Subscriber access provided by Stockholm University Library
Biological and Medical Applications of Materials and Interfaces
Hybrid Azo-Fluorophore Organic Nanoparticles as Emissive Turn-on Probes for Cellular Endocytosis Joanna Boucard, Tina Briolay, Thibaut Blondy, Mohammed Boujtita, Steven Nedellec, Philippe Hulin, Marc Grégoire, Christophe Blanquart, and Eléna Ishow ACS Appl. Mater. Interfaces, Just Accepted Manuscript • DOI: 10.1021/acsami.9b12989 • Publication Date (Web): 19 Aug 2019 Downloaded from pubs.acs.org on August 20, 2019
Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.
Page 1 of 8 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
ACS Applied Materials & Interfaces
Hybrid Azo-Fluorophore Organic Nanoparticles as Emissive Turn-on Probes for Cellular Endocytosis Joanna Boucard,† Tina Briolay,‡ Thibaut Blondy,‡ Mohammed Boujtita,† Steven Nedellec,§ Philippe Hulin,§ Marc Grégoire,‡ Christophe Blanquart,‡,* Eléna Ishow†,* †CEISAM–UMR
CNRS 6230, Université de Nantes, 2 rue de la Houssinière, 44322 Nantes, France. ‡CRCINA, INSERM, Université d’Angers, Université de Nantes, 44007 Nantes, France. §INSERM UMS 016-UMS CNRS 3556, 8 quai Moncousu, 44007 Nantes, France.
Supporting Information Placeholder ABSTRACT:
The development of fluorescent organic nanoparticles, serving as bioimaging agents or drug cargos, represents a buoyant field of investigations. Nevertheless, their ulterior fate and structural integrity after cell uptake remain elusive. To this aim, we have elaborated original photoactive organic nanoparticles (dTEM ~35-50 nm wide) with an off-on signal upon cellular internalization. Such nanoparticles are based on the noncovalent association of red-emitting benzothiadiazole (BDZ) derivatives and azo dyes, acting as fluorescence quenchers. Upon varying the azo:BDZ ratio, we found that quantitative emission quenching could be obtained with only a 0.2:1 BDZ:azo ratio and originated from exergonic oxidative and reductive photoinduced electron transfer from the azo units (elG0 = -0.21 eV and -0.29 eV resp.). Such results revisited the origin of emission quenching, often confusedly ascribed to Förster resonance energy transfer. A nonlinear and sharp drop of the emission intensity with the increase in the azo unit density n was observed and presents comparable evolution to a n-1/3 mathematical law. Thorough biological examinations involving cancer cells prove a receptor-independent endocytosis pathway, leading to progressive cell lighting upon nanoparticle accumulation in the late endosomal/lysosomal compartments. Complete emission recovery of the initially quenched azo:BDZ nanosystems could be achieved by using mefloquine which caused endosomal/lysosomal disruption, and release of their content in the cytoplasm. Such results demonstrate that the dot-like emission from endosomes actually stems from fully dissociated individual dyes and not integer nanoparticles. They conclude on the high spatial confinement promoted by organelles and finally question its severe impact on functional compounds or nanoparticles whose properties are strongly distance-dependent.
