Heavy-atom-free p-twisted photosensitizers for fluorescence bioimaging and photodynamic therapy

This journal is © The Royal Society of Chemistry 2024 J. Mater. Chem. B, 2024, 12, 8107–8121 | 8107 Cite this: J. Mater. Chem. B, 2024, 12, 8107 Heavy-atom-free p-twisted photosensitizers for fluorescence bioimaging and photodynamic therapy† Darı ´o Pucha ´nSa ´nchez,‡ a Korentin Morice,‡ a Monika G. Mutovska, b Lhoussain Khrouz, c Pierre Josse, a Magali Allain, a Fre ´de ´ric Gohier, a Philippe Blanchard, a Cyrille Monnereau, c Tangui Le Bahers, cd Nasim Sabouri, e Yulian Zagranyarski,* b Clement Cabanetos * a and Marco Deiana * ef As the field of preclinical research on photosensitizers (PSs) for anticancer photodynamic therapy (PDT) continues to expand, a focused effort is underway to develop agents with innovative molecular structures that offer enhanced targeting, selectivity, activation, and imaging capabilities. In this context, we introduce two new heavy-atom-free PSs, DBXI and DBAI, characterized by a twisted p-conjugation framework. This innovative approach enhances the spin–orbit coupling (SOC) between the singlet excited state (S 1 ) and the triplet state (T 1 ), resulting in improved and efficient intersystem crossing (ISC). Both PSs are highly effective in producing reactive oxygen species (ROS), including singlet oxygen and/ or superoxide species. Additionally, they also demonstrate remarkably strong fluorescence emission. Indeed, in addition to providing exceptional photocytotoxicity, this emissive feature, generally lacking in other reported structures, allows for the precise monitoring of the PSs’ distribution within specific cellular organelles even at nanomolar concentrations. These findings underscore the dual functionality of these PSs, serving as both fluorescent imaging probes and light-activated therapeutic agents, emphasizing their potential as versatile and multifunctional tools in the field of PDT. 1. Introduction Phototherapeutic approaches encompass a broad spectrum of medical treatments exploiting the use of light. 1–4 Among these approaches, photodynamic therapy (PDT) stands out as a wellestablished and extensively investigated method in the field of anticancer phototherapy. 5–11 PDT harnesses the unique attributes of photosensitizers (PSs) and light to generate highly cytotoxic reactive oxygen species (ROS), specifically targeting cancer cells while preserving neighboring healthy tissues through minimally invasive to non-invasive modalities. 12,13 Upon light absorption, the PS is activated, transitioning from its ground state (S 0 ) to the lowest excited singlet state (S 1 ). The excited S 1 state subsequently evolves into an excited triplet state (T 1 ) through a phenomenon referred to as intersystem crossing (ISC). Importantly, the extended lifetime of the generated T 1 state allows it to follow two distinct evolutions. It can directly generate highly cytotoxic singlet oxygen ( 1 O 2 )via a type-II mechanism or initiate a series of electron transfer reactions, through a type-I mechanism, resulting in the formation of radical intermediates such as superoxide (O 2 ) and hydroxyl radical (OH). 14 These dual pathways collectively contribute to the induction of apoptosis and/or necrosis in cancer cells. 6 In the context of PSs, a conventional practice has been to introduce halogens, such as Br, I, and second or third-row transition metal ions, into a target chromophore, with the aim of boosting ISC processes, via heavy atom-induced strong spin– orbit coupling (SOC). 15–18 However, this approach comes with certain limitations, such as inherent dark toxicity and potential high synthesis costs (in terms of price and number of a Univ Angers, CNRS, MOLTECH-ANJOU, SFR MATRIX, F-49000 Angers, France. E-mail: [email protected] b Faculty of Chemistry and Pharmacy, University of Sofia, 1 James Bourchier blvd., 1164 Sofia, Bulgaria. E-mail: [email protected] c ENS de Lyon, CNRS, Laboratoire de Chimie UMR 5182, F-69342 Lyon, France d Institut Universitaire de France, 5 rue Descartes, 75005 Paris, France e Department of Medical Biochemistry and Biophysics, Umeå University, SE-901 87, Umeå, Sweden f Institute of Advanced Materials, Faculty of Chemistry, Wrocław University of Science and Technology, 50-370 Wrocław, Poland. E-mail: [email protected] †Electronic supplementary information (ESI) available: 1 H NMR, 13 C NMR, ESIMS, X-ray. CCDC 2083069 (for DBI), 2335339 (for DBXI) and 2335340 (for DBAI). For ESI and crystallographic data in CIF or other electronic format see DOI: https://doi.org/10.1039/d4tb01014k ‡Contributed equally. Received 10th May 2024, Accepted 15th July 2024 DOI: 10.1039/d4tb01014k rsc.li/materials-b Journal of Materials Chemistry B PAPER Open Access Article. Published on 18 July 2024. Downloaded on 1/30/2025 10:46:47 AM. This article is licensed under a Creative Commons Attribution 3.0 Unported Licence. View Article Online View Journal | View Issue 8108 | J. Mater. Chem. B, 2024, 12, 8107–8121 This journal is © The Royal Society of Chemistry 2024 steps). 