A Rationally Designed Azobenzene Photoswitch for DNA GQuadruplex Regulation in Live Cells

Photopharmacology A Rationally Designed Azobenzene Photoswitch for DNA GQuadruplex Regulation in Live Cells Marta Dudek,* Lucía López-Pacios, Nasim Sabouri, Juan J. Nogueira, Lara Martinez-Fernandez, and Marco Deiana* Abstract: G-quadruplex (G4) DNA structures are increasingly acknowledged as promising targets in cancer research, and the development of G4-specific stabilizing compounds may lay a fundamental foundation in precision medicine for cancer treatment. Here, we propose a light-responsive G4-binder for precise modulation of drug activation, providing dynamic and spatiotemporal control over G4-associated biological processes contributing to cancer cell death. We developed a specialized fluorinated azobenzene (AB) switch equipped with a quinoline unit and a positively charged carboxamide side chain, Q-Azo4F-C, designed for targeted binding to G4 structures within cells. Biophysical studies, combined with molecular dynamics simulations, provide insights into the unique coordination modes of the photoswitchable ligand in its trans and cis configurations when interacting with G4s. The observed variations in complexation processes between the two isomeric states in different cancer cell lines manifest in more than 25-fold reversible cytotoxic activity. Immunostaining conducted with the structure-specific G4 antibody (BG4), establishes a direct correlation between cytotoxicity and the varying extent of G4 induction regulated by the two isoforms. Finally, we demonstrate the photo-driven reversible regulation of G4 structures in lung cancer cells by Q-Azo4F-C. Our findings highlight the potential of light-responsive G4-binders in advancing precision cancer therapy through dynamic control of G4-mediated pathways. Introduction Four-stranded G-quadruplex (G4) DNA structures are intricately linked to critical biological processes, such as DNA replication, transcription, and telomere maintenance, with profound implications for cancer.[1] Computational and experimental mapping of G4s in the human genome has revealed an enrichment of G4-forming sequences in genomic regions associated with disease states.[2] Within the dynamic milieu of live cells, G4s emerge as significant players capable of impeding DNA transactions, acting as stable roadblocks to the replicative helicase and polymerase processivity.[3] This unique characteristic opens a window of opportunities to introduce structure-specific DNA damage through the strategic use of tailored G4-stabilizers.[4] G4-interacting compounds show promising therapeutic potential in synthetic lethality strategies, and are particularly effective in DNA damage response (DDR)-deficient cells, especially in combination with DNA-damaging or DDR-inhibiting drugs. G4s have also become a significant focal point in phototherapeutic approaches for cancer treatment, such as photodynamic therapy due to their unique photophysical properties such as low-energy photoionization.[5] Their susceptibility to oxidation can indeed be harnessed in conjunction with G4-selective photosensitizers to optimize oxidative DNA damage.[6] With regard to small molecule design, integrating light-responsive moieties into the structure of G4-binding ligands enables the remote-controlled modulation of their activity.[7] Indeed, as an external targeting stimulus, light provides several distinct advantages, encompassing minimal toxicity, precise spatiotemporal control over drug activation by finely tuning the power, duration, and even the wavelength of the excitation source.[8] [*] Dr. M. Dudek, Dr. M. Deiana Institute of Advanced Materials, Faculty of Chemistry Wrocław University of Science and Technology Wyb. Wyspiańskiego 27, 50-370 Wrocław (Poland) E-mail: [email protected] [email protected] L. López-Pacios, Dr. J. J. Nogueira Departamento de Química, Facultad de Ciencias Universidad Autónoma de Madrid, Campus de Excelencia UAMCSIC Cantoblanco, 28049 Madrid (Spain) Prof. Dr. N. Sabouri, Dr. M. Deiana Department of Medical Biochemistry and Biophysics Umeå University SE-901 87 Umeå (Sweden) Dr. J. J. Nogueira Institute for Advanced Research in Chemistry (IAdChem) Universidad Autónoma de Madrid, Campus de Excelencia UAMCSIC Cantoblanco, 28049 Madrid (Spain) Dr. L. Martinez-Fernandez Departamento de Química Física de Materiales, Instituto de Química Física Blas Cabrera, CSIC 28006, Madrid (Spain) © 2024 The Authors. Angewandte Chemie International Edition published by Wiley-VCH GmbH. This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited. Angewandte Chemie Research Article www.angewandte.org How to cite: Angew. Chem. Int. Ed. 2025,64, e202413000 doi.org/10.1002/anie.202413000 Angew. Chem. Int. Ed. 2025,64, e202413000 (1 of 12) © 2024 The Authors. Angewandte Chemie International Edition published by Wiley-VCH GmbH In this context, the Balasubramanian lab reported the use of a caged derivative of the renowned G4-stabilizing compound pyridostatin (PDS) to regulate the transcription of G4containing genes.