Far-red BODIPY-based oxime esters: photo-uncaging and drug delivery

Abstract

Far-red BODIPY-based oxime esters for photo-uncaging were designed to release molecules of interest with carboxylic acids. The low power red LED light breaks the N–O oxime ester bond and frees the caged molecules. We studied the mechanism and kinetics of the uncaging procedure using a ¹H NMR spectrometer. More-over, the drug delivery strategy to release valproic acid (VPA) on demand was tested in vitro using this far-red BODIPY photo-uncaging strategy to induce apoptosis in tumor cells.

Photo-uncaging compounds are light-responsive molecules that can undergo bond cleavage once exposed to a certain wavelength of light.¹ This technique is also known as the use of photoprotecting groups (PPGs) and has been widely used in industry and organic syntheses.² The key components of photo-uncaging compounds are the chromophore (to harvest the energy of photons) and the caged molecule of interest, which are connected via a photolabile covalent bond.³ Upon light irradiation, the photolabile covalent bond cleaves and consequently frees the caged molecule with precise spatial and temporal control.⁴ However, most of the photocaged compounds only have a maximum absorbance from UV to blue (300–450 nm).⁵ Although sufficient for industry applications, short-wavelength light has severe photon cytotoxicity and low tissue penetration depth, and the applications in biomedical and biological studies are limited.⁶ Thus, it is still challenging to extend the wavelength into the far-red region and preserve the uncaging efficiency, and this endeavour has become a long-term goal.

The selection of the chromophore to construct photo-uncaging compound is important, as it is a key component that harvests the energy of incident light and is responsive to light wavelength. In general, chromophores have wide usages in chemosensors,^7,8 photodynamic therapy,⁹ 3D data storage,¹⁰ and bio-imaging.^11,12 Chromophores with an absorption maximum approaching the far-red window can passably take light in deep tissue and have broader usage in the biomedical field.¹³ Among these far-red fluorophores, boron-dipyrromethene (BODIPY) derivatives are an important branch due to their high fluorescent quantum yield, relative chemical inertness, exceptional photostability, and high molar absorptivity.^14,15 Furthermore, the synthesis and modification of BODIPY core structures have been extensively investigated in recent years.^16,17 Thus, several BODIPY derivatives have been developed as photocatalysts,¹⁸ photodynamic therapy agents,¹⁹ photo-controlled drug release systems,²⁰ and optical sensors.²¹ The photo-cleavage of the C–O bond based on BODIPY was well developed previously, with fair degradation quantum yield under green light irradiation.²² Recently, efforts have been made to extend the maximum absorption wavelength, and only a few photocages met this criterion and showed applications in biology. Considering the difficulty of structure modification to reach these goals, challenges remain in structure design and synthesis.^23–26 For example, Sitkowaska et al. reported an efficient and biocompatible BODIPY-based photocaged compound that could release amine to control heart rhythm.²³ In another example, Winter et al. reported a far-red/near-IR BODIPY photocage by blocking unproductive conical intersections.²⁷ The same group also reported a series of BODIPY photocages that could be cleaved under a one photon mechanism in the near-infrared light range via a C–O bond cleavage.²⁸ More strategies have been reported based on different bond dissociations on the BODIPY platform,²⁹ such as B–O bond cleavage in the BODIPY core structure to release histamine.³⁰

Recently, we reported a novel strategy to release carboxylic acid via the breakage of the N–O bond in a real-time manner.³¹ We connected the BODIPY core structure with the molecule of interest with a C＝N–O–R group and uncaged cargos upon green light irradiation. We monitored the photo-uncaging mechanism and kinetics and found the photodegraded product to be BODIPY-nitrile. Based on that, we explored the photo-uncaging activated by green light and capable of releasing VPA for histone deacetylases that caused the death of tumour cells.³¹ Encouraged by the versatile preparation and feasibility of photo-uncaging based on the N–O bond, in this report, we further conjugated a diphenyl-sulphide group to the BODIPY core structure and extended the absorption wavelength to the far-red light window (Scheme 1).

Scheme 1.

Synthesis route of compound 3a/3b. The labels correspond to the following: (a) DMF, piperidine, acetic acid, microwave irradiation 150 1C, 40 psi, 5 min; (b) POCl₃, DMF, dichloromethane; (c) NH₂OH-HCl, ethanol; (d) acetyl chloride/2,2-Di-n-propylacetyl chloride, triethyla-mine, dichloromethane.

