Exploiting Co(III)-Cyclopentadienyl Complexes To Develop Anticancer Agents

João Franco Machado; Sandra Cordeiro; Joana N Duarte; Paulo J Costa; Paulo J Mendes; Maria Helena Garcia; Pedro V Baptista; Alexandra R Fernandes; Tânia S Morais

doi:10.1021/acs.inorgchem.3c03696

. 2024 Mar 19;63(13):5783–5804. doi: 10.1021/acs.inorgchem.3c03696

Exploiting Co(III)-Cyclopentadienyl Complexes To Develop Anticancer Agents

João Franco Machado ^†, Sandra Cordeiro ^‡,^§, Joana N Duarte ^†, Paulo J Costa ^∥, Paulo J Mendes ^⊥, Maria Helena Garcia ^†,^*, Pedro V Baptista ^‡,^§,^*, Alexandra R Fernandes ^‡,^§,^*, Tânia S Morais ^†,^*

PMCID: PMC10988555 PMID: 38502532

Abstract

graphic file with name ic3c03696_0018.jpg

In recent years, organometallic complexes have attracted much attention as anticancer therapeutics aiming at overcoming the limitations of platinum drugs that are currently marketed. Still, the development of half-sandwich organometallic cobalt complexes remains scarcely explored. Four new cobalt(III)-cyclopentadienyl complexes containing N,N-heteroaromatic bidentate, and phosphane ligands were synthesized and fully characterized by elemental analysis, spectroscopic techniques, and DFT methods. The cytotoxicity of all complexes was determined in vitro by the MTS assay in colorectal (HCT116), ovarian (A2780), and breast (MDA-MB-231 and MCF-7) human cancer cell lines and in a healthy human cell line (fibroblasts). The complexes showed high cytotoxicity in cancer cell lines, mostly due to ROS production, apoptosis, autophagy induction, and disruption of the mitochondrial membrane. Also, these complexes were shown to be nontoxic in vivo in an ex ovo chick embryo yolk sac membrane (YSM) assay.

Short abstract

First half-sandwich complexes of Co(III) bearing an unsubstituted cyclopentadienyl and containing a monodentate phosphane and bidentate N,N-heteroaromatic ligands as potential anticancer agents.

Introduction

Cancer remains a major burden in society worldwide. Despite the much progress achieved in oncology over the last decades, medical science has not succeeded in providing an efficient solution to cure cancer. The serendipitous discovery of the anticancer properties of cisplatin has fueled the development of several metal complexes for cancer therapy.¹⁻³ In this regard, the anticancer properties of ruthenium, osmium, rhodium, copper, gold, iridium, and titanium metal complexes have been explored.^2,4−12 Among them, ruthenium complexes have drawn great attention due to their broad spectrum of activity, low toxicity, and lower drug resistance than cisplatin.¹³⁻¹⁷ Particularly, half-sandwich organometallic Ru-arene complexes have preceded other anticancer organic and inorganic drugs due to their equilibrium between hydrophilicity and lipophilicity which could facilitate cellular uptake. Indeed, several examples of ruthenium(II)-(η⁶-C₆H₆) complexes showed promising in vitro and in vivo antitumor properties against various tumor types, even in cisplatin-resistant cell lines.^6,18−23 We have been focusing on developing new half-sandwich cationic complexes derived from the Ru(II)-(η⁵-C₅H₅) fragment as prospective anticancer agents. These Ru(II)-(η⁵-C₅H₅) derivatives, containing mono/bidentate heteroaromatic and phosphane ligands, have proven to be strong cytotoxic agents against a large range of cancer cell lines in vitro,^5,24−31 revealing in some cases an antimetastatic behavior in vivo.^32,33 Nonetheless, and despite these important advances in ruthenium complexes, these compounds still present some toxicity.

Alternative therapeutic approaches based on designing new complexes containing bioessential metals may lead to lower systemic toxicity than conventional metallodrugs and those complexes may reach the specific biological targets more easily. Cobalt is an essential trace element for humans, playing a vital role in several critical biological processes, being mainly found in the form of vitamin B12 (cobalamin).^34,35 Since humans have developed mechanisms to overcome cobalt overload, cobalt is thus less toxic than nonessential metals like platinum and ruthenium. In the past decade, a large number of cobalt complexes have been developed for therapeutic applications, demonstrating its therapeutic potential.³⁵⁻⁴¹ Despite the extensive work developed on cobalt complexes, and besides vitamin B12, there are so far no reports on the usage of cobalt compounds in general clinical use. However, a Co(III) imine complex (Doxovir) has already reached phase II clinical trials for antiviral treatment.^42,43 Several in vitro studies suggest that cobalt complexes may become promising anticancer agents.⁴⁴⁻⁴⁹ The great diversity of physicochemical properties of cobalt complexes, such as its accessible redox behavior and rich photochemistry, results in complexes with different cytotoxic modes of action.^35,44,46,48