Keywords: organic nanoparticles, fluorescence, azo dyes, photoinduced electron transfer, bioimaging probes, nanoparticle endocytosis. INTRODUCTION The fabrication of emissive nanoparticles endowed with progressive biodegradability or excretion properties has recently been considered as an imperative criterion for bio-imaging and more generally theranostic applications.1,2,3 In this way, deleterious in vivo accumulation and correlative risks of cell toxicity, oxidative
stress or inflammation are strongly minimized.4,5,6 In this context, fluorescent organic nanoparticles (FONs) have appeared in the last ten years as particularly attractive probes since they are mainly composed of small hydrophobic or hydrogen-bonding fluorophores, self-assembled in a non-covalent fashion.7,8 They display advantageous features such as a high brightness (> 107 L.mol-1.cm-1) related to their high density of active units (104 to 105 molecules), a large synthetic modularity enabling straightforward emission color tuning (from the blue to the near-infrared (NIR) range),9,10,11 and eventually no cytotoxic heavy metals. The most commonly adopted strategy to generate FONs is based on aggregation induced emission (AIE) effects, causing rigidification of the molecular backbone of the fluorophores and dramatic increase in the emission after self-assembling into FONs.12,13 Although AIE-based nanoparticles have showed promising use for cell labelling,14 they face strong limitations when long-term followup over several cell cycles is required. Indeed, their expected disassembling into flexible single units within the cells is likely to progressively conduct to non-fluorescent molecules due to the recovery of prevailing radiationless relaxation processes of the excited state, hence further tracking is precluded. To avoid gradual fainting of the emissive labels, fluorophores showing emission signal as single or packed entities would represent an undoubtedly superior alternative. To this aim, the simplest way to avoid dark states issued from stacking is to prevent molecules from aggregating with each other by grafting bulky alkyl groups on the -conjugated backbone,15,16,17 or associating bulky organic counterions to charged fluorophores.18,19 The former approach has successfully been experimented especially with solvatochromic dyes, exhibiting a distinct and diffuse emission color when disassembled compared with the dot-like FON signal after cell internalization.20,21 Nevertheless, having in hand FONs amenable to switch from an off to an on state upon mere erosion after cell uptake, would represent a very ubiquitous approach to signal cell entrance, without requiring specific enzymatic or chemical intervention. To reach this goal, molecular beacons, lighting on upon hybridization with a determined DNA or RNA sequence, represent inspiring systems.22 This photophysical configuration actually imply a pair composed of a fluorophore and an azo-based quencher counterpart.23 The fluorescent and azo units are each grafted at one extremity of an oligonucleotide strand. The DNA strand hairpin structure brings them in close vicinity, leading to a dark state.24 Upon recognition of the targeted complementary oligonucleotide strand, the molecular beacon unfolds, placing the dyes far apart, which restores the fluorophore emission. Following
ACS Paragon Plus Environment
ACS Applied Materials & Interfaces 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
this principle and the ability of FONs to dissociate into individual components after cell endocytosis,25 we want herein to show that FONs with a turn-on signal after cell internalization can simply be constructed through the non-covalent incorporation of azo dyes, providing higher contrast for cell labelling than those with neat FONs. The reported studies also interrogate on the origin of the fluorophore quenching by the “black” azo dye, which has rarely been deeply considered and so often erroneously described via Förster Resonance Energy Transfer (FRET).26 They evidence the high confinement effects and electronic interactions of molecules when they are self-assembled as FONs and, more unexpectedly, once engulfed inside endosomal vesicles. Finally, the devised dual structures allowed us to unveil for the first time to our knowledge the specific cell internalization mechanism of FONs and their unexpected fate right after endocytosis. RESULTS AND DICUSSION Choice of FON precursors. FONs are mostly simply fabricated upon nanoprecipitation by quickly mixing a small amount of a concentrated solution of hydrophobic compounds with a large volume of water.27 An extensive hydrophobic backbone for the photoactive molecules, driving their self-assembling upon van der Waals interactions, is a mandatory prerequisite to induce formation of FONs following a nucleation-aggregation process upon solvent shifting.21,28 In this way, tight cohesion of FONs is promoted, while their erosion is prevented in water. Since FONs are further exposed to biological surroundings, especially saline and serum-completed cell culture media, they need to display large colloidal stability. The most common strategy consists in imparting nanoparticles with charged surface, which can easily be achieved by devising molecules with ionizable groups at physiological pH, like carboxylic acids causing little cell toxicity compared to ammonium salts. Finally, self-assembled -conjugated fluorophores must keep their intrinsic properties, which in our case implies the presence of sterically crowded substituents. To meet all the above-mentioned criteria, we designed the structures of the fluorescent25 and azo29 dyes with an extensive aromatic backbone, substituted with tertbutylphenyl groups and a carboxylic acid unit as well (Figure 1). As described infra, the latter acidic group advantageously endowed organic nanoparticles with a largely negative surface potential, without requiring the addition of anionic surfactants whose immunogenicity and cytotoxicity are currently fiercely debated. Furthermore, it also participates to the intrinsic integrity of the nanoparticles thanks to the establishment of favorable hydrogen bonds between the molecular constituents.