For the uncaging cargo, we chose acetyl acid (compound 3b) to represent universality and VPA (compound 3a) as an example of an anti-cancer medicine. VPA is a small molecular fatty acid that can be utilized to treat bipolar illness and epilepsy.^32,33 Also, VPA was discovered to inhibit histone deacetylase (HDAC) and was studied as an anticancer medication candidate.^34,35 When conjugated with a photocage, VPA can be selectively released while avoiding adverse effects. Herein, we showed that the as-prepared di-styryl-BODIPY oxime ester containing caged VPA (3a) can function as a light-controlled drug delivery platform for cancer treatment. We found that compound 3a can effectively release VPA into tumour cells upon far-red light irradiation and trigger tumour cell death.

Results and discussion

BODIPY oxime was synthesized using the approach that we previously reported³⁶ and used for further connection with molecules of interest (Scheme 1). First, the BODIPY core structure was conjugated with di-phenyl-sulphide aldehyde via microwave irradiation to afford di-styryl-BODIPY (compound 1) to extend the wavelength from green to red. Then, compound 1 was oxidized by POCl₃ to form an aldehyde derivative. A further reduction reaction by hydroxylamine hydrochloride was set up to afford the BODIPY oxime compound 2. Next, the reaction of compound 2 with acid chlorides generated BODIPY oxime ester derivatives, 3a and 3b, with caging molecules of VPA and acetic acid. Red solids were obtained with high yields of around 45% and 43% for 3a and 3b, respectively. The chemical shift of the proton located on CH=NO at 8.43 ppm confirmed the formation of both oxime esters (Fig. S1 and S4, ESI†), and corresponding high-resolution mass peaks at m/z of 976.3815 and 892.2844 were also observed (Fig. S3 and S6, ESI†).

After successful preparation of 3a and 3b, their UV absorption and emission spectra were measured in methanol. The conjugated structures with sulphur heteroatoms lead to strong push–pull systems with far-red absorbability. The absorption spectra of 3a and 3b show intense peaks centred at 626 nm (ε = 5.6 × 10⁴ M^–1 cm^–1) and 627 nm (ε = 5.9 × 10⁴ M^–1 cm^–1) (Fig. 1a). Meanwhile, the emission maximum wavelengths centred at 664 nm and 665 nm, respectively. Fluorescent quantum yields (Φ_flu) were calculated using crystal violet as the standard, and the Φ_flu values of 3a and 3b were measured to be 0.46 and 0.48. The high Φ_flu values are partially attributed to the conjugation with electron-rich group diphenyl sulphide. Compared with precursor compounds 1 and 2, 3a and 3b have similar photophysical properties regarding the absorbance and emission. The lifetime was calculated and 4.4 ns was obtained for compound 3a (Fig. 1b).

Fig. 1.

The photophysical spectrum of compound 3a: (a) normalized absorption and emission spectrum; (b) lifetime measurement.

Next, the photo-uncaging process and light sensitivity of compounds 3a and 3b were monitored using UV-vis and emission spectroscopy. We irradiated the solution of styryl-BODIPY oxime ester (1.0 × 10^–5 M of 3a) in MeOH/H₂O (4/1, v/v) with LED light (660 ± 30 nm, power density of ~20 mW cm²), and recorded UV-vis and emission spectra at different time intervals of light irradiation. The blue shift of the peak in the absorption and emission spectra was observed. For the absorption spectrum, it has a ratio-metric trend at a wavelength of around 588 nm (Fig. S7, ESI†) and for the emission spectrum, an apparent density increase was observed (Fig. 2), which was attributed to the photolysis of compound 3a. Similar properties were reported and have been used in super-resolution microscopy, such as stochastic optical reconstruction microscopy (STORM).³⁷ We then measured the photo-uncaging quantum yields (Φ_un) using a reported procedure.³¹ Furthermore, the rate of photo-uncaging was presented as a function of the quantum efficiency parameter (ɛΦ_un), and the ɛΦ_un values are 5.32 × 10³ and 7.08 × 10³ for compounds 3a and 3b, separately. The photophysical properties of 3a and 3b are summarized in Table 1.

Fig. 2.

The emission spectra of compound 3a with irradiation of 660 nm LED light in MeOH/H₂O (v/v, 4 : 1) within different time intervals (excitation wavelength at 590 nm).

Table 1.

Photophysical properties of compounds 1, 2, 3a and 3b.