The chemistry of cobalt complexes for therapeutic applications has been essentially dominated by classical coordination complexes. In contrast, the design and development of half-sandwich organometallic cobalt complexes remains scarcely studied and exploited in this context.³⁵ Our interest has focused on the exploitation of the “piano-stool” Co-cyclopentadienyl complexes as prospective anticancer agents. Notice that half-sandwich metal complexes with N,N-heteroaromatic ligands such as 2,2′-bipyridine or 1,10-phenantroline show promising anticancer and/or antimetastatic properties, both in vitro and in vivo, and are strong alternatives to current chemotherapeutic metallodrugs such as cisplatin.^5,32,50 Notably, the ruthenium complex [RuCp(PPh₃)(bipy)][CF₃SO₃] (TM34) has been shown to surpass the cytotoxicity of cisplatin against several cancer cell lines (breast, ovarian, prostate, leukemia) owing to a different mode of action that might overcome drug resistance to cisplatin by targeting the cell membrane, Golgi apparatus, and mitochondria instead of the nuclear DNA.^26,28 Encouraged by our previous results obtained with ruthenium compound TM34, we decided to explore an analogous family of Co(III)-cyclopentadienyl (CoCp) complexes with N,N-heteroaromatic bidentate ligands. Herein, we report the unprecedented synthesis and structural characterization of a new family of organometallic Co(III)(η⁵-C₅H₅) complexes of general formula [CoCp(PPh₃)(NN)][(CF₃SO₃)₂], where NN represents 2,2′-bipyridine, 4,4′-dimethyl-2,2′-bipyridine, 1,10′-phenanthroline, and 5-amino-1,10′-phenanthroline ligands. To the best of our knowledge, these are the first examples of half-sandwich complexes of Co(III) bearing an unsubstituted cyclopentadienyl and containing a monodentate phosphane and bidentate N,N-heteroaromatic ligands ever reported in the literature. After a complete characterization using experimental and computational techniques, their cytotoxicity was determined in vitro by the MTS assay in colorectal, ovarian, and breast human cancer cell lines, and in healthy fibroblasts. The cellular uptake was studied by ICP-AES. The determination of cell death mechanism and ROS production is presented and discussed. Also, studies using an ex ovo chick embryo yolk sac membrane (YSM) assay^51,52 were performed to evaluate the potential of these complexes to modulate the angiogenic process and their toxicity in vivo as angiogenesis is part of the invasion/metastization process in cancer but is also important for tumor growth and maintenance as it allows tumor cells to receive nutrients (glucose, oxygen, etc.).⁵²

Experimental Methods

Materials and Methods

All chemicals and solvents were of analytical reagent grade and used without further purification except dichloromethane and n-hexane, which were purified immediately before use with an MBraun SPS-800 solvent purification system. All manipulations involving air-free syntheses and purifications were carried out under a dinitrogen atmosphere using Schlenk techniques. The starting material [CoCp(CO)I₂] was synthesized according to previously reported procedures.⁵³ NMR spectra were recorded in acetone-d₆ at probe temperature using a Bruker Avance 400 spectrometer on 400.13 MHz (¹H NMR), 100.62 MHz (APT-¹³C{¹H} NMR) or 161.97 (³¹P{¹H} NMR). Chemical shifts (δ) are reported in parts per million (ppm), downfield from solvent peaks considering internal Me₄Si (0.00 ppm) in the ¹H and ¹³C spectra or referred from external 85% H₃PO₄ in the ³¹P spectra. All NMR resonances were unambiguously assigned using bidimensional complementary experiments (COSY, HSQC, and HMBC). Abbreviations: s = singlet; d = doublet; t = triplet; m = multiplet; J = coupling constant. FT-IR spectra (4000–400 cm^–1) were recorded in KBr pellets at room temperature, using a Shimadzu IRAffinity-1 spectrophotometer, only the most significant bands are cited in the text. Electronic spectra (233–900 nm) were recorded in dichloromethane (1 × 10^–5–1 × 10^–3 M) at room temperature, using a Jasco V-560 spectrometer and quartz cuvettes with a 1 cm optical path. Elemental analyses were performed at Laboratório de Análises at Instituto Superior Técnico, using a Fisons Instruments EA1108 system. Data acquisition, integration, and handling were made resorting to the software package EAGER-200 (Carlo Erba Instruments).

Syntheses of the complexes

The new cobalt(III) complexes of general formula [Co(η⁵-C₅H₅)(PPh₃)(NN)][(CF₃SO₃)₂], where NN represents the N,N-bidentate ligands 2,2′-bipy (1), phen (2), Me₂bipy (3), or NH₂phen (4), were synthesized following the general procedure: to a stirring solution of [CoCp(CO)I₂] (0.20g, 0.5 mmol) in acetonitrile (10 mL) was added dropwise a solution of the desired ligand (0.5 mmol) and AgCF₃SO₃ (0.26 g, 1.0 mmol) in acetonitrile (10 mL). The mixture was stirred at 0 °C for 10 min protected from light, turning from black to purple-red. The formation of an AgI precipitate was also observed. The solution was then cannula-filtrated into a stirring solution of triphenylphosphane (0.13 g, 0.5 mmol) in acetonitrile (10 mL). The resulting mixture was stirred at 0 °C for 5 min, turning to orange-red. The solvent was evaporated under vacuum, and the residue was washed with diethyl ether (2 × 10 mL) and cold dichloromethane (1 × 5 mL). The products were recrystallized by slow diffusion of dichloromethane/n-hexane or methanol/diethyl ether, affording crystalline products.

Data for [Co(η⁵-C₅H₅)(PPh₃)(bipy)][(CF₃SO₃)₂] (1)

Recrystallized from dichloromethane/n-hexane; orange crystals; yield: 56%.

