55.4, 14.6 ppm. The data are consistent with those reported in the literature [62].

Synthesis of thioaryl-BODIPYs (2a–e): To a suspension of N-chlorosuccinimide (267 mg, 2 mmol) in CH₂Cl₂ (10 mL) at r.t., we carried out a dropwise addition of a solution of the corresponding thiophenol (2 mmol) in dry CH₂Cl₂ (10 mL). After stirring for 30 min, the orange solution of the phenyl-sulfenyl chloride was used in situ. To the solution of phenyl-sulfenyl chloride at 0 °C, a solution of BODIPY 1 (50.0 mg, 0.135 mmol) in 10 mL of dry CH₂Cl₂ was added dropwise over 15 min. The mixture was stirred at r.t. for 15 min. Next, H₂O (20 mL) was added to the reaction mixture. After separation, the organic phase was washed with H₂O (20 mL), dried over Na₂SO₄, and the solvent was evaporated under reduced pressure. The residue was purified by column chromatography using hexane:CH₂Cl₂ (8:1) to give the products 2a–e.

4,4-Difluoro-8-(4-methoxyphenyl)-2,6-bis(phenylthio)-1,3,5,7-tetramethyl-4-bora-3a,4a-diaza-S-indacene (2a): The product was obtained as a red solid (94% yield; m.p. (°C) 225–228; ¹H NMR (500 MHz, CDCl₃) δ: 7.22–7.19 (m, 6H), 7.11–7.08 (t, J = 7.4 Hz, 2H), 7.04–7.03 (d, J = 8.7 Hz, 2H), 7.01–7.00 (d, J = 7.3 Hz, 4H), 3.87 (s, 3H), 2.62 (s, 6H), 1.54 (s, 6H). ¹⁹F NMR (470 MHz, CDCl₃) δ: −143.11 (q, J = 31.7 Hz). ¹³C NMR (75 MHz, CDCl₃) δ: 160.6, 160.1, 148.5, 143.3, 137.5, 132.0, 129.1, 126.7, 126.0, 125.3, 115.0, 77.1, 55.5, 13.6, 13.3. ESI⁺/HRMS (m/z) [M + H]⁺: 571.1855 Cald. for C₃₂H₃₀BF₂N₂OS₂, found: 571.1850. Data are consistent with those reported in the literature.

4,4-Difluoro-8-(4-methoxyphenyl)-2,6-bis(2-methylphenylthio)-1,3,5,7-tetramethyl-4-bora-3a,4a-diaza-S-indacene (2b): The product was obtained as a red solid (83% yield); m.p. (°C) 223-229; ¹H NMR (300 MHz, CDCl₃) δ: 7.25–7.22 (d, J = 8.7 Hz, 2H)), 7.14–7.11 (m, 2H), 7.05–7.00 (m, 6H), 6.63-6.60 (m, 2H), 3.78 (s, 3H), 2.52 (s, 6H), 2.33 (s, 6H), 1.43 (s, 6H).¹⁹F NMR (470 MHz, CDCl₃) δ: −143.11 (q, J = 31.7 Hz). ¹³C NMR (75 MHz, CDCl₃) δ: 160.6, 160.1, 148.7, 143.1, 136.5, 134.8, 130.2, 129.2, 126.7, 126.6, 124.9, 124.7, 115.0, 77.1, 55.5, 20.0, 13.5, 13.3. ESI⁺/HRMS (m/z) [M + H]⁺: 599.2168 Cald. for C₃₄H₃₄BF₂N₂OS₂, found: 599.2162.

4,4-Difluoro-8-(4-methoxyphenyl)-2,6-bis(4-methylphenylthio)-1,3,5,7-tetramethyl-4-bora-3a,4a-diaza-S-indacene (2c): The product was obtained as a pink solid (63% yield); m.p. (°C) 225–229; ¹H NMR (500 MHz, CDCl₃) δ: 7.21–7.20 (d, J = 8.6 Hz, 2H), 7.04–7.01 (m, 6H), 6.92–6.91 (d, J = 8.2 Hz, 4H), 3.86 (s, 3H), 2.27 (s, 6H), 1.53 (s, 6H). ¹⁹F NMR (470 MHz, CDCl₃) δ: −148.08 (q, J = 31.5 Hz). ¹³C NMR (125 MHz, CDCl₃) δ: 160.6, 160.0, 148.3, 143.2, 135.3, 133.9, 132.0, 129.9, 129.2, 126.8, 126.4, 115.0, 55.5, 13.6, 13.4. ESI⁺/HRMS (m/z) [M + H]⁺: 599.2168 Cald. for C₃₄H₃₄BF₂N₂OS₂, found: 599.2163.

