Light-responsive and Protic Ruthenium Compounds Bearing Bathophenanthroline and Dihydroxybipyridine Ligands Achieve Nanomolar Toxicity towards Breast Cancer Cells

Abstract

We report new ruthenium complexes bearing the lipophilic bathophenanthroline (BPhen) ligand and dihydroxybipyridine (dhbp) ligands which differ in the placement of the OH groups ([(BPhen)₂ Ru(n,n'-dhbp)]Cl₂ with n = 6 and 4 in 1_A and 2_A , respectively). Full characterization data are reported for 1_A and 2_A and single crystal X-ray diffraction for 1_A . Both 1_A and 2_A are diprotic acids. We have studied 1_A , 1_B , 2_A , and 2_B (B = deprotonated forms) by UV-vis spectroscopy and 1 photodissociates, but 2 is light stable. Luminescence studies reveal that the basic forms have lower energy ³ MLCT states relative to the acidic forms. Complexes 1_A and 2_A produce singlet oxygen with quantum yields of 0.05 and 0.68, respectively, in acetonitrile. Complexes 1 and 2 are both photocytotoxic toward breast cancer cells, with complex 2 showing EC₅₀ light values as low as 0.50 μM with PI values as high as >200 vs. MCF7. Computational studies were used to predict the energies of the ³ MLCT and ³ MC states. An inaccessible ³ MC state for 2_B suggests a rationale for why photodissociation does not occur with the 4,4'-dhbp ligand. Low dark toxicity combined with an accessible ³ MLCT state for ¹ O₂ generation explains the excellent photocytotoxicity of 2.

Figure 1.

Complexes [(BPhen)₂Ru(n,n′-dhbp)]Cl₂ (1_A (n = 6) and 2_A (n = 4)) are studied herein. Past work has involved complexes 3_A-7_A. Gasser et al. have studied complex 8.

Figure 2.

Deprotonation of X_A = [(N,N)₂Ru(6,6’-dhbp)]²⁺ complexes under physiological conditions changes the properties. X = 1-5 (see Figure 1) and A denotes the acidic form bearing OH groups and B denotes the basic form bearing O⁻ groups.

Figure 3.

Synthetic scheme for target complexes 1_A and 2_A.

Figure 4.

Molecular diagrams for 1_A crystallized as [(BPhen)₂Ru(6,6′-dhbp)]Cl₂(PhCN)₂. In both views, benzonitrile solvent (which co-crystallized with 1_A) and chloride counter anions (which are hydrogen bonded to the OH groups) have been removed for clarity. Hydrogen atoms are hidden on the left. Ellipsoids are shown at 50% probability.

Figure 5.

Representative immunofluorescence imaging of MCF7 (top), MDA-MB-231 (middle), and MCF10A (bottom) after overnight incubation with (A) [(BPhen)₂Ru(4,4′-dhbp)]Cl₂ (2_A) and (B) [(BPhen)₂Ru(6,6′-dhbp)]Cl₂ (1_A). Nuclei, blue; mitochondria, green; Ru complex, magenta. Scale bar: 20 μm. These images are from samples treated with light, but dark samples had similar results.

Figure 6.

Free energy diagram for energetically accessible excited states of 1_A (in blue) and 2_A (in red) from SMD-PBE0-D3BJ/BS1//PBE0-D3BJ/BS1. For both, the ³MLCT state is thermally accessible and may lead to singlet oxygen generation if oxygen is present, but the ³MC state which leads to photodissociation is also thermally accessible. Experimentally, photodissociation is observed for 1_A and it is not observed for 2_A, which suggests the lower energy ³MC state for 1_A plays a role in facilitating photodissociation.

Figure 7.

Free energy diagram for energetically accessible excited states of 1_B (in blue) and 2_B (in red) from SMD-PBE0-D3BJ/BS1//PBE0-D3BJ/BS1. For 1_B, the ³MLCT state is lowest in energy and likely leads to singlet oxygen generation if oxygen is present, but the ³MC state which leads to photodissociation is also thermally accessible. For 2_B, the ³MC state is thermally inaccessible and thus ³MLCT is the dominant excited state which leads to singlet oxygen formation without photodissociation.