¹H NMR [(CD₃)₂CO, Me₄Si, 400.13 MHz] δ/ppm: 9.72 [d, 2, H₁, ³J_HH = 5.00 Hz]; 8.38 [d, 2, H₄, ³J_HH = 7.62 Hz]; 8.29 [t, 2, H₃, ³J_HH = 7.29 Hz]; 7.80 [m, 2, H₂]; 7.67 [m, 3, H_para(PPh₃)]; 7.54 [m, 6, H_meta(PPh₃)]; 7.29 [t, 6, H_ortho(PPh₃), ³J_HH = 9.07 Hz]; 6.62 [s, 5, H_Cp]. APT-¹³C{¹H} NMR [(CD₃)₂CO, Me₄Si, 100.62 MHz] δ/ppm: 159.32 [C₁]; 158.36 [C₅]; 142.27 [C₃]; 134.27 [d, C_ortho(PPh₃), ²J_PC = 9.60 Hz]; 133.63 [d, C_para(PPh₃), ⁴J_PC = 2.99 Hz]; 130.82 [d, C_meta(PPh₃), ³J_PC = 10.89 Hz]; 129.48 [C₂]; 126.38 [d, C_ipso(PPh₃), ¹J_PC = 49.42 Hz]; 126.32 [C₄]; 94.30 [C_Cp]. ³¹P{¹H} NMR [(CD₃)₂CO, 161.97 MHz] δ/ppm: 41.16 [s, PPh₃]. FT-IR [KBr, cm^–1]: 3105–3066 (ν_C–H, aromatic rings); 1608 (ν_C=N, bipy); 1477–1433 (ν_C=C, aromatic rings); 1259 (νCF₃SO₃); 875–696 (δ_C–H, aromatic rings). Elemental analysis (%) found: C, 48.1; H, 3.2; N, 3.2; S, 8.0. Calculated for C₃₅H₂₈CoF₆N₂O₆PS₂·0.5CH₂Cl₂ (883.10 g/mol): C, 48.28; H 3.31; N, 3.17; S, 7.26.

Data for [Co(η⁵-C₅H₅)(PPh₃)(Me₂bipy)][(CF₃SO₃)₂] (2)

Recrystallized from methanol/diethyl ether; red crystals; yield: 74%.

¹H NMR [(CD₃)₂CO, Me₄Si, 400.13 MHz] δ/ppm: 9.50 [d, 2, H₁, ³J_HH = 4.38 Hz]; 8.21 [s, 2, H₄]; 7.68 [m, 3, H_para(PPh₃)]; 7.63 [m, 2, H₂]; 7.54 [m, 6, H_meta(PPh₃)]; 7.28 [t, 6, H_ortho(PPh₃), ³J_HH = 7.28 Hz]; 6.56 [s, 5, H_Cp]; 2.56 [s, 6, H₆]. APT-¹³C{¹H} NMR [(CD₃)₂CO, Me₄Si, 100.62 MHz] δ/ppm: 158.22 [C₁]; 157.84 [C₅]; 155.31 [C₃]; 134.35 [d, C_ortho(PPh₃), ²J_PC = 9.57 Hz]; 133.53 [d, C_para(PPh₃), ⁴J_PC = 2.90 Hz]; 130.77 [d, C_meta(PPh₃), ³J_PC = 10.80 Hz]; 130.37 [C₂]; 127.07 [C₄]; 126.62 [d, C_ipso(PPh₃), ¹J_PC = 48.73 Hz]; 93.92 [C_Cp]; 21.01 [C₆]. ³¹P{¹H} NMR [(CD₃)₂CO, 161.97 MHz] δ/ppm: 41.05 [s, PPh₃]. FT-IR [KBr, cm^–1]: 3120–3082 (ν_C–H, aromatic rings and methyl groups); 1624 (ν_C=N, Me₂bipy); 1487–1436 (ν_C=C, aromatic rings and δ_CH3, Me₂bipy); 1257 (νCF₃SO₃); 875–696 (δ_C–H, aromatic rings). Elemental analysis (%) found: C, 50.7; H, 3.7; N, 3.2; S, 7.5. Calculated for C₃₇H₃₂CoF₆N₂O₆PS₂ (868.69 g/mol): C, 51.16; H, 3.71; N, 3.22; S, 7.38.

Data for [Co(η⁵-C₅H₅)(PPh₃)(phen)][(CF₃SO₃)₂] (3)

Recrystallized from methanol/diethyl ether; red crystals; yield: 78%.

¹H NMR [(CD₃)₂CO, Me₄Si, 400.13 MHz] δ/ppm: 10.14 [d, 2, H₁ + H₁₀, ³J_HH = 4.07 Hz]; 8.86 [d, 2, H₃ + H₈, ³J_HH = 7.68 Hz]; 8.14 [m, 4, H₂ + H₅ + H₆ + H₉]; 7.54 [m, 3, H_para(PPh₃)]; 7.39 [m, 6, H_meta(PPh₃)]; 7.18 [m, 6, H_ortho(PPh₃)]; 6.72 [s, 5, H_Cp]. APT-¹³C{¹H} NMR [(CD₃)₂CO, Me₄Si, 100.62 MHz] δ/ppm: 160.41 [C₁ + C₁₀]; 148.92 [C₁₁ + C₁₂]; 140.62 [C₃ + C₈]; 133.95 [d, C_ortho(PPh₃), ²J_PC = 9.72 Hz]; 133.45 [d, C_para(PPh₃), ⁴J_PC = 2.90 Hz]; 132.38 [C₄ + C₇]; 130.41 [d, C_meta(PPh₃), ³J_PC = 10.92 Hz]; 128.81 [C₅ + C₆]; 128.47 [C₂ + C₉]; 125.99 [d, C_ipso(PPh₃), ¹J_PC = 49.05 Hz]; 94.10 [C_Cp]. ³¹P{¹H} NMR [(CD₃)₂CO, 161.97 MHz] δ/ppm: 41.25 [s, PPh₃]. FT-IR [KBr, cm^–1]: 3082–3076 (ν_C–H, aromatic rings); 1604 (ν_C=N, phen); 1583–1433 (ν_C=C, aromatic rings); 1261(νCF₃SO₃); 873–696 (δ_C–H, aromatic rings). Elemental analysis (%) found: C, 49.9; H, 3.3; N, 3.3; S, 7.0. Calculated for C₃₇H₂₈CoF₆N₂O₆PS₂·0.4CH₂Cl₂ (898.63 g/mol): C, 49.99; H, 3.23; N, 3.12; S 7.14.