4,4-Difluoro-8-(4-methoxyphenyl)-2,6-bis(4-methoxiphenylthio)-1,3,5,7-tetramethyl-4-bora-3a,4a-diaza-S-indacene (2d): The product was obtained as a deep red solid (90% yield); m.p. (°C) 234–238; ¹H NMR (500 MHz, CDCl₃) δ: 7.19–7.18 (d, J = 8.7 Hz, 2H), 7.03–7.02 (d, J = 8.7 Hz, 2H), 7.01–6.99 (d, J = 8.9 Hz, 4H), 6.78–6.76 (d, J = 8.9 Hz, 4H), 3.79 (s, 3H), 3.68 (s, 6H), 2.55 (s, 6H), 1.46 (s, 6H). ¹⁹F NMR (470 MHz, CDCl₃) δ: −143.17 (q, J = 31.7 Hz). ¹³C NMR (125 MHz, CDCl₃) δ: 160.6, 159.7, 158.2, 147.7, 143.0, 131.8, 129.2, 128.8, 127.9, 126.8, 122.3, 114.9, 114.8, 77.1, 55.5, 13.6, 13.4. ESI⁺/HRMS (m/z) [M + H]⁺: 631.2066 Cald. for C₃₄H₃₄BF₂N₂O₃S₂, found: 631.2057.

4,4-Difluoro-8-(4-methoxyphenyl)-2,6-bis(4-chlorophenylthio)-1,3,5,7-tetramethyl-4-bora-3a,4a-diaza-S-indacene (2e): The product was obtained as a deep red solid (52% yield); m.p. (°C) 230–235; ¹H NMR (500 MHz, CDCl₃) δ: 7.21–7.19 (d, J = 8.8 Hz, 2H), 7.18–7.17 (d, J = 8.7 Hz, 4H), 7.05-7.03 (d, J = 8.7 Hz, 2H), 6.93–6.91 (d, J = 8.7 Hz, 4H), 3.52 (s, 3H), 2.57 (s, 6H), 1.47 (s, 6H). ¹⁹F NMR (470 MHz, CDCl₃) δ: −148.01 (q, J = 31.7 Hz). ¹³C NMR (125 MHz, CDCl₃) δ: 160.7, 160.0, 148.6, 143.6, 136.0, 132.1, 131.2, 129.2, 129.1, 127.3, 126.5, 120.3, 115.1, 77.1, 55.5, 13.5, 13.3. ESI⁺/HRMS (m/z) [M + H]⁺: 639.1076 Cald. for C₃₂H₂₈BCl₂F₂N₂OS₂, found: 639.1076.

2.2. Photophysical Measurements

Absorption spectra were obtained on a UV-visible spectrophotometer (Thermo Scientific Evolution 600) at r.t. Fluorescence excitation and emission spectra were recorded on a Varian Cary Eclipse spectrofluorometer using 1 cm pathlength cuvettes at r.t. Fluorescence quantum yields (Φ_F) of the new compounds were obtained from a comparative method using Rhodamine 6G (Φ_F = 0.95 in ethanol, λ_exc = 500 nm) as a standard [63]. In this method, a series of solutions in different concentrations were used for measurements of absorbance and fluorescence emission, keeping a low absorbance (A < 0.1) to avoid inner-filter effects and intermolecular reabsorption [64]. The quantum yield was calculated by plotting the integrated fluorescence intensity vs. absorbance to obtain the slope of the curve. The photoluminescence quantum yield of the tested compound (Φ_x) was calculated using the following equation

$${\mathsf{\Phi }}_{\mathrm{x}}={\mathsf{\Phi }}_{\mathrm{st}}\left(\frac{{\mathrm{Grad}}_{\mathrm{x}}}{{\mathrm{Grad}}_{\mathrm{st}}}\right)\left(\frac{{\mathsf{\eta }}_{\mathrm{x}}^2}{{\mathsf{\eta }}_{\mathrm{st}}^2}\right)$$

(1)

where Φ_st is the quantum yield of the standard, Grad_x and Grad_st are the slopes for the test and standard compounds, respectively, and η_x and η_st are the refractive indexes of the solvents.

2.3. Biological Experiments

Viability assays: Cell lines were cultured in DMEM supplemented with 10% FBS, antibiotics (100 U mL⁻¹ penicillin and 100 mg mL⁻¹ streptomycin) and 1% L-Glutamine at 37 °C in a humidified atmosphere with 5% CO₂. Cells were regularly passaged in T-75 flasks using trypsin-EDTA. Cells were plated at a density of 10,000 cells per well and incubated overnight with the fluorophores (10 μM). Cell viability was assessed using an MTT cell proliferation assay following the manufacturer’s instructions. Values were normalized to the viability of untreated cells and analysed by GraphPad Prism.

