Ru(II) photocages enable precise control over enzyme activity with red light

Cytochrome P450s (CYPs) mediate xenobiotic metabolism and biosynthesis of signaling molecules; imbalances contribute to disease. Selective CYP inhibition has clinical utility (e.g., aromatase inhibitors for ER+ breast cancer), but off-target inhibition across the 57 human CYPs can cause toxicity and drug–drug interactions, complicated further by extensive polymorphisms. CYP1B1 is an extrahepatic enzyme overexpressed in many tumors; it generates mutagenic estrogen metabolites and is implicated in resistance to diverse chemotherapies (cisplatin, daunorubicin, taxanes). Achieving selective inhibition of CYP1B1, ideally localized to target tissues, is desirable. Photolabile prodrugs offer spatiotemporal control but organic photocages are problematic for CYP inhibitors, as common cages (e.g., coumarins) are CYP substrates and smaller organic cages are prone to metabolic cleavage. Prior attempts using large metal complexes retained significant inhibitory activity in the caged state, likely due to CYP active-site flexibility accommodating bulky groups. The study aims to create selective, light-activated CYP1B1 inhibitors using a coordination-based prodrug strategy and Ru(II) photocages activatable by red light.

Clinically used CYP inhibitors often employ Type II coordinating groups (e.g., anastrozole, letrozole, abiraterone, metyrapone, and azole antifungals for CYP51). For CYP1B1, coordinating inhibitors suitable for photocaging had not been reported. Organic photocages like coumarins, despite excellent photochemical properties, are metabolized by CYPs, undermining selectivity. Prior metal-based photocaging, including Ru(bpy)3-type scaffolds accommodating one or two inhibitors, showed residual activity of caged inhibitors and low photoactivity indices (often under 40) across targets such as cysteine proteases, NAMPT, tubulin polymerization, and CYPs; in some cases, intact Ru(II) complexes were more active than released inhibitors. Large Ru(II) complexes can bind CYPs due to active-site flexibility, but with unexpected selectivity. Shifting to Ru(II) scaffolds with longer wavelength activation (e.g., biquinoline ligands) improves tissue penetration and avoids UV-induced stilbene isomerization, and incorporation of carboxylates can improve biocompatibility.

Design and synthesis: A simple CYP1B1 inhibitor scaffold, tetramethoxystilbene (TMS, 1), was modified to introduce Lewis-basic metal-coordinating heterocycles distal to the heme-binding region: a diazine (compound 2) and a pyridyl (compound 3). The coordinating moiety anchors the inhibitor to the heme iron to enforce orientation; the diazine was expected to enhance Ru(II) photosubstitution yields while the pyridyl should enhance CYP1B1 potency due to higher basicity.

Photocage scaffolds: Two Ru(II) polypyridyl scaffolds were used. Scaffold II employed 2,2′-biquinoline (bq) functionalized as [2,2′-biquinoline]-4,4′-dicarboxylic acid (bicinchoninic acid, bca) to red-shift MLCT absorption and reduce cellular toxicity. A tridentate 2,2′;6′,2″-terpyridine (tpy) ligand occupied three sites, bca occupied two, and the monodentate inhibitor (2 or 3) occupied the remaining site, yielding octahedral complexes 4–6. A control complex [Ru(tpy)(bca)(pyridine)] (8) was prepared. Complexes varied in monodentate ligand (2 vs 3) and bidentate co-ligands (Scaffold I vs II).

Photophysics and photochemistry: UV/Vis spectra were recorded; MLCT maxima were between ~530–545 nm in water with tailing to 650–700 nm. Quantum yields for photosubstitution (ΦPS) were determined at 470 nm by optical methods and HPLC in aqueous and organic media. Stability was assessed in water at 37 °C over 24 h, including in the presence of glutathione, imidazole, and at low pH.

Biological assays: Cell-based CYP activity assays were developed in HEK293 T-Rex cells stably co-expressing inducible CYP1B1, CYP1A1, or CYP19A1 and P450 oxidoreductase (POR). Cells were dosed with compounds, irradiated for photoactivation when applicable, and enzyme activity measured using fluorogenic substrates: resorufin ethyl ether (REE) for CYP1B1/1A1 and dibenzylfluorescein (DBF) for CYP19A1. Red light irradiation was 660 nm, 58.7 J/cm² for 1 h. Dose-response IC50 values were calculated. Inhibition in pooled human liver microsomes (phLM) from 50 donors was measured using 7-benzoyloxy-4-trifluoromethylcoumarin; plates were either protected from light or irradiated (660 nm, 58.7 J/cm²) prior to initiating turnover with NADPH. Cellular viability was assessed by resazurin reduction following exposure (0–30 µM), in dark and after irradiation.

Mechanistic and biophysical studies: Singlet oxygen generation upon photoexcitation of complexes 4–8 was quantified to assess photodynamic contributions. Circular dichroism thermal melts provided CYP1B1 Tm shifts in the presence of inhibitors (ANF, 3, 6, and photoactivated 8). Structure-guided analysis included docking of inhibitor 3 into CYP1B1 (PDB 3PM0) and targeted mutagenesis (e.g., Ser127Ala, Phe134Leu, Gln332Glu, Asp333Asn, Ser269Ala) to probe key contacts, with IC50 shifts determined for inhibitors 1–3 and ANF in wild type vs mutants. HPLC characterized photoejection products (release of 3 and formation of complex 7).