Data for [CoCp(PPh₃)(NH₂phen)][(CF₃SO₃)₂] (4)

Recrystallized from methanol/diethyl ether; red crystals; yield: 54%.

¹H NMR [(CD₃)₂CO, Me₄Si, 400.13 MHz] δ/ppm: 10.08 [d, 1, H₁]; 9.58 [d, 1, H₁₀]; 8.99 [d, 1, H₃, ³J_HH = 8.33 Hz]; 8.35 [d, 1, H₈, ³J_HH = 8.23 Hz]; 8.07 [m, 1, H₂]; 7.80 [m, 1, H₉]; 7.57 [m, 3, H_para(PPh₃)]; 7.40 [m, 6, H_meta(PPh₃)]; 7.18 [m, 6, H_ortho(PPh₃)]; 7.02 [s, 1, H₆]; 6.68 [s, 5, H_Cp]; 6.50 [s, 2, H₁₃]. APT-¹³C{¹H} NMR [(CD₃)₂CO, Me₄Si, 100.62 MHz] δ/ppm: 159.83 [C₁]; 155.09 [C₁₀]; 149.43 [C₁₂]; 145.83 [C₅]; 143.05 [C₁₁]; 136.63 [C₈]; 135.53 [C₃]; 134.43 [C₇]; 133.83 [d, C_ortho(PPh₃), ²J_PC = 9.53 Hz]; 133.27 [d, C_para(PPh₃), ⁴J_PC = 2.74 Hz]; 130.24 [d, C_meta(PPh₃), ³J_PC = 10.84 Hz]; 127.76 [C₉]; 126.83 [C₂]; 125.95 [d, C_ipso(PPh₃), ¹J_PC = 49.29 Hz]; 125.10 [C₄]; 102.88 [C₆]; 93.86 [C_Cp]. ³¹P{¹H} NMR [(CD₃)₂CO, 161.97 MHz] δ/ppm: 41.61 [s, PPh₃]. FT-IR [KBr, cm^–1]: 3250 (ν_N–H, NH₂phen); 3084–3064 (ν_C–H, aromatic rings); 1645–1598 (ν_C=N and δ_N–H, NH₂phen); 1519–1435 (ν_C=C, aromatic rings); 1261 (νCF₃SO₃); 867–696 (δ_C–H, aromatic rings). Elemental analysis (%) found: C, 50.6; H, 3.2; N, 4.8; S, 7.1. Calculated for C₃₇H₂₉CoF₆N₃O₆PS₂ (879.67 g/mol): C, 50.52; H, 3.32; N, 4.78; S, 7.29. RP-HPLC: t_R = 29.02 min (A: H2O, B: ACN; method 0–1 min: 10%B; 1–18 min: 10–90% B; 18–21 min: 90% B; 21–24 min: 90–10% B; 24–25 min: 10% B); purity: 95,94%.

Stability Assays by NMR

The stability of the CoCp complexes in aqueous medium was evaluated by ¹H and ³¹P{H} NMR, using a Bruker Avance 400 spectrometer on 400.13 and 161.97 MHz, respectively. Solutions of the complexes in 90% D₂O/10% DMSO-d₆ were prepared at 2.5 mM and analyzed over 24 h. The spectra were acquired with intervals of 1 h within the first 6 h, followed by a final acquisition by completing 24 h. The samples were kept at room temperature and protected from light in between measurements. Upon competition of the assay, the spectra were analyzed regarding the number, chemical shift, integration, and multiplicity of each ¹H and ³¹P resonances for each complex.

Octanol–Water Partition Coefficients (logP)

The lipophilicity of the CoCp complexes was estimated by the shake-flask method.⁵⁴ Before the assays, n-octanol was vigorously mixed with distilled water for 24 h at room temperature to promote solvent saturation of both phases. After separating the phases, the complexes were dissolved in the n-octanol phase to obtain solutions 1.0 × 10^–5 to 2.5 × 10^–4 M. Then, each solution was equilibrated with water for 4 h in a mechanical shaker, at a phase ratio of 2 mL n-octanol:2 mL water. Afterward, the aqueous and octanol layers were carefully separated by centrifugation (5000 rpm, 10 min), and the UV–vis absorption spectra of the complexes were recorded in the n-octanol phase. For each sample, the concentration was determined by using a calibration curve in n-octanol. All experiments were performed in triplicate for each complex, being the averages used for calculation purposes. Finally, the values of the partition coefficients were calculated by resorting to the following eq (eq 1):

where logP_oct/water represents the octanol–water partition coefficient, [complex]_oct represents the concentration of the complex in the n-octanol phase, and [complex]_water represents the concentration of the complex in the water phase.

DFT and TD-DFT Calculations

All DFT and TD-DFT calculations were performed with Gaussian16⁵⁵ using the PBE0 functional.⁵⁶ Cobalt and phosphorus were represented, respectively, by the LANL2TZ(f) and LANL08(d) basis sets, along with the associated effective core potential.^57,58 For the other elements, the 6-311G** basis set was employed. The geometry was optimized without symmetry constraints. Using the optimized geometry, TD-DFT calculations were performed, typically requesting the lowest allowed 20 excitations. Solvent effects (dichloromethane) were included by employing the SMD solvation model.⁵⁹ The analysis of the UV–vis spectra and generation of natural transition orbitals cube files was performed with Multiwfn 3.8.⁵⁹

Cell Culture

Human colorectal carcinoma cell line (HCT116), human breast cancer cell lines (MDA-MB-231 and MCF-7), and normal human primary dermal fibroblasts neonatal (PCS-201-010) cell line were grown in DMEM (Dulbecco’s modified Eagle’s medium) (Invitrogen Corp., Grand Island, NY, USA) supplemented with 10% (v/v) fetal bovine serum and 1% (v/v) antibiotic/antimycotic solution (Invitrogen Corp.). The human ovarian carcinoma (A2780) cell line was cultivated using RPMI (Roswell Park Memorial Institute) medium supplemented as in DMEM medium. The supplemented medium is named a complete medium. Cells were grown in an incubator with a humidified atmosphere at 5% (v/v) CO₂ and 37 °C. All cell lines were purchased from ATCC (ATCC, American Type Culture Collection).