15,19 Additionally, the introduction of heavy atoms can concomitantly boost ISC from S 1 -T 1 and T 1 -S 0 , resulting in a shorter lifetime of the T 1 state, potentially detrimental for the generation of ROS. 15,18,20 Consequently, these concerns have significantly driven researchers’ interest towards the development of heavy-atomfree PSs with optimized ISC and efficient T 1 population. In recent years, three main alternative mechanisms to enhance ISC of organic PSs have been proposed which consist in: 15,21 (i) Employing spin–orbit charge-transfer intersystem crossing (SOCT-ISC) by fostering, for instance, the interaction between orthogonal molecular donor–acceptor (D–A) structures. (ii) Enhancing the SOC by involving non-bonding orbitals in the electronic transitions by replacing, for instance, the oxygen atom in the carbonyl group of conventional fluorophores with a sulfur atom. (iii) Promoting torsional motions (twisting-induced ISC) by introducing, for instance, a nonplanar p-conjugated chromophore. To date, a wide array of D–A dyads and thionatedbased compounds that exhibit improved ISC have been reported, utilizing various structural frameworks such as anthracene, 22,23 benzothioxanthene imide, 24 boron dipyrromethene (BODIPY), 9,20,25–28 coumarin, 29 cyanine, 30 imidazolium, 31 naphthalimide, 29,32 phenoxazine, 33 pyrene, 34 perylene 35,36 and others. 37–39 Though efficient in generating ROS, such strategies have shown adverse effects on the emissive properties of these compounds, particularly in the case of thionated derivatives. 24,26,32,35 thus limiting their usefulness for imaging-guided PDT. Indeed, the inherent fluorescence properties of PS are crucial for several reasons. 40 They enable real-time tumor visualization, allowing precise boundary delineation for accurate targeting and comprehensive treatment. 41–44 Fluorescence aids in monitoring PS distribution and concentration within the tissue, ensuring adequate tumor accumulation before activation. 41–44 This selective activation minimizes damage to healthy tissues, enhancing treatment safety and effectiveness. Additionally, fluorescent imaging is essential for dosimetry, determining the appropriate light dose for effective PDT without excessive damage. 45 Post-treatment, fluorescence imaging assesses therapy response by evaluating signal changes, guiding further therapeutic decisions. 46,47 This advancement facilitates precise delineation of the treatment area and enables continuous monitoring of therapeutic progress. Moreover, PSs derived from thiones are prone to oxidation when exposed to light, leading to structural alterations. This raises safety concerns regarding their suitability for potential clinical applications. 24 In contrast to these design principles, there has been a recent significant focus on the exploration of heavy-atom-free PSs that incorporate a twisted p-conjugated system. 15 Indeed, considerable research efforts have been directed towards investigating the photophysical properties of extended p-conjugated helicoidal systems, which includes triply linked bayfused diperylene bisimides, 48 chromophores derived from hexa-peri-hexabenzocoronenes, 49 phenanthrene-fused twisted perylenebisimide 50 and BODIPY 51–53 derivatives. However, despite their promise, these materials have primarily been restricted to test tubes, with only few practical applications in cellular models. 52,53 Furthermore, they present several drawbacks such as their intricate synthesis, high molecular weight, limited fluorescence quantum yield (F F ), and, for some of them, a weak absorbance in the visible spectral range (above 400 nm). 15,54 Drawing on this, we reported on a rational design of a dibenzothioxanthene imide, abbreviated as DBI, 55 stemming from a classic highly luminescent dye, initially reported in the late 1970’s for plastic staining and textile industry, namely the benzothioxanthene imide (BTI). Characterized by a distorted pconjugated core, resulting from the selective inclusion of an extra fused phenyl ring, that facilitates efficient ISC through a significant SOC, DBI demonstrated remarkable ROS production and phototherapeutic efficacy at nanomolar concentrations. However, although it remains fluorescent, its F F of ca. 8%, was found to be at the lower limit for efficient detection and localization of in vivo tumor accumulation, a key feature for therapy-guiding PSs. 55 To overcome this obstacle, and provide a second generation of DBI-related molecules that attain simultaneously high fluorescence emission and effective ROS generation, we herein present derivatives, named DBXI and DBAI in which the sulfur heteroatom bridging the upper and lower part of the molecules was selectively replaced by either a less bulky chalcogen (O) or an unsubstituted amine, thus strongly affecting (i) the molecular strain, and thereby, (ii) the distortion induced SOC thus above all (iii) their relevant spectroscopic properties. By conducting a comprehensive examination that incorporates experimental, theoretical, and biological investigations, we thoroughly scrutinized the photophysical and in cellulo properties of these newly synthesized PSs, yielding highly consistent findings. With a strong visible centered emission, these compounds can be readily monitored in real-time, disclosing their accumulation within specific cellular organelles. Moreover, these PSs display minimal dark toxicity and efficiently inhibit cell proliferation in a dose-dependent manner with a halfmaximum inhibitory concentration (IC 50 ) in the nanomolar range (B100 nM) upon light exposure associated to morphological changes in cells, leading to apoptosis. This study further highlights the key role of molecular design and promising applications of such new generation of p-distorted heavyatom-free small PSs for imaging-guided photodynamic cancer treatment. 