• Coordinating CYP1B1 inhibitors: Introducing a metal-binding heterocycle produced highly potent CYP1B1 inhibitors. Mutational analysis identified key residues influencing potency: Ser127Ala, Gln332Glu, and Asp333Asn mutations decreased inhibitor efficacy by 2–3 orders of magnitude, implicating I-helix contacts (Q332 unique to CYP1B1) in selectivity. Phe134Leu had moderate impact; Ser269Ala minimal. Controls ANF and TMS (1) showed different sensitivity patterns consistent with distinct binding interactions.
• Ru(II) photocages and photophysics: Complexes 4–6 exhibited MLCT maxima ~530–545 nm in H2O with absorption tails to 650–700 nm, enabling red-light activation. ΦPS values at 470 nm varied: complexes bearing diazine ligands (4, 5) showed higher ΦPS (e.g., 4 ≈ 0.055; 5 ≈ 0.02), whereas complex 6 with a pyridyl ligand had low ΦPS in MeCN (~0.00043) but increased in less polar environments (up to ~0.008), indicating environmental sensitivity. Stability at 37 °C over 24 h: 4 (~43% remaining), 5 (~50%), 6 (~98.6%), supporting 6 as the preferred photocage due to thermal robustness.
• Photocontrol of CYP1B1 inhibition: Complexes 4 and 5 achieved photocontrol with photoactivity indices (PI; IC50 dark/IC50 light) of 16–102 upon 660 nm irradiation; dark IC50 values were 0.2–2 µM. Despite their size, intact Ru(II) complexes exhibited CYP1B1 inhibition but showed little to no inhibition of CYP19A1 or CYP1A1 at ≥10 µM. Control complex 8 (releasing pyridine) showed no CYP inhibition up to 30 µM, in dark or after irradiation to form 7.
• Lead system (complex 6): Incorporation of inhibitor 3 into Scaffold II yielded complex 6, activatable with low-energy red light (660 nm). Photolysis cleanly released inhibitor 3 and formed inert complex 7, confirmed by HPLC and UV/Vis. Post-irradiation IC50 for CYP1B1 inhibition was ~300 pM with PI > 6300, representing exceptional photocontrol (10–1000-fold higher PI than prior Ru(II) photocages). Selectivity indices over other CYPs were 4,000–100,000. In the dark, 6 inhibited CYP1B1 but had minimal effects on other CYPs up to 30 µM.
• Mechanism and safety: Minimal singlet oxygen generation by 4–8 suggests photodynamic inactivation is unlikely. CD thermal melts showed stabilization of CYP1B1 by ANF (+3 °C) and by 3 (+2 °C), while 6 slightly decreased Tm at 20 µM, consistent with distinct binding of the intact complex. The Ru(II) scaffold with carboxylated biquinoline reduced off-target interactions and cytotoxicity; model complex 8 showed no impact on cell health up to 30 µM.

The study demonstrates that precise, red-light-controlled inhibition of CYP1B1 is achievable using Ru(II) photocages combined with rationally designed coordinating inhibitors. Achieving high PI values has been a persistent challenge in photopharmacology, particularly for Ru(II) systems where dark IC50 values often remain in the micromolar range and intact complexes can be as active as released inhibitors. Here, by simultaneously optimizing the inhibitor (to sub-nanomolar/picomolar potency and high CYP1B1 selectivity via key I-helix contacts) and the Ru(II) scaffold (to ensure biocompatibility, red-shifted activation, and thermal stability), the authors overcame these limitations. The lead system (complex 6) achieved picomolar potency upon red-light activation with a PI > 6300 and outstanding selectivity over other CYPs, enabling spatiotemporal control with deep tissue-penetrant light. While intact Ru(II) complexes retained some dark inhibition of CYP1B1, they did not broadly inhibit other CYPs, indicating selective engagement potentially near the active-site channel rather than nonspecific protein interactions. These findings refine understanding of CYP–inhibitor interactions, particularly the role of coordinating groups and unique CYP1B1 residues (e.g., Gln332), and suggest broader applicability of Ru(II) photocages to other coordinating CYP inhibitors.

The authors developed a coordination-mediated prodrug strategy for CYP1B1 inhibition using biocompatible Ru(II) photocages activatable by 660 nm light. Rational redesign yielded extremely potent and selective coordinating inhibitors (compound 3), and coupling to an optimized Ru(II) scaffold (complex 6) provided exceptional photocontrol (PI > 6300) and picomolar potency upon activation, with 3–4 orders of magnitude selectivity over other CYPs. This represents, to the authors’ knowledge, the highest photoactivity index for a Ru(II) or organic photocaged enzyme inhibitor and the most selective CYP inhibitor reported. The approach is anticipated to generalize to other coordinating CYP inhibitors. Future work should focus on elucidating the binding mode of intact Ru(II) complexes to minimize dark activity, engineering fully inert scaffolds, expanding to additional CYP targets and in vivo models, and optimizing light delivery parameters for therapeutic contexts.

• Residual dark activity: Intact Ru(II) complexes (including 6) retain CYP1B1 inhibition in the absence of light, limiting PI and necessitating highly potent released inhibitors to achieve large on/off ratios.
• Mechanistic uncertainty: The precise binding site and mode of intact Ru(II) complexes remain unclear; docking of large complexes into CYP1B1 failed, hindering structure-guided reduction of dark binding.
• Stability vs photosubstitution trade-off: Complexes with higher ΦPS (4, 5) showed lower thermal stability over 24 h, while the most stable complex (6) had low ΦPS in polar media and environmental dependence, complicating optimization.
• Generalizability: While the strategy should apply to coordinating CYP inhibitors, off-target inhibition risk remains for CYPs sensitive to coordinating groups; careful selectivity profiling is required.
• Photochemical constraints: Although activation at 660 nm avoids UV-induced stilbene isomerization and improves tissue penetration, light dose requirements (58.7 J/cm²) and tissue optics may limit in vivo applications without further optimization.