[9] Vilar’s group has developed a rotaxane incorporating a square planar PtII-salphen complex with photon-dependent G4-binding activity.[10] Recently, we investigated the photoinduced macrocyclization of a helicenoid quinoline derivative of binaphthol, resulting in a πextended dicationic compound.[7g] The latter demonstrated the capacity to stall DNA polymerase by selectively stabilizing G4 structures linked to oncogenic promoters and telomere repeat units. In a series of papers, the Galan lab explored the G4-interacting binding properties of a photoresponsive stiff-stilbene ligand using both experimental and computational approaches.[11] While these studies have demonstrated some degree of success, their limitations, such as the requirement for UV light (�365 nm) to excite the ligand’s active form and/or the absence of a reversible process, make them far from ideal for usage in cells. Molecular switches, incorporating the versatile azobenzene (AB) unit,[7e] offer a promising avenue to overcome the existing limitations. In this context, the pioneering work of Zhou’s group opened up new possibilities for regulating G4 structures derived from telomeric sequences using AB switches decorated with quaternary ammonium groups, including aliphatic and aromatic species.[7b,c,e,12] The authors demonstrated robust G4 folding-unfolding activity associated with the isomeric states of the photochrome between the trans and cis configurations.[7b,c,e] The same group further applied this principle to manipulate enzymatic reactions, demonstrating the potential to regulate thrombin activity by modulating the conformation of a DNA-based inhibitor system.[7e,12] This system consists of a central telomeric G4forming sequence acting as a regulatory element fused with two thrombin-binding aptamers that cooperatively prevent fibrinogen from binding to thrombin. Depending on either the geometric state of the photoswitch[7e] or host-guest supramolecular interactions with a macro-cavitant,[12] the telomeric G4-acting element could fold or unfold, providing reversible control over the thrombin-fibrinogen reaction. However, up to now, these studies have relied on UV light to induce trans-to-cis isomerization and have been limited to test-tube settings, not yet demonstrating their full potential in biological systems, such as cell studies. AB functionalized with ortho-fluorine atoms has a lower energy of the n-orbital of the cis-isomer, resulting in a separation of the trans and cis isomer’ n!π* absorption bands, corresponding to absorption to the S1electronic state.[13] This provides switching possibilities with visible light, avoiding the need for UV irradiation. Recently, we demonstrated that these types of fluorinated ABs could target G4s, with the recognition process and optical output modulated by changing the isomeric states of the switch using only visible light.[7f] However, these results were again limited to test-tube settings. Consequently, we decided to investigate the potential of this class of photochromes in live cell studies. In this work, we developed an ortho-fluorinated AB switch with nearly quantitative two-way isomerization using visible light, making it suitable for live cell applications. This compound (Q-Azo4F-C, Figure 1A), equipped with a quinoline unit on one side and a positively charged carboxamide side chain on the other, is designed to enhance both its G4-interacting binding properties and its uptake in cancer cells. The trans and cis forms of Q-Azo4F-C exhibit distinct coordination modes associated with the isomerization process. Biophysical studies, molecular dynamics (MD) simulations, and G4 immunodetection techniques in cancer cells show that the cis isomer promotes G4 formation through external binding. In contrast, the trans isomer partially unfolds G4s by intercalating into the central guanine (G)-tetrad. The ability to fold and unfold G4s correlates positively with the cytotoxic effects of the trans and cis isomers, respectively, across various cancer cell lines. Moreover, the compound enables in-cell photoregulation of G4s, which is achieved by converting the G4-forming and cytotoxic cis isomer into a non-toxic, non-G4-forming trans state. Results and Discussion Design and Photochemical Characterization of the Molecular Switch Q-Azo4F-C compound consists of a quinoline unit and a