Work with peripheral blood cells: The study protocol to work with human peripheral blood leukocytes was approved by the Accredited Medical Regional Ethics Committee (AMREC, 20-HV-069, University of Edinburgh). Human blood leukocytes were isolated from healthy volunteers and processed as described previously [65]. Briefly, whole blood was anticoagulated with sodium citrate 0.4% (w/v) and centrifuged at 350×  g for 20 min. Polymorphonuclear cells (>95% neutrophils) were harvested from the 63%/72.9% interface and cultured in RPMI with 5% FBS, 100 U mL⁻¹ penicillin, and 100 µg mL⁻¹ streptomycin. Neutrophils were stimulated with 10 nM PMA for 3 h at 37 °C. Monocytes were separated from the 49.5%/63% interface and cultured for 10 days in IMDM supplemented with 5% autologous serum to yield monocyte-derived macrophages (MDMs). Cells were plated and cultured with or without 5 ng/mL LPS at 37 °C.

2.4. Fluorescence Confocal Microscopy

Cells were plated in glass chamber slides (Nunc^TM Lab-Tek^TM II) and incubated with compounds 1 or 2a (10 μM) and Hoechst 33342 (0.1 µg/mL) for 15 min. Cells were washed and imaged using a Leica SP8 confocal microscope using 488 nm laser for excitation. All resulting images were analysed using FIJI.

2.5. Crystallography

Single, clear, light, colourless block-shaped crystals of 2d were obtained by recrystallization from slow evaporation. A suitable crystal 0.22 × 0.20 × 0.15 mm³ was selected and mounted on a suitable support on an XtaLAB Mini (ROW) diffractometer. The crystal was kept at a steady T = 293(2) K during data collection. The structure was solved with the ShelXT [66] structure solution program using the intrinsic phasing solution method and by using Olex2 [67] as the graphical interface. The model was refined with version 2017/1 of ShelXL 2017/1 [68] using least-squares minimization.

3. Results

The synthesis of the target compounds started with the preparation of BODIPY structure 1 as shown in Scheme 2. Compound 1 was synthesized from 2,4-dimethyl-1H-pyrrole and 4-methoxybenzoldehyde condensation, followed by oxidation with p-chloranil and complexation with BF₃·OEt₂. The one-pot reaction was performed by grinding the starting materials using a simple mortar affording the compound 1, and the optimized reaction procedure raised the recovery yields from 19 to 45% [62]. The use of neat conditions also led to faster reaction times (15 min) in comparison to conventional BODIPY synthesis.

With the BODIPY derivative 1 in hand, we prepared 2-, and 6-thioaryl-substituted BODIPY dyes as outlined in Scheme 3. Arylthiol chlorides were prepared in situ from N-chlorosuccinimide and commercially available thiophenols, and then used for the selective electrophilic substitution of the positions 2 and 6 of compound 1, which are known to be more susceptible to electrophilic attack [59]. The new 2,6-substituted thioaryl-BODIPYs 2a–e were obtained in moderate to high yields (52–94%), and fully characterized by ¹H, ¹³C and ¹⁹F NMR spectroscopy and high-resolution mass spectrometry. Suitable crystals of compound 2d were obtained by the slow evaporation method using chloroform as solvent. With appropriate red crystals in hand, the structure of 2d was solved and reconfirmed by X-ray crystallographic analysis. Representative Oak Ridge thermal ellipsoid plot (ORTEP)-3 diagram for 2d is displayed in Figure S1 and the crystallographic data is shown in Table S1.

The photophysical properties (absorption/emission wavelengths and relative photoluminescence quantum yields) of compounds 2a–2e were determined in different relevant solvents, namely toluene, tetrahydrofuran (THF), acetone, methanol (MeOH), dimethylsulfoxide (DMSO), and water (including 1% DMSO). These results are summarized in Figure 1 and

Table 1. Photophysical properties of BODIPY compounds (2a–2e) and the standard BODIPY FL.