Antiproliferative Activity

The evaluation of the cytotoxic potential of complexes 1–4 was obtained through in vitro measurements of the antiproliferative activity of these complexes. Cells were plated in 96-well plates at 7.5 × 10⁴ cells/mL and incubated at 37 °C with 5% (v/v) CO₂. After 24 h, the culture medium was replaced with fresh medium (100 μL) containing 0.1–50 μM of the different complexes and incubated for 48 h in tumor cell lines or containing 0.1–100 μM of the different complexes and incubated for 48 h in fibroblasts. DMSO 0.1% (v/v) was used as the negative control and cisplatin and doxorubicin (DOX) were used as positive controls (common chemotherapeutic agents) for all cell lines. The fibroblast cell line was used as a control of the cytotoxicity of the complexes for healthy cells. Cell viability was evaluated using Cell Titer 96 Aqueous Non-Radioactive Cell Proliferation Assay (Promega, Madison, WI, USA) using 3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium inner salt (MTS), as described previously.⁶⁰ In metabolically active cells, enzymes present in the mitochondria catalyze a reaction in which NADPH/NADH is produced. These enzymes reduce the MTS reagent into a brownish product, Formazan, which can be quantified by measuring the absorbance at 490 nm. The quantity of formazan produced is directly proportional to the number of viable cells in the culture. This quantification was executed with a Biorad microplate reader Model 680 (Bio-Rad, Hercules, CA, USA). Half maximal inhibitory concentration (IC₅₀) was calculated using GraphPad Prism 6 (GraphPadSoftware). We have calculated the ratio between the IC₅₀ of fibroblasts and the IC₅₀ of HCT116, A2780, MDA-MB-231, or MCF-7 cell lines for complexes 1–4 that was named herein as the Selectivity index (SI). To ensure the solubility of the complexes in all biological assays, complex solutions were freshly prepared and well solubilized before addition to the cells.

Inductively Coupled Plasma Atomic Emission Spectroscopy (ICP-AES)

Internalization of the cobalt complexes was evaluated by ICP-AES in the A2780 cell line. For this, the A2780 cell line was seeded in 25 cm² T-flasks at a density of 5 × 10⁵ cells/T-flask and incubated for 24 h at 37 °C with 5% (v/v) CO₂. After the incubation, the culture media was replaced by fresh medium with 10× the IC₅₀ of the different complexes or 0.1% (v/v) of DMSO. As the IC₅₀ values in A2780 cells are in the low micromolar range and to ensure ICP-AES sensitivity in terms of limit of detection, the 10× IC₅₀ concentrations were used. Cells were then incubated for 3, 6, and 12 h at the same conditions as before and for 12 h at 4 °C. The culture media was recovered to a new tube (supernatant fraction) and cells were detached from the T-flask with 2 mL of TripLE Express (TE) (Gibco by life technologies) and centrifuged at 750 × g for 5 min. Then, fresh aqua regia was added to the supernatant tube and to the cellular fractions and incubated overnight at room temperature (RT) in the hood fume. Samples were delivered to Laboratório de Análises/LAQV and the levels of cobalt were evaluated by ICP-AES.

Cell Fractioning

Internalization of the complexes in different cellular factions (cytosolic, mitochondrial, or nuclear) of A2780 cells was evaluated by using a cell fractionation kit (Abcam). Cells were seeded in a 24-well plate at a density of 2 × 10⁵ cells/mL and incubated for 24 h at 37 °C with 5% (v/v) CO₂. After the incubation, the culture media was replaced by fresh medium with 10× the IC₅₀ of the different complexes or 0.1% (v/v) of DMSO, as for ICP-AES. Cells were then incubated for 12 h at the same conditions as before. The culture media was recovered to a new tube (supernatant fraction) and cells were detached from the T-flask with 250 μL of TE (Gibco by life technologies). Then, we followed the instructions provided by the manufacturer of the cell fractionation kit (Abcam). At the end of the procedure, fresh aqua regia was added to the supernatant tube and to the cellular fractions and incubated overnight at room temperature (RT) in the hood fume. Samples were delivered to Laboratório de Análises/LAQV and the levels of cobalt were evaluated by ICP-AES.

UV–Visible Spectrum in Biological Medium

UV–visible spectra (220–700 nm) were used to analyze the stability and solubility of complexes 1–4. Complexes were solubilized in DMSO and later diluted in RPMI medium without phenol red, at final concentrations of 25/50 μM. The spectra were obtained after 0, 24, and 48 h of incubation at 37 °C on the spectrophotometer Shimadzu mini-UV-1240 (Izasa Scientific) with a quartz cuvette with a 1 cm path length. Additionally, a Glutathione (GSH) interaction assay was also performed through the UV–vis spectra (220–700 nm) on the spectrophotometer Evolution 300 UV–vis (Thermo Fischer Scientific, Waltham, MA, USA) with a quartz cuvette with 1 cm path length. As above, RPMI medium without phenol red and FBS was used to dilute the complexes 2–4 at a final concentration of 50 μM with/without 50 μM of GSH. A control solution of GSH was also prepared. All solutions were incubated for 24 h at 37 °C and then analyzed.