positively charged carboxamide side chain, directly linked to the para position of the AB core. This synthetic approach aims to functionalize the molecule with two flexible scaffolds, each potentially influencing the G4 in a complementary or opposing manner, depending on the switch‘s geometric configuration. In fact, while the quinoline moiety, exemplified by PhenDC3[14] and PDS,[15] is known for stabilizing G4 structures, positively charged amino groups[16] can destabilize/unfold G4s. Building on this knowledge, QAzo4F-C (Figure 1A) was synthesized in nine steps, including a bromination/cyanation/hydrolysis/esterification/oxidative coupling/hydrolysis/followed by two final amide coupling reactions and deprotection of the amine groups (see Supporting Information pp. S6–S9 for details concerning synthesis). The UV/Vis spectrum of Q-Azo4F-C in the trans state exhibited a prominent π!π* absorption peak at 335 nm and a less intense n!π* band around 472 nm (Figure 1B). Exposure to visible light (λ�550 nm) induces trans-to-cis isomerization, leading to a reduction in the π!π* band and an increase in the n!π* absorption, coinciding with a notable blue-shift (~425 nm) of approximately 50 nm. The distinction in the n!π* bands between the two isomers facilitates the isomerization of the switch using visible light in both directions. This process generates photostationary states (PSSs) comprising 93% cis isomer when exposed to green light (λ�550 nm) and 92% trans isomer under blue light (436 nm), as quantitatively determined by 1H and 19F nuclear magnetic resonance (NMR) spectra analysis (Figures 1C, S5 and S6). Notably, Q-Azo4FCin the cis form exhibited high thermal stability (Figure 1D, Supporting Information pp. S11), boasting a half-life of approximately 62 days at 25°C, indicative of an activation Angewandte Chemie Research Article Angew. Chem. Int. Ed. 2025,64, e202413000 (2 of 12) © 2024 The Authors. Angewandte Chemie International Edition published by Wiley-VCH GmbH 15213773, 2025, 1, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/anie.202413000 by Wroclaw University Of Science, Wiley Online Library on [30/01/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License barrier of 102 kJ/mol, as estimated by the Arrhenius equation (Table S1). Photoregulation of G4-Associated Binding Mechanisms The binding affinity of Q-Azo4F-C for G4s in either trans or cis configurations (Table S2) was quantified using a fluorescence quenching assay (Figure 2A). This assessment involved monitoring the change in fluorescence intensity induced by Q-Azo4F-C on a Cyanine 5 (Cy5)-labeled G4 oligonucleotide (c-MYC Pu22) derived from the promoter region of the cancer-associated MYC gene (Figure 2A).[17] Nonlinear curve-fitting procedures, based on fluorescence quenching mediated by proximal ligand binding at the 5’-Gtetrad end, were applied to both isomers. These procedures demonstrated robust binding strengths to the G4 template, with association constants of 2.5×106M-1 for the trans isomer and 3.5×105M1for the cis-rich mixture. These findings were further corroborated by microscale thermophoresis (MST) experiments using the natural G4-forming sequence found in the promoter region of hypoxia-inducible factor 1 alpha (HIF-1α), a master regulator in the pathogenesis of cancer.[18] These experiments provided association constants of 1.0×106M1for the trans isomer and 8.4×104M1for the cis-rich mixture (Figure S8). The difference in binding strength can be attributed to the distinct degrees of planarity exhibited by these two isomeric forms. The structure of the trans form, indeed, facilitates easier establishment of πstacking interactions with the G-tetrads. Conversely, the distorted configuration of the cis form partially impedes πstacking interactions resulting in a coordination mode that likely engages the G-tetrads only partially, primarily relying on groove-binding and electrostatic attractions. As discussed later, this fact is corroborated by the computational simulations. To gain a more comprehensive insight into the structural basis of the interaction between Q-Azo4F-C and G4s, 1H NMR experiments were conducted (Figures S9– S15) by using the well-characterized telomeric G4 sequences Tel22-Na+(Figures S10–S12) and Tel23-K+(Figures S13– Figure 1. (A) Chemical structures of Q-Azo4F-C in trans or cis forms. (B) Absorption spectra of Q-Azo4F-C recorded in DMSO solution under excitation with light of different wavelengths (cQ-Azo4F-C =30 μM). (C) Quantification of the PSSs of Q-Azo4F-C (cQ-Azo4F-C =3 mM, in methanol-d4 at 25°C) by 19F NMR spectroscopy. The contents of trans and cis isomers in solution were calculated from the intensity ratios of the integrals of the