Compound	Solvent	λabs [nm]	λem [nm]	Stokes Shift [cm−1]	ΦPL
2a	Toluene	521	595	2387	0.27
	THF	520	593	2340	0.18
	Acetone	516	597	2629	0.1
	MeOH	516	600	2713	0.1
	DMSO	517	605	2813	0.07
	Water (1% DMSO)	516	606	2878	0.02
2b	Toluene	523	599	2426	0.24
	THF	520	599	2536	0.13
	Acetone	519	600	2601	0.08
	MeOH	517	602	2731	0.05
	DMSO	518	613	2991	0.07
	Water (1% DMSO)	519	612	2928	0.03
2c	Toluene	524	599	2389	0.2
	THF	521	602	2583	0.12
	Acetone	518	607	2831	0.05
	MeOH	519	607	2793	0.05
	DMSO	522	616	2923	0.02
	Water (1% DMSO)	520	614	2944	< 0.01
2d	Toluene	519	618	3087	0.14
	THF	517	619	3187	0.03
	Acetone	516	619	3225	< 0.01
	MeOH	514	619	3300	< 0.01
	DMSO	513	620	3364	< 0.01
	Water (1% DMSO)	515	622	3340	< 0.01
2e	Toluene	521	590	2245	0.26
	THF	517	589	2364	0.2
	Acetone	512	592	2639	0.12
	MeOH	515	591	2497	0.11
	DMSO	518	599	2610	0.05
	Water (1% DMSO)	513	600	2826	0.01
BODIPY-FL	DMSO	503	509	234	0.97

. In general, no relevant solvatochromic properties were observed (Figure S2 and Table 1). All compounds presented a large Stokes shift (70–100 nm, 2200–3200 cm⁻¹) with absorption maxima around 515–525 nm and emission maxima around 595–620 nm in toluene. Notably, this large Stokes shift was not observed for the unsubstituted parent compound 1, highlighting the increased electronic delocalization in thio-containing compounds. We also observed that the emission wavelengths shifted from ~590 nm for the most electron-poor thioaryl group (2e, 4-Cl-Ph, toluene) to ~618 nm for the most electron-rich substituent (2d, 4-OMe-Ph, toluene); however, the absorption maxima remain unchanged (519 and 521 nm for 2d and 2e). In most BODIPY dyes, it has been reported that the HOMO is localized on the π-system of the pyrrole rings, whereas the LUMO sits on a mix of the π-system and the meso position [69]; however, with the introduction of the thioaryl groups on positions 2 and 6 of the BODIPY core, the compound may undergo heavy-atom-free intersystem crossing via photoinduced electron transfer (PET) as observed in some BODIPY dyads [45].

Next, to evaluate the applicability of our new BODIPY fluorophores for live-cell imaging, we performed cell viability assays. For these experiments, we used breast cancer cells MDA-MB-231 as a representative cell line. All our new compounds (2a–2e) did not show any significant cytotoxicity when incubated with the cells for 16 h at 10 μM, and adverse effects on cell viability were only detected at very high concentrations (i.e., >100 μM, Figure S3). Because the photophysical properties of our new BODIPY fluorophores are relatively similar, we decided to assess the application of compound 2a, the simplest disubstituted thioaryl-BODIPY compound, for live-cell imaging experiments.

First, we evaluated the ability of compound 2a to label different human cancer cell lines, namely MDA-MB-231 and MCF7 (both breast cancer cells) and HT29 (colorectal carcinoma). Cells were incubated with compound 2a (10 μM) and Hoechst 33342 for 15 min prior to acquiring images under the fluorescence confocal microscope (λ_exc: 488 nm, λ_em: 550–680 nm). Figure 2 shows that compound 2a was taken up by cancer cells, with slightly brighter staining being observed in breast cancer cells than in colorectal cancer cells, which could be due to differences in the expression of transporters or efflux pumps between these cell lines. The parent compound 1 was also able to stain MCF7 cells with a similar intracellular distribution than compound 2a (Figure S4).

Next, we also examined the utility of compound 2a to label human primary cells, such as healthy immune cells derived from peripheral blood. We examined the labeling capabilities of compound 2a in human neutrophils (both unstimulated and stimulated with phorbol myristate) and in monocyte-derived macrophages (both unstimulated and stimulated with liposaccharide). We acquired Z-stack fluorescence microscopy to analyze the cell labelling as well as the intracellular localization. As shown in Figure 3, we observed bright intracellular punctate structures in neutrophils, whereas the labeling in macrophages was distributed across the whole cytoplasm.

4. Conclusions

In conclusion, we reported a new synthetic approach to generate BODIPY derivatives under neat conditions and the preparation of novel 2,6-thioaryl-derivatized BODIPY compounds. The straightforward and selective functionalization of the BODIPY core by thiophenol chloride prepared in situ allows for easy modification and addition of targeting moieties. Analysis of the photophysical properties shows that thioaryl-BODIPY derivatives display large Stokes shifts, which can be advantageous for optical imaging. Finally, we have confirmed that thioaryl-BODIPYs are not toxic in mammalian cells at their working concentrations and can be used for fluorescence confocal microscopy in human cells of multiple origin.