2. Experimental Synthetic procedures and compounds characterization All reagents and chemicals from commercial sources were used without further purification unless specified. Solvents were dried and purified using standard techniques. Flash chromatography was performed with analytical-grade solvents using Paper Journal of Materials Chemistry B Open Access Article. Published on 18 July 2024. Downloaded on 1/30/2025 10:46:47 AM. This article is licensed under a Creative Commons Attribution 3.0 Unported Licence. View Article Online This journal is © The Royal Society of Chemistry 2024 J. Mater. Chem. B, 2024, 12, 8107–8121 | 8109 Aldrich silica gel (technical grade, pore size 60 Å, 230–400 mesh particle size). Flexible plates Alugram Xtra SIL G UV254 from Macherey-Nagel were used for thin layer chromatography (TLC). Compounds were detected by ultraviolet (UV) irradiation (Bioblock Scientific). Nuclear magnetic resonance (NMR) spectra were recorded with a Bruker Avance III 300 ( 1 H, 300 MHz and 13 C, 76 MHz) or a Bruker Avance DRX500 ( 1 H, 500 MHz and 13 C, 125 MHz). Chemical shifts are given in ppm relative to the residual 1 H resonance of the deuterated solvent and coupling constants Jin Hz. High-resolution mass spectrometry (HRMS) was performed with a JEOL JMS-700 B/E. Matrix Assisted Laser desorption/ionization was performed on MALDI-TOF MS BIFLEX III Bruker Daltonics spectrometer using trans-2-[3- (4-tert-butylphenyl)-2-methyl-2-propenylidene]malononitrile (DCTB+) as matrix (Bruker, Billerica, MA, USA). The purity of the final compounds was 495% as assessed by NMR, HRMS (mass accuracy o5 ppm) and X-ray diffraction. 6-(2-Methoxynaphthalen-1-yl)-2-(pentan-3-yl)-1H-benzo[de] isoquinoline-1,3(2H)-dione (2). In a Schlenk tube, 1(500 mg, 1.44 mmol) was added to (2-methoxynaphthalen-1-yl)boronic acid (875 mg, 4.33 mmol), sodium carbonate (459 mg, 4.33 mmol) and Pd(PPh 3 ) 4 (83 mg, 0.07 mmol). After vacuum pumping and flushing with argon (3 10 minutes), THF (40 mL) flushed with argon was added and the mixture was heated at 95 1C overnight. After completion of the reaction, the solvent was evaporated and the mixture was dissolved in CH 2 Cl 2 for purification by silica-gel column chromatography (eluent: CH 2 Cl 2 ) to afford 2as a green oil (562 mg, 92% yield). 1 H NMR (300 MHz, CDCl 3 )d8.70 (d, J= 7.5 Hz, 1H), 8.57 (d, J= 7.2 Hz, 1H), 8.05 (d, J= 9.1 Hz, 1H), 7.90 (d, 1H), 7.73–7.66 (m, 2H), 7.54 (dd, J= 8.4, 7.2 Hz, 1H), 7.47 (d, J= 9.1 Hz, 1H), 7.37 (ddd, J= 8.1, 6.8, 1.3 Hz, 1H), 7.27 (dd, 1H), 7.09 (dd, 1H), 5.11 (td, J= 15.5, 9.8, 5.9 Hz, 1H), 3.79 (s, 3H), 2.36–2.23 (m, 2H), 2.02–1.87 (m, 2H), 0.95 (t, J= 7.5 Hz, 6H). 13 C NMR (76 MHz, CDCl 3 )d154.51, 141.98, 133.78, 132.50, 131.42, 130.71, 129.88, 129.04, 128.89, 128.22, 127.15, 126.89, 124.93, 124.03, 121.02, 113.35, 57.48, 56.66, 56.62, 25.24, 11.50. HRMS (MALDI-TOF): m/zcalculated for C 28 H 25 NO 3 : 423.1843, found: 423.1840 (s= 0.69 ppm). 6-(2-Hydroxynaphthalen-1-yl)-2-(pentan-3-yl)-1H-benzo[de] isoquinoline-1,3(2H)-dione (3). In a Schlenk tube, 2(500 mg, 1.18 mmol) was dissolved in CH 2 Cl 2 (30 mL), put at 0 1C and degassed with argon. After 15 minutes, a 1 M solution of BBr 3 in CH 2 Cl 2 (3.54 mL, 3.54 mmol) was added dropwise and the solution turned dark. After 10 minutes of stirring, the reaction was allowed back to room temperature and stirred overnight as the color became progressively clearer. After completion of the reaction, the mixture was treated with an aqueous solution saturated of NaHCO 3 and extracted with CH 2 Cl 2 . The solvent was evaporated and the crude was dissolved in CH 2 Cl 2 for purification by silica-gel column chromatography (eluent: CH 2 Cl 2 /EtOAc 9/1) to afford 3as a green oil (458 mg, 95% yield). 1 H NMR (300 MHz, CDCl 3 )d8.74 (d, J= 7.4 Hz, 1H), 8.61 (d, J= 7.1 Hz, 1H), 7.96 (d, J= 8.9 Hz, 1H), 7.89 (d, J= 8.1 Hz, 1H), 7.82 (d, J= 7.4 Hz, 1H), 7.76 (dd, J= 8.5, 1.2 Hz, 1H), 7.60 (dd, J= 8.4, 7.2 Hz, 1H), 7.40–7.27 (m, 3H), 7.05 (d, J= 8.4 Hz, 1H), 5.10 (td, 1H), 4.92 (s, 1H), 2.36–2.21 (m, 2H), 2.02–1.86 (m, 2H), 0.94 (t, J= 7.4 Hz, 6H). 13 C NMR (76 MHz, CDCl 3 )d 151.25, 140.24, 133.72, 132.43, 131.38, 130.81, 130.54, 129.05, 128.95, 128.30, 127.46, 127.16, 124.57, 123.80, 117.98, 117.52, 57.64, 25.10, 11.40. HRMS (MALDI-TOF): m/zcalculated for C 27 H 23 NO 3 : 408.1602, found: 408.1605 (D=0.82 ppm). 2-(Pentan-3-yl)-1H-benzo[7,8]xantheno[2,1,9-def]isoquinoline-1,3(2H)-dione (DBXI). 3 (250 mg, 0.61 mmol) and K 2 CO 3 (337 mg, 2.44 mmol) were degassed and flushed with argon (3 10 minutes) in a 20 mL microwave vial. Argon flushed nitrobenzene (20 mL) was added and the mixture was stirred at 200 1C under microwaves for 40 h. The resulting crude was cooled down and put directly in a silica-gel column (eluent: CH 2 Cl 2 /petroleum ether 4/6 to remove most of nitrobenzene, then 6/4 to collect the product). The solvent was then evaporated to afford DBXI as an orange solid (135 mg, 54% yield). 1 H NMR (300 MHz, CDCl 3 )d8.76 (d, J= 8.6, 0.9 Hz, 1H), 8.51 (d, J= 8.1 Hz, 1H), 8.46 (d, J= 8.3 Hz, 1H), 8.33 (dd, J= 8.1, 0.6 Hz, 1H), 7.85–7.77 (m, 2H), 7.58 (ddd, J= 8.6, 6.9, 1.5 Hz, 1H), 7.47 (ddd, J=8.0,6.9,1.1Hz,1H),7.31(d,J= 8.9 Hz, 1H), 7.16 (d, J=8.4Hz,1H),5.03(td,J= 9.5, 5.9 Hz, 1H), 2.31–2.10 (m, 2H), 1.96–1.76 (m, 2H), 0.85 (t, J=7.5Hz,6H). 13 CNMR(76MHz, CDCl 3 )d155.26, 152.53, 133.88, 133.35, 131.83, 130.36, 129.59, 128.39, 125.95, 124.18, 120.78, 119.31, 117.78, 113.63, 109.43, 57.39, 25.13.HRMS (MALDI-TOF): m/zcalculated for C 27 H 21 NO 3 :407.1523, found: 407.1516 (D=1.69ppm). 