Figure 1. Structures of BDZ fluorophore and azo dye. We thus based the azo structure on the backbone of a recently synthesized red-emitting benzothiadiazole (BDZ) derivative whose corresponding FONs were harnessed to specifically deliver doxorubicin, an anticancer drug, into malignant cells. BDZ fluorophores, containing strong electron-donating moiety, display high brightness,30 allowing for large emission dynamics when selfassembled with the azo quenchers. Moreover, they have showed no significant cytotoxicity, in the same way as azo dyes, broadly used to color food and textiles. The particularity of the retained
fluorescent and azo compounds relies on their glassy behavior in the solid state (Tg(BDZ) = 173 °C; Tg(Azo) = 115 °C) as revealed by differential scanning calorimetry. Such amorphous behaviors favor a high mixing entropy of the azo and BDZ units, especially since their structures are very similar, leading to spherical nanoparticles in solution to minimize their interfacial tension. Uniform composition is thus expected, especially for bicomposite nanoparticles, which will further simplify the statistical treatments of the quenching mechanism. Dispersions of pristine FONs and azo organic nanoparticles (AONs) were thus directly obtained by adding into water (2.5 mL) a 1.45 mmol.L-1 THF solution (50 L) of the corresponding compound under rapid stirring. Mixed nanoparticles, dubbed FAONs, were obtained by preparing first an organic solution with a fixed 1:1 BDZ:azo ratio that was further injected in water (Figure 2a). All resulting organic nanoparticles FONs, AONs and FAONs display a remarkable well-defined spherical geometry from transmission electron microcopy characterizations. Their mean diameters were found to be 36 to 49 nm with a relatively narrow dispersion given by a polydispersity index value of around 0.24-0.27 (Table S1, Figure S1). Compared to most of organic nanoparticles exclusively made out of aromatic and alkyl units, the presence of carboxylic acids imparts nanoparticles with a largely negative surface potential ( = -33 to 43 mV), hence long-term stability over months in aqueous solution is obtained (Figure S2a).
Figure 2. a) Nanoprecipitation of organic nanoparticles upon solvent shifting (exemplified with mixed FAONs). b) TEM images of FON, AON and FAONs. Inset: fluorescence from FONs in quartz cells upon excitation using laser diode at 480 nm. Photophysical properties of the photoactive units. Both compounds in THF solution strongly absorb in the visible range around 455 nm due to a marked charge transfer from the electrondonating triphenylamino unit to the electron-withdrawing benzothiadiazolyl30 or azo moieties. -conjugation of the latter units with the carboxylic acid unit reinforced the strength of the charge transfer. Only, BDZ emits an intense unstructured band (f = 0.25) centered at 665 nm and showing strong Stokes shift, which is appreciable to limit the autofluorescence contribution of live cells (Figure S3). The absorption characteristics were found almost similar for pristine FONs and AONs with a slight 10 nm bathochromic shift of the absorption maximum, as a result of the polar surroundings exerted by the neighboring dyes. As for emission, FONs were found quite intense (f = 0.11), especially in the context of a strong Stokes shift and red emission (Table S2). This moderate decrease in emission had earlier been ascribed to radiationless deactivation through hydrogen-bond coupling with the surrounding water molecules.25 The most remarkable feature relies on the fully quenched fluorescence signal of the azo-doped FONs comprising a 1:1 BDZ:azo (Figure 3b).