corresponding peaks. (D) UV/Vis absorption spectra for 30 μM solution of Q-Azo4F-C at 70°C in DMSO in the PSS after�550 nm irradiation (blue curve), and spectral evolution during the cis-trans thermal return. The inset presents absorption changes monitored at 334 nm during cis-trans thermal relaxation. Angewandte Chemie Research Article Angew. Chem. Int. Ed. 2025,64, e202413000 (3 of 12) © 2024 The Authors. Angewandte Chemie International Edition published by Wiley-VCH GmbH 15213773, 2025, 1, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/anie.202413000 by Wroclaw University Of Science, Wiley Online Library on [30/01/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License S15) as well as c-MYC Pu22 (Figures 2B–D and S9).[17,19] In its unbound state, c-MYC Pu22 displayed 12 imino proton peaks, with their assignments having been reported previously.[19c] Addition of Q-Azo4F-C in its trans configuration induced a chemical shift alteration in almost all imino protons associated with the terminal G-tetrad ends, along with certain ones connected to the central G-tetrad (Figure 2B). Specifically, our findings revealed that QAzo4F-C significantly impacted G16, G11, and G20, along with G9, G13, and G18 linked to the 5’- and 3’-ends, respectively. Additionally, it affected G21 and G17 associated with the central G-tetrad. Notably, a progressive broadening of the imino signals was observed during the titration, resulting in the detection of only 6 out of the initial 12 imino protons at equimolar concentrations of Q-Azo4F-C and c-MYC Pu22. These observations suggest the potential for the trans form to coordinate the c-MYC Pu22 template through not only the two G-tetrad ends but also the central G-quartet. A comparison of these findings with recently reported NMR data for photochromic compounds interacting with G4s hints at the possibility that the trans isomer might have induced a partial unfolding or opening of the G4 structure via an unconventional intercalative binding mechanism.[11b] Interestingly, 1H NMR experiments performed with Tel23-K+and Q-Azo4F-C in trans configuration show not only broadening and annulation of specific resonances but also the appearance of new signals, which may be attributed to the eventual emergence of misfolded Figure 2. Biophysical data demonstrate the interaction between Q-Azo4F-C and G4s. (A) Fluorescence changes of Cy5-c-MYC Pu22 G4 (cG4 =50 nM, 50 mM Tris pH 7.2, 100 mM KCl) in the presence of the incremental addition of Q-Azo4F-C in either trans (red dots) or cis-rich PSS (blue dots). The competitive assay was performed in the presence of 100 eq. of dsDNA for trans (pink dots) or cis-PSS (pale blue dots). (B–C) 1H NMR spectra of the G-tetrad imino protons in the absence (0.0 eq) or presence (0.2, 0.4, 0.8, and 1.0 eq) of Q-Azo4F-C trans (B) and of Q-Azo4F-C cis-rich mixture (C). (D) 1H NMR spectra of c-MYC Pu22 alone (grey), c-MYC Pu22 with 1.0 eq of Q-Azo4F-Ctrans (red) or Q-Azo4F-CPSS>550 nm (blue). 1H NMR spectra representing in situ photoswitching were taken after irradiation with 436 nm (cis-to-trans isomerization, pink), and clearly show the change in imino protons shift, or after irradiation with �550 nm (trans-to-cis isomerization, light blue), and show minor change in imino protons shift. The red and blue circles are provided to aid in data interpretation, highlighting which spectra should align if the system achieves full reversibility. Angewandte Chemie Research Article Angew. Chem. Int. Ed. 2025,64, e202413000 (4 of 12) © 2024 The Authors. Angewandte Chemie International Edition published by Wiley-VCH GmbH 15213773, 2025, 1, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/anie.202413000 by Wroclaw University Of Science, Wiley Online Library on [30/01/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License states, as previously reported (Figure S13).[11b] Similar experiments carried out with the c-MYC Pu22 titrated with the cis-rich mixture of Q-Azo4F-C showed minimal impact on the guanines linked to the 5’-end (Figure 2C). However, it notably affected G13 and G18 associated with the 3’-end, as well as G17 linked with the central G-tetrad. Additionally, all 12 imino proton peaks attributed to c-MYC Pu22 remained distinct at the conclusion of the titration involving the cis-rich mixture. These data suggest that the binding of the cis form to the c-MYC Pu22 template likely involved the 3’-end, resulting in minimal perturbation of the G4 structure. It is noteworthy that the cis-to-trans isomerization, triggered by light, causes the broadening of imino proton signals together with attenuation of G9 and G16 resonances particularly observed in the trans-c-MYC Pu22 system (Figure 2D). However, the