6-Bromo-7-nitro-2-(pentan-3-yl)-1H-benzo[de]isoquinoline1,3(2E)-dione (5). To a suspension of 4(500 mg, 1.55 mmol) in 2 mL of NMP was added dropwise a solution of 3-aminopentane (0.51 mL, 3.1 mmol) in 1.33 mL of acetic acid. The mixture was heated at 120 1C for 20 minutes. Then the reaction was poured into water and extracted with CH 2 Cl 2 (315 mL). The organic phase was dried over MgSO 4 and evaporated. The product was purified by column chromatography (eluent: CH 2 Cl 2 /petroleum ether 8/2) to afford 5as a pale brown powder was obtained (334 mg, 54% yield). 1 H NMR (300 MHz, CDCl 3 )d8.68 (d, J= 7.9 Hz, 1H), 8.50 (d, J= 8.0 Hz, 1H), 8.21 (d, J= 8.0 Hz, 1H), 7.92 (d, J=7.9Hz,1H),5.07–4.91(m,1H),2.31–2.10 (m, 2H), 2.01–1.81 (m, 2H), 0.89 (t, J=7.5Hz,6H). 6-(Naphthalen-1-yl)-7-nitro-2-(pentan-3-yl)-1H-benzo[de] isoquinoline-1,3(2H)-dione (6). Argon flushed 1,4-dioxane (3.1 mL) and water (0.75 mL) were added over a mixture of 4 (300 mg, 0.77 mmol), 1-naphthylboronic acid (198 mg, 1.15 mmol), potassium carbonate (318 mg, 2.3 mmol) and Pd(PPh 3 ) 4 (44 mg, 0.04 mmol) in a sealed tube under argon. The reaction mixture was stirred at 110 1C for 6 h and monitored by TLC (eluent: CH 2 Cl 2 ). The reaction was cooled at room temperature and extracted with water and CH 2 Cl 2 . The organic phase was collected and dried over MgSO 4 and the solvent was evaporated. The resulting product was purified by column chromatography (eluent: CH 2 Cl 2 ) to obtain 5as a yellow-orange powder (310 mg, 92% yield). 1 H NMR (300 MHz, CDCl 3 )d8.78 (d, J= 7.6 Hz, 1H), 8.68 (d, J= 7.8 Hz, 1H), 8.22– 7.88 (m, 3H), 7.84 (d, J= 7.8 Hz, 1H), 7.64 (d, J= 8.4 Hz, 1H), 7.60–7.51 (m, 1H), 7.50–7.37 (m, 2H), 7.25 (m, 1H), 5.68– 4.57 (m, 1H), 2.65–2.13 (m, 2H), 2.13–1.77 (m, 2H), 0.95 Journal of Materials Chemistry B Paper Open Access Article. Published on 18 July 2024. Downloaded on 1/30/2025 10:46:47 AM. This article is licensed under a Creative Commons Attribution 3.0 Unported Licence. View Article Online 8110 | J. Mater. Chem. B, 2024, 12, 8107–8121 This journal is © The Royal Society of Chemistry 2024 (t, J= 7.5 Hz, 6H). 13 C NMR (76 MHz, CDCl 3 ) 152.1, 142.4, 135.5, 133.9, 133.6, 131.4, 130.5, 129.7, 128.7, 127.3, 127.0, 126.5, 125.3, 124.8, 123.3, 122.6, 58.2, 25.1, 11.5. HRMS (MALDITOF) m/zcalculated for C 27 H 22 N 2 O 4 : 438.1589, found 438.1585 (D= 0.88 ppm). 2-(Pentan-3-yl)benzo[lmn]naphtho[2,1-c][2,8]phenanthroline1,3(2H,6H)-dione (DBAI). Argon flushed 1,2-dichlorobenzene (7 mL) was added to a mixture of 2(250 mg, 0.57 mmol) and triphenylphosphine (449 mg, 1.7 mmol) in a sealed tube under argon. The mixture was stirred for 7 h at reflux. The mixture was cooled down to room temperature and was directly injected on a silica-gel chromatographic column (eluent: CH 2 Cl 2 /EtOAc). A red powder was obtained (yield 91%, 210 mg). 1 H NMR (300 MHz, DMSO-d 6 )d:12.29(s,1H),9.00(d,J= 8.6 Hz, 1H), 8.43 (t, J= 6.8 Hz, 2H), 8.31 (d, J= 8.5 Hz, 1H), 8.09 (d, J= 8.9 Hz, 1H), 8.03 (d, J= 8.0 Hz, 1H), 7.77–7.66 (m, 1H), 7.66–7.38 (m, 2H), 7.08 (d, J= 8.4 Hz, 1H), 5.79–4.51 (m, 1H), 2.29–2.05 (m, 2H), 1.89–1.74 (m, 2H), 0.78 (t, J= 7.4 Hz, 6H). 13 C NMR (126 MHz, DMSO-d 6 )d 170.3, 142.9, 138.4, 137.7, 133.2, 131.2, 130.6, 129.9, 129.4, 128.8, 128.1, 125.3, 123.2, 122.2, 117.4, 116.9, 112.3, 105.7, 79.2, 59.7, 55.6, 24.3, 20.8, 14.1, 11.2. HRMS (MALDI-TOF) m/zcalculated for C 27 H 22 N 2 O 2 406.16763, found 406.16758 (D=0.11ppm). Crystallographic data Crystal data were collected on a Rigaku Oxford Diffraction SuperNova diffractometer equipped with an Atlas CCD detector and micro-focus Cu-Karadiation (l= 1.54184 Å). The structures were solved by dual-space algorithm and refined on F2 by full matrix least-squares techniques using SHELX package (G.M. Sheldrick, ShelXT2018/2, ShelXL2018/3-2019/3). All nonhydrogen atoms were refined anisotropically and the H atoms were included in the calculation without refinement. Multiscan empirical absorption was corrected for DBI and DBAI by using CrysAlisPro program (CrysAlisPro, Rigaku Oxford Diffraction, 2019–2023). Gaussian absorption corrections were applied for DBXI. Deposition number(s) 2083069 (for DBI), 2335339 (for DBXI) and 2335340 (for DBAI) contain(s) the supplementary crystallographic data for this paper.† Photophysical characterization Absorption spectra were obtained using a Jasco V-650 spectrophotometer for diluted solutions (B10 or 1 mM), employing spectrophotometric grade solvents. Fluorescence spectra were acquired with a Horiba Jobin Yvon Fluorolog-3. Steady-state luminescence measurements utilized unpolarized light from a 450 W xenon continuous wave lamp as the light source, with luminescence detected at a 901angle. These measurements were conducted in diluted solutions within a 10 mm Hellma quartz cuvette, employing a Hamamatsu R928 detector for visible light and a liquid nitrogen-cooled, solid-state Indium Gallium Arsenide detector (850–1600 nm) for near-infrared measurements of singlet oxygen phosphorescence signals. Spectra corrections accounted for variations in the excitation source light intensity and emission spectral responses. F F were determined in diluted chloroform (CHCl 3 ) solutions with an absorbance below 0.1, using the following equation for approximation: jfxðÞ jfrðÞ ¼Ar Ax nx 2 nr2Dx Dr (1) where Arepresents the absorbance (or optical density) at the excitation wavelength, nis the refractive index of the solvent, and Dis the integrated luminescence intensity, with ‘‘r’’ and ‘‘x’’ denoting the reference and sample, respectively. The reference compound was coumarin-153 in methanol (F F = 0.45). The excitation of both reference and sample compounds was ideally at the same wavelength. The results reported are the averages of 4–5 independent measurements at various absorbances (ranging from 0.01–0.1) for both the sample and reference. A linearization process was applied to derive an average value with a high coefficient of determination R 2 40.99. Singlet oxygen quantum yield (F D ) determinations followed the same methodology, with the exception that CDCl 3 was consistently used as the solvent (with n x =n r ). 