ACS Paragon Plus Environment
Page 2 of 8
Page 3 of 8 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
ACS Applied Materials & Interfaces
Figure 3. Photophysical properties of FAONs, AONs and FONs. a) Absorption and emission spectra; b) Fluorescence decays (exc = 480 nm, em = 636 nm). Control experiments consisting in the mixing of aqueous dispersions of pre-fabricated FONs and AONs following a 1:1 ratio show by contrast no significant emission quenching whereas the absorption spectra of the mix and FAON (1:1) dispersions appeared very similar. These elements clearly let us suggest the absence of dark intermolecular aggregates in the ground state and the existence of a quenching process arising from the excited state. Confined electron transfer. Very efficient energy transfer in large core-shell lanthanide nanocrystals (20-30 nm) coated with fluorescein isothiocyanate (FITC) has already been reported.31 In this case, efficient FRET resulting from IR-pumped nanocrystals down to the surface-attached fluorophores is achieved thanks to the small dimensions of the system with regard to the spatial extension of the energy transfer process and the favorable overlap between the upconverted emitted photon and the FITC absorption spectra. Such FRET has also been exploited to photoswitch the fluorescence of organic nanoparticles comprising BDZ-based emitters and diarylethene photochromes.32 Nevertheless, in our case, the azo absorption band displays insignificant spectral overlap with the BDZ emission band. Hence, no such FRET can operate, contrary to numerous assertions regarding the quenching origin of combined azo dyes and fluorophores, even in the absence of convenient spectral overlap. To gain insight into the underlying mechanism, we fabricated FAONs with a BDZ:azo ratio varying from 0.006:1 to 1:1 while keeping constant the fluorophore content. Comparative solutions in THF containing the same dye composition and concentration were prepared to evaluate the impact of the dye confinement in nanoparticles. Steady-state and time-resolved fluorescence measurements revealed considerably distinct results between mixed BDZ and azo solutions in THF and FAON dispersions. For the former, no significant emission change was noticed after internal filter correction while the lifetime of the BDZ excited state remained unchanged at 5.72 ns whatever the azo:BDZ ratio varying from 0 up to 380 % (Figure 4a-b). By contrast, dramatic reductions of the signal intensity and fluorescence decay of FAONs upon increasing the azo ratio were observed (Figure 4c). Similar to the above-mentioned FRET-based nanoparticles, the decrease of FAON emission did not follow a linear relationship with the photochromic content (Figure 4d).
Figure 4. a) Evolution of the absorption and emission spectra of BDZ in THF solution (6.510-6 mol.L-1) upon adding azo solution in THF (from 0 to 3.8 eq.). b) Lifetime decay – exc = 480 nm; em = 636 nm). c) Absorption and emission spectra of FAON dispersions in water as a function of the azo:BDZ ratio (namely from 0 to 10 mol. %). d) Evolution of the fluorescence intensity (red curve) and n-1/3 where n is the density of azo units (blue curve) as a function of the azo:BDZ ratio (from 0 to 100 mol. %) in FAONs. Interestingly, the very rapid decrease in intensity seems to follow an evolution similar to n-1/3 where n designates the density of azo units, assuming a homogeneous distribution throughout the sphere and a constant sphere diameter (here 49 nm). This first simple theoretical modeling indirectly points out the effects exerted by the intermutual distance between the azo quenchers and BDZ fluorophores whose accurate calculations would require longer and complex developments which are fully beyond the scope of the current studies33 Emission quenching upon energy transfer to the azo unit, amenable to undergo photoisomerization occuring in the subpicosecond range, was first envisaged.29 However, it could be ruled out because no change in the emission intensity was noticed after irradiation at 480 nm in the azo absorption band under optimized conditions (10 min, P = 9 mW.cm-2).34 In this condition, we put forward the possibility of photoinduced electron transfer (PET) between both BDZ and azo units. Compared to FRET occurring up to 100 Å, PET is limited to 10 Å, namely over a much shorter distance requiring larger orbital overlap.35,36 It turns out that in our case, although 83 % of emission quenching were obtained for only a 10 % molar content of photochromic dyes, such ratio fluorophore/photochrome remains higher than that reported for the diarylethene-based nanoparticles involving FRET.32 To confirm the assumption of a PET process, we carried out electrochemical measurements to access the redox potentials of each compound. From cyclic voltammetry performed in CH2Cl2 solution and referred to ferrocene (E0(Fc+/Fc) = 0.408 V), both BDZ and azo units showed a first