reversal of trans-to-cis isomerization did not undo the alterations associated with the cis-c-MYC Pu22 complexation process. This is likely due to the trans isomer‘s tendency to remain firmly bound within the G4 intercalation site, consistent with its high binding strength. Overall, these findings indicate a significant level of reversibility in the process, albeit not easily attainable bidirectionally. To further explore the binding modes and to compute the interaction free energies of Q-Azo4F-C to c-MYC Pu22, we conducted molecular docking and classical molecular dynamics (MD) simulations. During the first docking studies, using as target receptor the c-MYC Pu22 structure resolved by NMR (PDB code 1XAV.pdb[20]), two main external poses, based on their abundance, were obtained for both cis and trans forms (Figure 3A and B). Aiming to investigate the intercalated possibility, the c-MYC Pu22 structure was opened through a combination of MD and umbrella sampling techniques (see pp. S26–S27 of the SI). After equilibration of the final opened structure, new docking calculations provided the most populated internal intercalated cis and trans poses (Figure 3A and B). Then, for the three docking poses (two external and one internal), MD simulations combined with free-energy calculations (see pp. S37–S38 of the SI) were run to calculate the binding free energies of the cis and trans isomers and to decompose the energy into the contributions from the different nucleobases. By comparison of the most stable pose of each isomer (Table S11), the trans state interacts in a stronger way (106.6 kcal/mol) than the molecule in the cis state (94.8 kcal/mol), in qualitative agreement with the association constants determined by the MST measurements. For the trans isomer, the intercalative pose is energetically favored over the two external poses. The energy decomposition analysis shown in Figure 3C reveals that the intercalative binding is dominated by the interaction with the three tetrads (186.6 kcal/mol), while the interaction with the external nucleobases is less than half in magnitude (82.3 kcal/mol). In a more specific way, the intercalative Figure 3. Schematic representation of the most probable cis (A) and trans (B) poses obtained during the docking studies together with the binding energies from subsequent MD simulations. The densities of the point clouds indicate the occurrence of the different binding poses. Per-nucleobase energy decomposition analysis for the intercalated (C) and the most favourable external (D) poses. The numbers above the bars indicate the contribution from the external and tetrad nucleobases. In turn, the last one is split into contributions from the central and 5’- and 3’-end tetrads for the cis and trans isomers. Angewandte Chemie Research Article Angew. Chem. Int. Ed. 2025,64, e202413000 (5 of 12) © 2024 The Authors. Angewandte Chemie International Edition published by Wiley-VCH GmbH 15213773, 2025, 1, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/anie.202413000 by Wroclaw University Of Science, Wiley Online Library on [30/01/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License binding of the trans isomer is controlled by the interaction with the central tetrad (115.9 kcal/mol), especially with G17, although the interactions with the 5’- and 3’-end ones are not negligible, again in consonance with the shifts observed in the NMR experiments. On the contrary, for the cis isomer, the external and intercalative poses will co-exist, in the case that both are kinetically favored, although with a slight predominance of the external one, which is 2 kcal/mol more stable. In this case, as displayed in Figure 3D, the interaction between the cis structure and G4 in the external pose has similar contributions from the external nucleobases (121.6 kcal/mol) and from the ones forming the three tetrads (129.7 kcal/mol). As mentioned above, the NMR data suggest that the binding of the trans isomer might have induced a partial unfolding of the G4, while the binding of the cis isomer resulted likely in only a small structural perturbation. This hypothesis is corroborated by the MD simulations, where the G4 structure is better preserved along the dynamics for the external pose of the cis isomer than for intercalated trans isomer (see Figures S45 and S61 of SI). The structural data obtained for Q-Azo4F-C complexed with various G4 structures prompted us to further investigate G4 regulation using complementary biophysical methods. Circular dichroism (CD) titration experiments, performed on parallel (HIF-1α,c-MYC Pu22 and c-MYC Pu24), hybrid (Tel-22 in K+), and antiparallel (TBA and Bom17) G4 structures, revealed that the trans isomer