1 O 2 emission was monitored at B1270 nm. Phenalenone (F D = 0.95) served as the reference compound in these measurements. EPR experiments 5,5-Dimethyl-1-pyrroline N-oxide (DMPO) was purchased from TCI Chemicals. All samples were prepared under air atmosphere using capillary tubes filled with 1 10 4 M solutions of each chromophore with 5 10 3 M of the spin trap in DMSO. The irradiations were conducted using a Thorlab LED with a wavelength of 455 nm, which was directly aimed into the EPR cavity during the recording of the spectrum. All EPR assays were performed at room temperature, utilizing a Bruker E500 spectrometer that operated at the X-band frequency of 9.44 GHz, equipped with a standard cavity and a modulation frequency of 100 kHz. The settings of the instrument were adjusted as follows: microwave power at 2–2–69 mW and modulation amplitude at 1 G. The hyperfine coupling constants, including aand g values, were determined through the simulation of experimental spectra using the EasySpin toolbox in Matlab. Computational data The Gaussian16 code 56 was used to optimize the geometries at the ground and excited states along with the global hybrid functional PBE0. 57 This functional was chosen because of its accuracy to reproduce spectroscopic properties of these molecules. Structural optimizations and subsequent frequency calculations for both the ground and excited states were performed using an all electron Pople triple zeta basis set with one polarization function on all atoms and one diffuse function of heavier atoms, known as 6–311+G(d,p), for H, C, N, O and S atoms. Bulk solvent effects were included using the polarizable continuum Model (PCM) of Tomasi and co-workers. 58 Default radii (from the UFF, scaled by 1.1) were used. Excited state geometry was obtained by minimizing the forces of the S 1 state computed at the TD-DFT level by considering the 3 first excited states. The Orca program 59 was used to compute the SOC between the three first triplet states (namely T 1 and T 2 ) and Paper Journal of Materials Chemistry B Open Access Article. Published on 18 July 2024. Downloaded on 1/30/2025 10:46:47 AM. This article is licensed under a Creative Commons Attribution 3.0 Unported Licence. View Article Online This journal is © The Royal Society of Chemistry 2024 J. Mater. Chem. B, 2024, 12, 8107–8121 | 8111 the S 1 state at the S 1 optimized geometry using the quadricresponse Tamm–Dancoff approximation of TD-DFT at the PBE0/PCM level with the Def2-TZVP basis set adapted for the Douglas–Kroll calculations. The spin–orbit coupling was computed using the Douglas–Kroll Hamiltonian along with the spin–orbit mean field approach. Live cell imaging The day prior to treatment, a total of 10 10 4 HeLa cells were plated on a glass-bottom microwell dish (MatTek Corp.) in DMEM medium enriched with 1penicillin–streptomycin and 10% fetal bovine serum, and they were maintained at 37 1C in a 5% CO 2 environment. HeLa cells were washed with DMEM medium (twice), and then treated with DBXI (500 nM) or DBAI (500 nM) for 24 hours. After that, the cells were washed twice with 1phosphate-buffered saline (PBS) before being incubated with the organelle-selective stains. Nuclear staining was accomplished using Hoechst 33342 (500 nM; SigmaAldrich, ref. no.: B2261), while ER-TrackertRed (500 nM; ThermoFisher Scientific, ref. no.: E34250), Lyso-TrackertRed (100 nM; ThermoFisher Scientific, ref. no.: L7528), or MitoTrackertRed (100 nM; ThermoFisher Scientific, ref. no.: M7512) were utilized for organelle-specific staining. These dyes were dissolved in a live cell imaging solution (Molecular Probest, ref. no.: A14291DJ) and incubated with the cells for 30 minutes at 37 1C in a 5% CO 2 environment. Before imaging, the cells were washed twice with the live cell imaging solution. Imaging was conducted using a Leica SP8 FALCON confocal microscope equipped with an incubation chamber set to 37 1C and 5% CO 2 . Maximum intensity projections of Z-stack images were generated for data presentation. All data were subsequently processed using ImageJ software. Intracellular detection of ROS The day prior to treatment, 10 10 4 HeLa cells were cultured on glass-bottom microwell dishes (MatTek Corp.). Subsequently, these cells were exposed to DBXI (500 nM), DBAI (500 nM) or an equivalent volume of DMSO (0.04% v/v) and then incubated at 37 1Cin5%CO 2 for 24 hours. Then, the cells were exposed to CellROXtgreen reagent (5 mM; ThermoFisher Scientific, ref. no.: C10444) and Hoechst 33258 (500 nM) dissolved in the live cell imaging solution, and incubated at 37 1C for 30 minutes. Subsequently, the cells were imaged at two time points: before irradiation and 5 minutes post-irradiation, using a 480 nm supercontinuum white light laser (WLL) operating at 80% for a duration of 5 minutes. Phototoxic