quasi-reversible reduction potential at -1.45 and -1.47 V respectively (Figure S4, Table S3). As for oxidation, they also showed an anodic peak at 0.85 and 0.95 V respectively. Strong adsorption was found for the azo compound at potential higher than 1.5 V while BDZ showed a second oxidation potential for which reversibility could be achieved only at a higher scan rate, so we did not consider the latter peak. Given the high density of photoactive units within each nanoparticle, we considered the Sandros-Boltzmann approach that takes into account the formation of exciplexes37 and revisited the historical Rehm-Weller equation38. Both oxidative and reductive photoinduced electron transfer, based on prior excitation of the azo units, were then found exergonic and characterized by Gibbs free energy values elG0 equal to -0.21 and -0.29 eV respectively (see SI for calculations, Table S4). Such clear evidence of electron transfer recalled earlier studies involving azo dendrimers doped with eosin.39 Nevertheless, in the latter case, contribution of the azo units was ruled out and
ACS Paragon Plus Environment
ACS Applied Materials & Interfaces 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
emission quenching of eosin was attributed to reductive PET from the branched alkylamino scaffold. In the case of FAONs, tight structural confinement of the azo and fluorescent units within the nanoparticles renders PET phenomenon highly favorable, enabled by extensive orbital overlap between the aromatic backbones and appropriate design of the dyes. Interestingly, whereas fluorescence still operates for FONs, the photochromic properties for AONs were fully annihilated. Since azo E-Z photoisomerization requires enough free volume,40 the lack of photochromic properties confirms the assumption of extensive intermolecular interactions in the excited state, reinforced by the participation of hydrogen bonding. To conclude on this part, the choice of the azo unit relatively to the fluorophore has to be regarded through mutual redox activities, rather than through spectral overlap considerations, which opens a wide range of possible azo:fluorophore pairs in the framework of the herein reported concept. Probing the receptor-free endocytosis pathway. In order to envisage FAONs as potential light-up indicators of endocytosis mechanisms, known to promote the uptake of nanomedicines, we first studied in solution the ability to recover fluorescence after nanoparticle engulfment and dissociation. Such phenomenon has already been reported for various FONs, once endocytosed. To this aim, dissolution of FONs was undertaken by dichloromethane treatment, after adding a small amount of THF to favor phase transfer. Whatever the initial azo content, fluorescence measurements of all resulting chlorinated solutions showed similar emission intensity almost equal to that of solutions mainly constituted of BDZ fluorophores (Figure S5). We then applied such concept of turn-on emitters to live cells. We chose malignant mesothelioma cells (Meso 34 cell line issued from pleural fluids of patients), known to slowly develop after asbestos inhalation due to airborne fiber environmental pollution,41 and causing fatal cancers for which early treatment and diagnosis are severely lacking.42 We focused on FAONs with a 1:1 azo/BDZ ratio and incubated Meso cells with FAONs, AONs, and FONs, at a respective organic nanoparticle:cell ratio of 4105:1 per mL of culture medium. No cell toxicity was observed for any kind of nanoparticles. Colloidal stability for BDZ FONs was checked over 7 days using wide-field microscopy since any DLS investigations were precluded due to the presence of serum proteins (Figure S6, S2b-c). Flow cytometry was performed after various times and showed fast fluorescence staining for FON-incubated cells. As expected, no fluorescence could be detected for AONs and FAONs at the very beginning. However, after a few hours, progressive lighting of the cells was observed (Figure 5a-b, Movie S1). Surprisingly, while mean fluorescence intensity for cells incubated with FONs reached a plateau after 72 h, such signal, albeit quite dimmer, kept growing for cells treated with FAONs even after 72 h (Figure 5c).
Figure 5. a) Fluorescence microcopy imaging of Meso 34 cells after a 1, 3 and 5 h incubation time with FONs and FAONs (1:1) (the cell nuclei were stained in blue using Hoechst and FONs appeared as orange dots). b) Kinetic of internalization of FON and FAON performed over 72 h incubation time using flow cytometry (exc = 480 nm; (em > 670 nm). c) Comparative fluorescence intensities for cell incubation carried out at 37 and 4 °C for 3 h after pre-incubation of cells on ice for 30 min. d) Incidence of cytochalasin D (Cyto D) after cell pretreatment for 2 h and nanoparticle incubation for 3 h. Graphics represent means of ratio of mean fluorescence intensity (RMFI) of three independent experiments. *, p