induced pronounced spectral changes, especially in hybrid and antiparallel G4 topologies, indicative of a G4 unfolding binding event (Figures S16–S17). In contrast, the cis-rich mixture exhibited minimal spectral changes to these G4 structures compared to the trans isomer (Figures S16–S17). It is likely that the changes observed for the cis-rich mixture may be partially attributed to the residual trans isomer present in the solution (~10%) after photoconversion. Interestingly, almost no or very weak changes were detected for Q-Azo4F-C bound to duplex DNA in either configuration, providing initial evidence of a degree of selectivity for G4s over duplex DNA (Figure S18). To further investigate the roles of the trans and cis isomers in controlling the folding and unfolding states of G4s, we conducted CD-based thermal melting assays, measuring molar ellipticity as a function of increasing temperature. The melting temperatures (ΔTm) indicated a higher degree of G4 stabilization in the presence of the cis-rich mixture of Q-Azo4F-C compared to the trans form, while both isomers displayed a negligible effect on duplex melting temperature (Figure 4A). Overall, the CD data clearly support the potential foldingunfolding event mediated by the photochrome and demonstrate selectivity for G4 over duplex DNA. This selectivity was further investigated through competitive fluorescence titration experiments to test the binding selectivity of G4 ligands. These experiments employed a large excess of 100 equivalents of duplex DNA compared to the G4 concentration. As depicted in Figure 2A, the presence of dsDNA did not affect the binding affinity of the trans and cis-rich mixtures of Q-Azo4F-C, indicating a higher binding affinity for G4 DNA over dsDNA. To rule out any potential interference from duplex DNA on the coordination mode of the photochrome to G4s, we conducted 1H NMR competitive binding experiments using c-MYC Pu22 as a G4 model template. The detected chemical shift perturbations of the imino protons in the Q-Azo4F-C (trans or cis-rich PSS)-G4 complexes were unaffected by duplex DNA, confirming the ligand’s higher affinity for G4 structures and demonstrating that isomeric states can precisely trigger conformational changes in the G4 template despite the presence of duplex DNA (Figures 4B and S19). The complex intracellular milieu contains various metal ions that can promote conformational alterations in G4 structures, potentially affecting or hindering ligand binding.[7b,c] To address this, we investigated the role of the most abundant cellular ions (K+, Na+, Mg2+, and Ca2+),[21] both individually and in combination, on the recognition ability of Q-Azo4F-C toward the c-MYC Pu22 template. As depicted in Figure 4C–D and Supplementary Figures S20– S25, the presence of these different ions did not cause significant changes in the coordination mode of the photochrome to c-MYC Pu22. These findings further support the potential of Q-Azo4F-C for translation into live cell experiments. Visible Light Modulation of Cytotoxicity and G-Quadruplex Folding/Unfolding in Cancer Cells The almost complete two-way isomerization and the ability to alter Q-Azo4F-C’s binding mode on G4 structures using visible light led us to explore regulating cytotoxicity through the switch‘s state interconversion (Figures 5 and S26). For this purpose, four different human cancer cell lines originating from various tumors were utilized, including cervical cancer HeLa cells, osteosarcoma U2OS cells, breast cancer MCF-7 cells, and lung adenocarcinoma A549 cells. In its trans configuration, Q-Azo4F-C exhibited mild cytotoxic effects, displaying varying half-maximum inhibitory concentrations (IC50) across different cell lines: IC50(trans)(U2OS)= 41.1�10.0 μM, IC50(trans)(HeLa)=40.4�5.7 μM, IC50(trans) (MCF-7)>100 μM, and IC50(trans)(A549)>100 μM. Conversely, the cis-rich PSS, Q-Azo4F-C demonstrated heightened cytotoxicity with IC50(cis-rich PSS) (U2OS)=3.7�0.8 μM, IC50(cis-rich PSS) (HeLa)=2.6�0.2 μM, IC50(cis-rich PSS) (MCF-7)= 4.0�0.5 μM, and IC50(cis-rich PSS) (A549)=4.8�0.2 μM. These findings yielded high trans IC50/cis-rich PSS IC50 ratios ranging from 11 to >25, indicating the potential to modulate cytotoxicity through the application of visible light. To examine this hypothesis, we exposed the cells to cytotoxic levels of Q-Azo4F-C in its cis-rich PSS for 1 hour. Subsequently, the cells underwent a 10-minute exposure to blue light at 436 nm to induce the cis-to-trans isomerization. In this scenario, we observed a viability recovery of approximately 40% in U2OS, 50% in MCF-7, 70% in HeLa, and complete restoration (100%) in A549 cells (Figure 5). We then examined whether the light-dependent