experiments HeLa cells, at a density of 5 10 3 cells per well, were placed onto 96-well plates a day prior to the treatment with DBXI or DBAI.DBXI or DBAI were dissolved in complete medium at different concentrations, with DMSO reaching a maximum value of 0.5% v/v, and subsequently added to the cells for 24 hours. Where indicated, the cells underwent photo-irradiation using the EVOS s FL cell imaging system equipped with a customizable LED cube (Invitrogen, ref. no.: AMEP4651; excitation: 470/22 nm) operating at 27 mW cm 2 for 6 minutes. Following photo-irradiation, the cells were further incubated for 24 hours at 37 1C in a 5% CO 2 atmosphere. After 48 hours from the initiation of DBXI or DBAI treatment, PrestoBluet(Invitrogen, ref. no.: A13261) was introduced to each well, and the cells were incubated at 37 1Cin5%CO 2 for an additional three hours. The assessment of cell viability was performed by measuring the fluorescence signal emitted by PrestoBlue (l exc /l em : 560/590 nm) using a Synergy H4 microplate reader (Biotek). The photo-induced cell death caused by DBXI was corroborated using the LIVE/DEADtviability/cytotoxicity kit (Invitrogen, ref. no.: L34960). Cells were plated on glass-bottom microwell dishes (MatTek Corp.) a day prior to treatment. HeLa cells received a treatment of DBXI (500 nM) or an equivalent volume of DMSO and were then incubated at 37 1Cina5%CO 2 atmosphere for 24 hours. Where necessary, the cells underwent photoirradiation. Following this, an additional 24-hour incubation at 37 1Cin5%CO 2 was conducted. 48 hours post-DBXI treatment, live/dead fixable red stain (1 mLmL 1 ) was applied to the cells for 30 minutes at 37 1C before fixation with 4% paraformaldehyde (PFA). Prior to imaging, cells were washed with 1PBS enhanced with 1% bovine serum albumin. Imaging was performed using a Leica SP8 FALCON confocal microscope, employing maximum intensity projection of Z-stack images for data representation. ImageJ software was utilized for all data processing. Morphological changes associated with light irradiation and PSs treatment 24 hours prior to DBXI or DBAI treatment, 5 10 3 cells per well were seeded in complete medium on 96-well plates. DBXI or DBAI, at a concentration of 1 mM, or an equivalent volume of DMSO (0.08% v/v), were dissolved in complete medium and added to the cells for an additional 24 hours at 37 1Cina5% CO 2 environment. Subsequently, the cells were subjected to photo-irradiation using the EVOS s FL cell imaging system equipped with adjustable-intensity LED cubes (excitation = 470/22 nm operating at 30 mW cm 2 ). 3. Results and discussion Molecular design and synthesis The synthetic route initially engineered for the synthesis DBI had to be entirely redesigned for the preparation of the new Scheme 1 Synthetic routes and chemical structures of DBXI (top) and DBAI (bottom). Journal of Materials Chemistry B Paper Open Access Article. Published on 18 July 2024. Downloaded on 1/30/2025 10:46:47 AM. This article is licensed under a Creative Commons Attribution 3.0 Unported Licence. View Article Online 8112 | J. Mater. Chem. B, 2024, 12, 8107–8121 This journal is © The Royal Society of Chemistry 2024 target derivatives. Starting with the oxo derivative, preparation of the latter was initiated by reacting the bromonaphtalimide 1 with the commercially available (2-methoxynaphthalen-1-yl) boronic acid 2under Suzuki–Miyaura coupling conditions. The methoxy group of the resulting intermediate 3was then deprotected with BBr 3 prior to ring close the system under basic conditions to afford the desired DBXI derivative (Scheme 1). Regarding its NH counterpart, a different approach was developed, relying on the use of the 4-bromo-5-nitro-1,8naphthalic anhydride. Starting with its imidization in presence of the 3-aminopentane, the resulting intermediate 6was thereafter engaged in a Suzuki–Miyaura cross coupling reaction with the commercial 1-naphthaleneboronic acid 7to afford 8. Finally, 8was cyclized under Cadogan conditions providing the desired DBAI compound. The structures of both DBXI and DBAI were investigated using NMR spectroscopy and HRMS (Fig. S1–S16, ESI†). These data were confirmed through X-ray diffraction performed on crystals grown using the slow evaporation technique (Table S1, ESI†and experimental section for details). Both molecules exhibited overall good solubility in common organic solvents, with DBAI outperforming DBXI, facilitating the investigation of their spectroscopic and photophysical properties. Photophysical properties The absorption and emission spectra of DBXI and DBAI were recorded in CHCl 3 and compared with those of the parent DBI compound (Fig. 1). This revealed a significant impact of the substituent change on the spectral shape and the position of the maxima within the series (Fig. 1). Whereas the parent DBI exhibited structureless absorption and emission bands, with a maximum absorption (l max ) at 481 nm and maximum emission (l em ) at 544 nm, the two new compounds displayed notable differences. Both DBXI and DBAI had absorption and emission profiles characterized by pronounced vibronic progressions, indicating a more localized character of the p–p*transition as typically observed in closely related naphthalene diimide molecules. Specifically, DBXI exhibited a notable blue shift, with l max at 447 and 472 nm and l em at 494 nm, compared to DBI. Conversely, the spectral maxima of DBAI (l max = 487 and 517 nm and l em = 543 nm) were very similar