Evaluation of Induction of Apoptosis

The induction of apoptosis was evaluated by flow cytometry in the A2780 cell line, through the Alexa Fluor Annexin V/Dead Cell Apoptosis Kit (ThermoFisher Scientific). A2780 cells were seeded at a density of 2 × 10⁵ cells/well in 6-well plates and incubated for 24 h. After that time, the medium was replaced by medium mixed with complexes 2–4 at their IC₅₀. DMSO at 0.1% (v/v) was used as the negative control and 0.4 μM DOX and 5 μM cisplatin were used as positive controls. After a period of 48 h of incubation, cells were washed with PBS 1× and TE was used to detach the cells. The cells were resuspended in annexin-binding buffer 1× and incubated for 15 min at RT with Alexa Fluor 488 annexin V and Propidium Iodide. All samples were analyzed by an Attune Acoustic Focusing Flow Cytometer (ThermoFisher Scientific) and respective software.

Determination of BAX and BCL-2 Protein Expression by Western Blot (WB)

The quantification of BAX and BCL-2 protein expression in the A2780 cell line incubated with different cobalt complexes in the A2780 cell line was determined by WB. For this, the A2780 cell line was cultivated in 25 cm² T-flasks at a density of 2 × 10⁶ cells/mL and incubated for 24 h in the conditions described above. Cells were incubated for 24 h, and, after that period, the medium was replaced by fresh medium with DMSO 0.1% (v/v) or the IC₅₀ c of the complexes 2–4. After 48 h incubation, the cells were washed and collected using cold PBS 1× and a cell scraper. The samples were centrifuged for 5 min at 700 × g and resuspended in fresh lysis buffer (150 mM NaCl, 5 mM ethylenediaminetetraacetic acid (EDTA), 50 mM Tris–HCl at pH 8.0, 2% (v/v) NP-40, 1× phosphatase inhibitors (PhosStop, Roche), 1× Protease inhibitors (complete ULTRA tablets, Mini, easypack, Roche), 1 mM phenylmethylsulfonylfuoride (PMSF) and 0.1% 1,4-dithiothreitol (v/v), DTT) and then storage at −80 °C for 2 h or until required. Cells were submitted to 5 ultrasound cycles and centrifuged at 1000 × g for 5 min. The total protein extract (supernatant) was quantified with Pierce Protein Assay Reagent (ThermoFisher Scientific) at 660 nm.

The SDS-PAGE was loaded with 20 ug of protein on 10% polyacrylamide gel and transferred to a 45 μm PVDF membrane (GE Healthcare). A NZYColour Protein Marker II (NZYTech) was used as a molecular weight marker. Each membrane was incubated for 2 h under constant agitation with 5% (w/v) nonfat milk in TBST 1× buffer (50 mM Tris–HCl at pH 7.5, 150 mM NaCl and 0.1% (w/v) Tween 20) and later incubated for 1 h, at RT under constant agitation, with the primary antibodies diluted in 5% nonfat milk in TBST 1×, anti-BAX (1:5000, reference 32503, Abcam) and anti-BCL-2 (1:100; reference B3170, Sigma). After that period, the membranes were washed three times with TBST 1× for 5 min each, under constant agitation, and then incubated with the secondary antibodies, antimouse IgG, horseradish peroxidase HPR-linked antibody (1:3000) and antirabbit IgG, HPR-linked antibody (1:2000) (both from Cell Signaling Technology, USA), for identification of BCL-2 and BAX protein, respectively.

The membranes were incubated with WesternBright ECL subtract (Advansta) for 5 min in the dark and the film was exposed to the membranes in a dark room. Membranes were later incubated two times with stripping buffer (0.1 M glycine, 20 mM magnesium acetate, 50 mM KCl at pH 2.0) during 10 and 20 min, respectively, with agitation and then incubated with anti-β-actin (1:5000; reference A5441, Sigma) as a control to normalize the expression results of the proteins. Protein quantification was made by densitometry with ImageJ software.

Evaluation of the Mitochondrial Membrane Potential (ΔΨM)

ΔΨM was evaluated in the A2780 cell line using the JC-1 Mitochondrial Membrane Potential Assay Kit (Abnova, Taipei, Taiwan). Cells were seeded at a density of 2 × 10⁵ cells/well in 6-well plates and incubated for 24 h. Posteriorly, the medium was replaced by medium with DMSO 0.1% (v/v), vehicle control, 0.4 μM DOX and 5 μM cisplatin, positive controls, or IC₅₀ of the complexes 2–4. After an incubation of 48 h, the cells were washed with PBS 1×, detached with TE, and incubated with JC-1 in RPMI medium without phenol red + 5% (v/v) FBS for 20 min at 37 °C. The cells were then resuspended in the same medium and the ΔΨM was evaluated with the Attune acoustic focusing flow cytometer (ThermoFisher Scientific) and respective software.

Evaluation of the Levels of Caspase 8 Activity

The A2780 cell line was seeded in 25 cm² T-flasks at a density of 2 × 10⁶ cells/T-flasks. After 24 h incubation, at the same conditions as described previously, the medium was replaced by medium with DMSO 0.1% (v/v), 0.4 μM DOX, a 5 μM cisplatin or IC₅₀ of the complexes 2–4, and incubated for 48 h. After that, cells were detached using cold PBS 1× and a cell scraper and centrifuged at 500 × g for 5 min. Then, we followed the instructions provided by the manufacturer of the Caspase-8 assay kit (Abcam). The data was obtained by measuring the absorbance of each sample at 400 nm and each value was normalized to the value of the DMSO sample.