cytotoxicity displayed by the trans and cis isomers correlates with their ability to fold G4s inside cells (Figure 6). For this Angewandte Chemie Research Article Angew. Chem. Int. Ed. 2025,64, e202413000 (6 of 12) © 2024 The Authors. Angewandte Chemie International Edition published by Wiley-VCH GmbH 15213773, 2025, 1, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/anie.202413000 by Wroclaw University Of Science, Wiley Online Library on [30/01/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License purpose, we selected A549 cells as a model system, as the trans form of Q-Azo4F-C demonstrated minimal cytotoxicity, while the cis form exhibited substantial cytotoxicity, providing a wide therapeutic window. To visualize and quantitatively determine the G4 levels in A549 cells, we utilized the extensively characterized G4-specific antibody BG4.[2a,b] BG4 demonstrates a robust affinity for G4 structures across a spectrum of topologies and molecularities, with binding constants in the low nanomolar range.[2b] This strong binding affinity ensures minimal competition with Q-Azo4F-C for G4 structures, as the latter exhibits a significantly lower binding affinity, thereby facilitating its easy displacement from these structures. Furthermore, BG4’s ability to bind to both fully folded and partially folded G4 structures, coupled with its inability to interact with unfolded G4s, provides a reliable method to assess the effects of the isomeric forms of the switch on G4 structures within a cellular context.[2b] In these experiments, cells were exposed to different concentrations (6, 7, and 8 μM) of QAzo4F-C in either trans form or cis-rich mixture, after which immunofluorescence (IF) analyses were conducted. The chosen concentrations of Q-Azo4F-C in its cis-rich PSS were slightly higher than the previously calculated IC50 value of ~5 μM obtained in cell viability experiments. However, in the IF experiments, significantly different seeding densities were employed—approximately 100,000 cells per dish compared to 5,000 cells per well in cytotoxic studies, known to influence a drug‘s sensitivity.[22] Nevertheless, under the conditions of the IF experiment, the analyzed cells were confirmed to be viable. Compared to mock-treated cells, cells treated with Q-Azo4F-Ctrans, displayed a decreased number of BG4 foci (Figure 6B, C), suggesting an unfolding Figure 4. (A) Thermal stabilization analysis of Q-Azo4F-C trans and cis-rich PSS (cQ-Azo4F-C =8 μM) on Tel22-K+, Tel22-Na+, TBA, and dsDNA (cG4/ duplex =2 μM). Tris-HCl buffer=10 mM, KCl=5 mM for Tel22-K+, TBA, dsDNA, and NaCl=15 mM for Tel22-Na+. Results are presented as an average of three independent experiments, the error bars indicate SD. (B) 1H NMR spectra highlighting the G-tetrad imino protons of c-MYC Pu22 (black line, bottom) interacting with Q-Azo4F-C in trans form (red line) at different G4 to duplex DNA ratios. (C, D) 1H NMR spectra highlighting the G-tetrad imino protons of c-MYC Pu22 (cKCl =33 mM) interacting with 1 eq. of Q-Azo4F-C in its trans (C) or cis-rich PSS (D) forms, in the presence of 1 mM MgCl2. For comparison, the 1H NMR spectra of Q-Azo4F-C in either trans or cis-rich PSS forms bound to c-MYC Pu22 in the presence of 33 mM KCl but in the absence of MgCl2are also shown. The spectra, representing both the binding process and in situ photoswitching, indicate that MgCl2has no influence on the binding of the ligand to the G4 structure. Angewandte Chemie Research Article Angew. Chem. Int. Ed. 2025,64, e202413000 (7 of 12) © 2024 The Authors. Angewandte Chemie International Edition published by Wiley-VCH GmbH 15213773, 2025, 1, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/anie.202413000 by Wroclaw University Of Science, Wiley Online Library on [30/01/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License effect on G4 structures. Specifically, the trans isomer demonstrated a concentration-dependent impact on G4s, leading to reductions of approximately 0.7-fold, 0.6-fold, and 0.3-fold in the levels of nuclear BG4 foci at concentrations of 6, 7, and 8 μM, respectively. Conversely, the photoinduced cis-rich PSS of Q-Azo4F-C led to increased levels of nuclear BG4 foci, resulting in enhancements of about 2.1fold, 10.1-fold, and 5.6-fold at concentrations of 6, 7, and 8 μM, respectively. These findings suggest that the development of G4-binders should not focus exclusively on enhancing the binding affinity of compounds to G4 structures. Instead, it is crucial to prioritize the addition of substituents and structural modifications that enable diverse coordination modes. Although the trans isomer exhibits a higher affinity for G4 structures and was expected to aid in G4 folding, its