to those of the parent DBI. Additionally, DBAI demonstrated a minor Fig. 1 Absorption and emission spectra of 10 mM CHCl 3 solutions of DBAI, DBXI,andDBI.l exc (DBAI) = 480 nm, l exc (DBXI) = 446 nm and l exc (DBI)= 480 nm. Fig. 2 Summary of TD-DFT calculations for (A) DBAI,(B)DBXI and (C) DBI. Variation of the electron density, Dr, presenting the electron density depletion and the electron increase in blue and green respectively (isovalue 0.004 a.u.) as long as the extract D CT index. Jablonsky diagram is presented with S 0 ,S 1 and T n levels in black, blue and green respectively. The torsion angle of the p-systems is presented in inset. Paper Journal of Materials Chemistry B Open Access Article. Published on 18 July 2024. Downloaded on 1/30/2025 10:46:47 AM. This article is licensed under a Creative Commons Attribution 3.0 Unported Licence. View Article Online This journal is © The Royal Society of Chemistry 2024 J. Mater. Chem. B, 2024, 12, 8107–8121 | 8113 transition of low intensity with a similar band structure, peaking at 420 nm. This evolution was corroborated by vertical transition energies computed at the TD-DFT level, which yielded absorption wavelengths of 461 nm for DBAI, 436 nm for DBXI, and 470 nm for DBI, along with emission wavelengths of 527 nm for DBAI, 506 nm for DBXI, and 545 nm for DBI. The more localized character of the electronic transition was further confirmed by the computation of the D CT index, which is related to the variation in electron density between the ground and excited states and quantifies the strength of a charge transfer transition. The later was found to be larger for DBI than for DBXI and DBAI, indicating a stronger charge transfer transition in DBI as illustrated in Fig. 2. 60,61 The most striking effects of changing the bridging heteroatom were observed in the photophysical properties of the molecule. While DBI exhibited very efficient intersystem crossing (ISC), with F D of approximately 0.95 and minimal residual luminescence, the ISC efficiency was significantly reduced for DBXI and DBAI to F D values of ca 0.21 and 0.33, respectively. As an antagonistic effect, the F F significantly increased to 0.68 for DBXI and up to 0.72 for DBAI, associated to fluorescence lifetimes of 4.95 ns for DBXI and 6.7 ns for DBAI. Quantum chemical calculations provided atomistic insights into these observations. The ISC, responsible for populating the triplet state necessary for generating 1 O 2 , is still promoted by SOC Fig. 3 EPR signal of a solution of DBAI and DMPO in DMSO under 455 nm irradiation. Experimental (blue) and simulated (red) spectra of DMPO superoxide adduct. g= 2.006, aN = 12.85 G, aH = 10.37 G, aH = 1.28 G. Fig. 4 Cellular localization of DBXI and DBAI. CLSM images of live HeLa cells treated with DBXI (500 nM, green signal) (A) or DBAI (500 nM, green signal) (B) and incubated for 24 hours. After 24 hours, the cells were co-stained with the nuclear dye Hoechst 33342 (500 nM, blue signal) and Lyso-Tracker Red (100 nM, red signal). 2D scatter diagrams depicting DBXI or DBAI and Lyso-Tracker Red channels are presented, and the quantification of colocalization has been accomplished using the Pearson correlation coefficient (PCC). Image settings: l exc /l em : 405/420–470 nm for Hoechst; l exc /l em : 480/495– 560 nm for DBXI or DBAI;l exc /l em : 577/600–710 nm for Lyso-Tracker Red. The scale bar is set at 25 mm for regular images (top panel) and 10 mm for the enlarged images (bottom panel; indicated zoom). Journal of Materials Chemistry B Paper Open Access Article. Published on 18 July 2024. Downloaded on 1/30/2025 10:46:47 AM. This article is licensed under a Creative Commons Attribution 3.0 Unported Licence. View Article Online 8114 | J. Mater. Chem. B, 2024, 12, 8107–8121 This journal is © The Royal Society of Chemistry 2024 resulting from the distortion of the p-systems, also referred to as the twisting induced triplet state population. This distortion was computed for all three molecules (Fig. 2) and observed experimentally from X-ray structures. As expected, this effect was more pronounced for DBI than for DBAI and DBXI as characterized by the largest dihedral angle (by more than 101) measured between the upper and lower naphthyl rings (Fig. 2). This feature is directly related to the higher SOC computed for DBI, compared to the other two molecules, and ultimately the variations in 1 O 2 production efficiency that were experimentally evaluated. To determine if, beyond a type-II mechanism driven by 1 O 2 , a type-I mechanism could also play a role in generating ROS, electron paramagnetic resonance (EPR) experiments were conducted on both DBAI and DBXI. Utilizing 5,5-dimethyl-1pyrroline N-oxide (DMPO) as a radical scavenger, capable of reacting with O 2 , no radical adducts were detected for DBXI. In contrast, DBAI exhibited a well-resolved hyperfine splitting structure, indicative of a DMPO-superoxide adduct signal, corroborated by simulated EPR spectra (Fig. 3). These findings suggest that the photocytotoxic potential of these compounds might be exclusively mediated through a type-II mechanism for DBXI, while DBAI could involve a combined type-I and type-II mechanism. Live cells fluorescence imaging Organelles play pivotal roles in preserving cell structure and function, and any disruption can result in cellular dysfunction and eventual cell death. 