Evaluation of Autophagy Induction

The autophagy induced by complexes 2–4 was evaluated using the Autophagy Assay Kit (ab139484, Abcam, Cambridge, United Kingdom) according to the manufacturer’s instructions. For this, the A2780 cell line was seeded at a density of 2 × 10⁵ cells/well in 6-well plates and incubated for 24 h at the conditions mentioned before. Posteriorly, the medium was replaced by medium with DMSO 0.1% (v/v) (vehicle control), 0.4 μM DOX and 5 μM cisplatin, positive controls, and the IC₅₀ of the complexes 2–4. Cisplatin can induce autophagy through the activation of the BECN1 and DOX high cytotoxicity, which also activates this process.⁶¹⁻⁶³ The positive control Rapamycin at 500 nM was incubated for only 15 h. After 48 h incubation, the cells were washed with PBS 1×, detached with TE, and treated with the Autophagy detection kit (Abcam, Cambridge, UK) according to the manufacturer’s instructions. The evaluation of autophagy induction was done through the Attune acoustic focusing flow cytometer (ThermoFisher Scientific) and respective software.

Production of Reactive Oxygen Species (ROS)

The production of ROS was evaluated in the A2780 cell line, seeded in 6-well plates at a density of 2 × 10⁵ cells/well, and incubated for 24 h at the same conditions as described above. After that, the medium was replaced by medium with DMSO 0.1% (v/v), vehicle control, 0.4 μM DOX, 5 μM cisplatin, and 22.2 μM TBHP (as positive controls) or IC₅₀ of the complexes 2–4 and incubated for 48 h. Later, cells were washed with PBS 1×, detached from the wells with TE, and incubated with 10 mM of 2′,7′-dichlorodihydrofluorescein diacetate (H₂DCF-DA) (ThermoFisher Scientific, Waltham, MA, USA) in PBS 1× for 20 min at 37 °C. The data were obtained through an Attune acoustic focusing flow cytometer (ThermoFisher Scientific) and respective software.

Cell Cycle Progression

The interference of the copper complexes 2–4 in the cell cycle progression was evaluated in A2780 cells, seeded in 6-weel plates, with a density of 2 × 10⁵ cells/well, and incubated for 8 h at the same conditions as described above. After that incubation time, the cells were synchronized in early S-phase with a 2 mM thymidine solution (Sigma, St. Louis, USA) for a 16 h period. Media with thymidine was replaced by fresh media and, after 8 h, another 2 mM thymidine solution was added to the wells (double thymidine block) and incubated for 16 h. The cell medium was replaced with IC₅₀ concentrations of complexes 2–4, DMSO 0.1% (v/v), vehicle control, 0.5 μM DOX, or 5 μM cisplatin (positive controls). Cells were incubated for 9, 18, or 24 h, and, at each hour, the cells were treated with TE and centrifuges at 650 × g for 5 min at 4 °C. The pellet was resuspended in PBS 1× and centrifuged at 3000 × g for 5 min at 4 °C. Later, the pellet was resuspended in 100 μL of PBS 1×, and 1 mL of a solution of 80% (v/v) ethanol was gently added to the cells. Cells were stored at 4 °C for at least 12 h and then centrifuged for 10 min at 7500 × g at 4 °C, treated with 250 μL of 50 μg/mL RNase A, and incubated for 30 min at 37 °C. After incubation, 100 μL of PI (25 μg/mL) was added to the cells as well as 650 μL of PBS 1×. The DNA content was evaluated on an Attune Acoustic Focusing Flow Cytometer (ThermoFisher Scientific) and results were analyzed by the respective software.

Escherichia coli Culture and Plasmid DNA Extraction

E. coli transformed with pUC18 was inoculated onto a plate of LB-agar medium (Luria–Bertani) (Applichem, Darmstadt, Germany) supplemented with ampicillin (100 μg/mL) (Bioline, London, UK) and incubated for 24 h at 37 °C. Then, the E. coli cells were inoculated in liquid LB medium, supplemented with ampicillin (100 μg/mL) for 24 h at 37 °C. After incubation, the plasmid DNA (pDNA) was extracted by using the Kit NZYSpeedy Miniprep, following the instructions provided by the manufacturer.

Evaluation of the Interaction with pDNA

The extracted pDNA (100 ng) was incubated with 5, 25, 50, and 100 μM concentrations of complexes 2–4, with DMSO 0.1% or only pUC18. The solutions were prepared in 5 mM Tris–HCl, 50 mM NaCl, pH 7.02, buffer, and incubated for 24 h at 37 °C. A control of pDNA cleaved with Hind III endonuclease was also used. Electrophoresis was performed in 0.8% (w/v) agarose gel (NZYtech) in TAE buffer at 70 V constant voltage for 1 h. Electrophoresis gel image was acquired in the Gel Doc EZ Imager (Bio-Rad).

Determination of the DNA Cleavage Mechanism

The pDNA (100 ng) was incubated with 25 μM of complexes 2–4, in the absence or presence of 50 μM of NaN₃, 4 units of catalase from the bovine liver (2000–5000 U/mg, Sigma-Aldrich, Spain), and 25 mM of D₂O. The solutions were prepared in 5 mM Tris–HCl, 50 mM NaCl, pH 7.02, buffer, and incubated for 24 h at 37 °C. A control of pDNA cleaved with Hind III endonuclease was also used. Electrophoresis was performed in 0.8% (w/v) agarose gel (NZYtech) in TAE buffer at 70 V constant voltage for 1 h. Electrophoresis gel image was acquired in Gel Doc EZ Imager (Bio-Rad).