atypical binding mechanism—likely involving partial intercalation into the G4 cavity—actually unfolds/ opens the G4 architecture, leading to reduced BG4 signal. On the other hand, the cis form, despite its lower affinity for G4 binding, maintains the integrity of G4 structures due to its non-disruptive coordination mode. This allows it to effectively fold G4 structures within cells without significantly altering the G4 framework. If, on one hand, G4 stabilization can serve as a strategic means to cause substantial damage to the genome of cancer cells, an overabundance of G4s resulting from impaired G4helicase function or the presence of G4-stabilizing ligands can contribute to various age-related disorders, such as certain neurodegenerative conditions (e.g., amyotrophic lateral sclerosis (ALS) and frontotemporal dementia Figure 5. Antiproliferative activity of Q-Azo4F-C on various cancer cell lines (A549, HeLa, MCF-7, and U2OS) using both trans and cis-rich PSS, followed by evaluation post-photoconversion into the trans-rich PSS. The in-cell cis-to-trans isomerization was achieved by incubating the cis-rich PSS for 1 hour followed by exposing the cells to 10 minutes of 436 nm blue light irradiation. The yellow bars illustrate the transformation of cytotoxic cis-rich PSS doses into less or even non-toxic trans-rich PSS concentrations. Error bars indicate mean�SD (n=3). Angewandte Chemie Research Article Angew. Chem. Int. Ed. 2025,64, e202413000 (8 of 12) © 2024 The Authors. Angewandte Chemie International Edition published by Wiley-VCH GmbH 15213773, 2025, 1, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/anie.202413000 by Wroclaw University Of Science, Wiley Online Library on [30/01/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License (FTD)), and specific genetic disorders associated with premature aging (e.g., Werner syndrome).[23] Hence, special caution must be exercised regarding the use of G4 ligands within the realm of tumor biology, as they could accumulate and potentially damage neurons.[24] Consequently, the regulation of G4-folding and unfolding using a singular molecular tool holds promise for interventions aimed at both inhibiting or enhancing DNA repair efficiency. To explore the potential of controlling G4 formation within cells using light, we conducted in-cell photomodulation experiments (Figure 7). As anticipated, using the cis-rich PSS of QAzo4F-C at a concentration of 7 μM resulted in an 8.8-fold increase in nuclear BG4 foci, while the trans form decreased the number of G4 foci by 0.5-fold. Interestingly, the photomodulation achieved by incubating the cis-rich PSS in the cell for 1 hour and exposing the cell solution to 10 minutes of blue light at 436 nm, inducing the cis-to-trans isomerization, almost completely restored the G4 levels to their native state (Figure 7A). It is important to note that even if the cis-to-trans isomerization did not cause the unfolding of Figure 6. G4 visualization in A549 cells utilizing the G4-specific antibody BG4. (A) A schematic illustration delineating the workflow for G4 immunodetection within cells. The cells were exposed to various concentrations of Q-Azo4F-C in either trans or cis-rich PSS forms prior to BG4 immunostaining and imaging. Image created with Biorender.com (B–D) Immunofluorescence staining of A549 cells treated with differing concentrations (6, 7, and 8 μM) of Q-Azo4F-C in trans or cis-rich PSS, or with equivalent amounts of DMSO. A549 cells were co-stained with the nuclear dye Hoechst 33342 (500 nM, blue). Red foci indicate BG4-mediated G4 recognition. λexc/λem: 405/420–460 nm for Hoechst (blue signal) and 598/620–750 for BG4 (red signal). The scale bar is set at 20 μm for regular images and 5 μm for the enlarged images. (E) Quantification of BG4 signal in the nucleus of A549 cells in the experimental conditions provided in B–D. Data represent populations of individual cells (for the 6 μM system: Nmock =201 cells, Ntrans =164 cells, Ncis-rich PSS =153 cells; for the 7 μM system: Nmock =279 cells, Ntrans =341 cells, Ncis-rich PSS =232 cells; for the 8 μM system: Nmock =277 cells, Ntrans =131 cells, Ncis-rich PSS =176 cells). The error bars represent mean�SD. Data analysis was performed using a two-sample t-test, and the significance level is denoted by the asterisks (****p<0.0001). Angewandte Chemie Research Article Angew. Chem. Int. Ed. 2025,64, e202413000 (9 of 12) © 2024 The Authors. Angewandte Chemie International Edition published by Wiley-VCH GmbH 15213773, 2025, 1, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/anie.202413000 by Wroclaw University Of Science, Wiley Online Library on [30/01/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License