62 Achieving precise targeting of PSs to organelles offers several advantages, including the potential to lower PS dosages, mitigate side effects, reduce the risk of drug resistance, and enhance the effectiveness of PDT. 11 Indeed, given that ROS typically have an exceedingly brief lifetime and a limited diffusion range in biological systems, 63 PSs designed to precisely target organelles often exhibit superior PDT outcomes. 64 In this context, a number of organelle-specific PSs already have been reported with localization to the nuclei, 31,41,65–67 mitochondria, 20,68–70 multivesicular bodies, 55 lysosomes 19,71–74 and endoplasmic reticulum (ER). 75–79 Next, we thus investigated if the high fluorescence emission demonstrated by these new PSs could be used to probe their cellular emission fingerprint by confocal laser scanning microscopy (CLSM). Live HeLa cells were treated with DBXI or DBAI at a concentration of 500 nM, followed by a 24-hour incubation. Fig. 5 Cellular co-localization of DBXI and DBAI with ER tracker. CLSM images of live HeLa cells treated with DBXI (500 nM, green signal) (A) or DBAI (500 nM, green signal) (B) and incubated for 24 hours. After 24 hours, HeLa cells were co-stained with the nuclear dye Hoechst 33342 (500 nM, blue signal) and ER-Tracker Red (500 nM, red signal). 2D scatter diagrams depicting DBXI or DBAI and ER-Tracker Red channels are presented, and the quantification of colocalization has been accomplished using PCC. Image settings: l exc /l em : 405/420–470 nm for Hoechst; l exc /l em : 480/495–560 nm for DBXI or DBAI;l exc /l em : 587/600–710 nm for ER-Tracker Red. The scale bar is set at 25 mm for regular images and 10 mm for the enlarged images. Paper Journal of Materials Chemistry B Open Access Article. Published on 18 July 2024. Downloaded on 1/30/2025 10:46:47 AM. This article is licensed under a Creative Commons Attribution 3.0 Unported Licence. View Article Online This journal is © The Royal Society of Chemistry 2024 J. Mater. Chem. B, 2024, 12, 8107–8121 | 8115 The incubation time was carefully selected based on the results of phototoxicity experiments (detailed below) to coincide precisely with the period during which light is applied to induce cell death. This synchronization ensures consistent PS accumulation and distribution within cellular compartments, thereby avoiding variability in PS concentration and localization that could affect the interpretation of the experimental outcomes. This treatment initially revealed an accumulation of the PSs in the cytoplasm, confirming that the molecules were well dissolved within the cellular environment and isolated as highly emissive species (Fig. 4–6). Thereafter, the intracellular distribution of the PSs was further explored by using four different commercially available organelle-selective live-cell trackers, namely Hoechst 33342, ER-Tracker Red, LysoTracker Red, and Mito-Tracker Red to specifically label the nuclei, ER, lysosomes, and mitochondria, respectively. No relevant nuclear colocalization was observed between the PSs and Hoechst 33342, suggesting negligible uptake in the nucleus (Fig. 4–6). Also, low to moderate Pearson’s correlation coefficients (PCC) were observed between the PSs and Lyso-Tracker Red, indicating their limited accumulation within the lysosomes (Fig. 4). On the other hand, and as depicted in Fig. 5, the green fluorescence of DBXI and DBAI exhibited significant overlap with the fluorescence of ER-Tracker Red, yielding PCC values of 0.59 and 0.82 for DBXI and DBAI, respectively whereas the degree of colocalization between DBXI or DBAI and MitoTracker Red was lower, with PCC values of 0.3 and 0.61 for DBXI and DBAI, respectively (as seen in Fig. 6). These results highlight a good extent of PS internalization within the ER. Moreover, it is noteworthy that DBXI exhibited distinct, intense fluorescent clusters with a punctate nature in the cytoplasm, which did not overlap with the signals emitted by any of the organelle-specific trackers employed, suggesting further accumulation into vesicular bodies (Fig. 4–6). Indeed, the PCC value of 0.59 calculated for the co-localization of DBXI and ERTracker Red would be expected to be higher if the punctate signal of DBXI were excluded from the analysis. We believe that the slight differences in the staining patterns observed for DBXI and DBAI can be partially attributed to their differing hydrophobic properties. DBAI, which is fairly soluble, predominantly demonstrated uniform accumulation in the ER. In contrast, DBXI, due to its higher hydrophobicity, also tended to accumulate in vesicles with likely lipophilic character. These Fig. 6 Cellular localization of DBXI and DBAI. CLSM images of live HeLa cells treated with DBXI (500 nM, green signal) (A) or DBAI (500 nM, green signal) (B) and incubated for 24 hours. After 24 hours, cells were co-stained with the nuclear dye Hoechst 33342 (500 nM, blue signal) and Mito-Tracker Red (100 nM, red signal). 2D scatter diagrams depicting DBXI or DBAI and Mito-Tracker Red channels are presented, and the quantification of colocalization has been accomplished using PCC. Image settings: l exc /l em : 405/420–470 nm for Hoechst; l exc /l em : 480/495–560 nm for DBXI or DBAI;l exc /l em : 580/ 590–715 nm for Mito-Tracker Red. The scale bar is set at 25 mm for regular images and 10 mm for the enlarged images. Journal of Materials Chemistry B Paper Open Access Article. Published on 18 July 2024. Downloaded on 1/30/2025 10:46:47 AM. This article is licensed under a Creative Commons Attribution 3.0 Unported Licence. View Article Online