Compound	λabs (nm)	λem (nm)	ɛ (M−1 cm−1)	Φ flu	Φun × 103	ɛΦun × 10−3
1	639	662	47500	0.55	—	—
2	640	675	45800	0.31	—	—
3a	626	664	56000	0.46	9.5	532
3b	627	665	59000	0.48	12.0	708

λ_abs is the wavelength of absorbance maxima, λ_em is the emission maximum; ɛ is the extinction coefficient; Φ_flu is the fluorescent quantum yield; Φ_un is the photo-uncaging quantum yield. All measurements were performed in methanol

To further study the photo-uncaging pathway, we then used compound 3a as the model compound and monitored the photolysis mechanism and kinetics using ¹H NMR. 3a solution in methanol/water (4/1, v/v, 1 × 10^–3 M) was irradiated with LED light (660 ± 30 nm, ~20 mW cm ²), and the mixed solution was dried at regular time intervals for ¹H NMR analysis. Before light irradiation (0 min), 3a showed a CH = NO proton peak at 8.43 ppm and BODIPY-β-proton at 6.74 ppm. After 20 min irradiation under red LED light, we observed a decrease in the peaks of 3a and an increase of new peaks nearby (Fig. 3). After purifying the mixture of photolyase, we confirmed that the photo-product was styryl-BODIPY-CN (BD-CN) by ¹H NMR, ¹³C NMR, and mass spectrum (Fig. S8–S11, ESI†). According to the photo product, we proposed a mechanism the same as the green light BODIPY oxime ester that we reported before (Scheme 2).³¹ We further measured the photophysical properties of BD-CN and found that it had a high fluorescent quantum yield of 0.74 (Fig. S15, ESI†), much higher than compound 3a, which correlated well with the emission increase in photolysis (Fig. 2), and the maximum emission wavelength of 659 nm that is lower than that of compound 3a, which correlated well with the blue shift in photolysis (Fig. 2).

Fig. 3.

¹H NMR spectra of 3a, BD-CN, and the photolysis mixture after 20 min irradiation with red LED light (λ = 660 ± 30 nm, ~20 mW cm^–2).

Scheme 2.

Proposed photo-uncaging mechanism of compound 3a.

After completing the photophysical measurement and photo-uncaging analysis, we then confirmed the dark stability of 3a in the cell culture medium (Fig. S20, ESI†) and used HeLa cells for in vitro studies. To check the cellular uptake of compound 3a, we incubated different concentrations of 3a in HeLa cells for 1 h, then recorded cell images using confocal laser scanning microscopy. The red emission cell image showed that compound 3a could be uptaken by HeLa cells (Fig. 4). It is worth mentioning that even at a low incubation concentration of 1 μM, compound 3a distributed in the cytoplasm can still be observed under a fluorescent microscope.

Fig. 4.

Cellular uptake of 3a within different concentrations (1, 5, and 10 μM) incubated for 1 h with HeLa cells. Cell images were excited at 633 nm and recorded at emission wavelengths 665 ± 20 nm.

We then measured the dark cell viability so that cells could endure the concentration of 3a to 20 μM (Fig. S21, ESI†). Finally, photo-toxicity experiments were performed in 660 nm LED light (660 ± 30 nm, power density of ~20 mW cm^–2) using HeLa cells and 20 μM 3a. The results showed that 3a killed more than 70% of HeLa cells after 30 min light irradiation. In a control experiment, HeLa cells were treated with a photolysis mixture at 20 μM under irradiation by 660 nm LED light. As shown in Fig. 5, the photolysis mixture has no cytotoxicity to HeLa cells (blue), which indicates that the mixture of free VPA and BD-CN has no apparent toxicity at the experimental concentration. After the treatment of 3a and LED light, we also monitored the morphology change of the cells using optically computed phase microscopy that quantitatively measures the thickness of cells.^38,39 Apparent rounding and increase in height were observed for the treated samples compared with the untreated ones that were elongated and flat (Fig. S22, ESI†).

Fig. 5.

Cell viability with irradiation time-dependent experiments using HeLa cells and 20 μM 3a (red) and the photolysis product (blue) (irradiated by 660 nm LED light).

Conclusions

We have designed and developed a new far-red BODIPY-based oxime ester photo-uncaging system for the light-controlled delivery of small molecules with the functional group of carboxylic acid. The photophysical and photochemical studies were characterized and a plausible mechanism was proposed according to the structure of the photoproduct. We also showed the light-triggered delivery of an anti-cancer drug and observed the light-induced therapeutic effect due to high cell uptake and targeted release.