Single-Cell Gel Electrophoresis Assay (Comet Assay)

A2780 cells were seeded in 6-well plates with a density of 1 × 10⁶ cells/well and incubated for 24 h in an incubator with a humidified atmosphere at 5% (v/v) CO₂ and 37 °C. After that time, the medium was replaced by a medium with 10× IC₅₀ concentrations of complexes 2–4, 0.1% (v/v) DMSO as vehicle control or 0.05% (v/v) H₂O₂. The complexes and DMSO were incubated at 37 °C for 12 h and H₂O₂ for 30 min at RT. Cells were harvested with a cell scraper and solutions with 1 × 10⁵ cells/mL in PBS were prepared. It was withdrawn 10 μL from the solutions and added to 90 μL of 1.5% (w/v) low melting point agarose in 1× PBS and dropped on the slides prepared previously with a coating of 1.5% (w/v) agarose (normal melting point agarose) in 1× TAE buffer. After drying at 4 °C for 15 min, the slides were dipped into lysis solution (450 mM NaCl, 3.72% EDTA, 5 mM Tris; to which 10% (v/v) DMSO and 1% (v/v) Triton-X were added just before use) for 1 h at 4 °C, followed by their being dipped for 40 min into cold electrophoresis solution (1 mM EDTA, 300 mM NaOH, pH 13). Electrophoresis was performed at 4 °C for 30 min at 25 V using a Sub-Cell model 96 apparatus (Bio-Rad). Afterward, slides were placed in 0.1 M Tris–HCl buffer (pH 7.5) at 4 °C for 15 min for neutralization, followed by 15 min in methanol at 4 °C and drying at 37 °C. For visualization of comets, slides were hydrated with distilled water at 4 °C for 30 min, stained with 20 μL of GelRed working solution (3×), and covered with new coverslips. The slides were observed under a Ti–U Eclipse inverted microscope (Nikon Instruments, Japan) and analyzed by CometScore version 2.1 (TriTek). About 100 cells per sample were analyzed. The percentage of DNA in the tail was used as a measure of the total DNA strand breakage.

Calf Thymus DNA (CT-DNA) Binding Assays

Calf Thymus DNA (CT-DNA) binding assay was performed through the UV–vis spectra from 230 to 700 nm on the spectrophotometer Evolution 300 UV–vis (Thermo Fischer Scientific, Waltham, MA, USA). It was used a filtered buffer (5 mM Tris–HCl, 50 mM NaCl, pH 7.02) for the dilution of 0–50 μM of CT-DNA and 25 μM of the complexes 2–4. Control solutions of CT-DNA, CT-DNA + DMSO, and DMSO were also prepared. All solutions were incubated for 24 h at 37 °C and then analyzed. CT-DNA concentration in base pairs had been determined at 260 nm in a NanoDrop2000 spectrophotometer (Thermo Scientific), using an extinction coefficient of 6600 M^–1 cm^–1. The dilution effect because of the addition of the DNA solution was corrected and the affinity constants were calculated according to eq 2:

where [DNA] is the concentration of CT-DNA (per nucleotide phosphate), ε_a = Abs/[complex], ε_f is the extinction coefficient for the free complex, and ε_b is the extinction coefficient for [Co(η⁵-C₅H₅)(PPh₃)(NN)][(CF₃SO₃)₂] when fully bound to DNA. DNA concentration (expressed as the molarity of phosphate groups) was determined with a NanoDrop 2000 spectrophotometer assuming ε₂₆₀ = 6600 M^–1 cm^–1. A melting profile of 10 μM CT-DNA and 10 μM CT-DNA with 10 μM of the complexes 2–4, was also performed. Evagreen dye was added to all samples after a 1h 30 min incubation of CT-DNA with the complexes at 37 °C. Melting curves were analyzed in a Corbet RotorGene 6000 (Biorad, Spain) by a first incubation step at 37 °C for 5 min followed by a ramp from 37 to 95 °C with 90 s hold in the first step and 5 s hold in the next steps with green acquisition.

Cell Migration Assay

Fibroblasts were seeded in 24-well plates at a density of 4 × 10⁵ cells/mL and grown in the same conditions as described previously until a confluent monolayer was obtained. The medium was changed by a medium with DMSO 0.1% (v/v) or the IC₅₀ of complexes 2–4 and a scratch was made on the surface of each well. Images of the Fibroblasts were captured right after the exposure (0 h) to the complexes and after 24 h of incubation at 37 °C and 5% (v/v) CO₂ in a humidified atmosphere, using CytoSMART Lux2 Live Cell Imager (Axion biosystems, USA) and Ti–U Eclipse Inverted microscope (Nikon Instruments, Japan). Images were analyzed with the ImageJ MRO Wound Healing Tool or with the scratch assay measurement of the CytoSMART cloud portal for the Lux system.

Ex Ovo Chick Embryo Yolk Sac Membrane (YSM) In Vivo Assay

Ex ovo chick embryo yolk sac membrane (YSM) in vivo assay was performed as previously described.^51,52 First, the chicken embryos were incubated for 24 h at 37 °C, 5% (v/v) CO₂, and humidified atmosphere, to stabilize the embryos, and after that, it was placed O-rings with a solution containing DMSO 0.1% (v/v), as the negative control, or the IC₅₀ of complexes 2–4 in PBS 1× in the middle. The embryos were incubated for 48 h at 37 °C and images were captured after 0, 24, and 48 h. Newly formed blood vessels were manually counted, via ImageJ software.

The ex ovo YSM assay fulfills the Directive 2010/63/EU of the European Parliament to protect animal models for scientific purposes.

Statistical Analysis

GraphPad Software (GraphPad Prism version 8.01 for Windows, GraphPad Software, La Jolla, CA, USA, www.graphpad.com) was used to perform the statistical analysis of the data performing a comparison using one-way ANOVA with a confidence interval of 95%.

Results and Discussion

Design and Synthesis of New CoCp Complexes

A new family of dicationic Co(III)-cyclopentadienyl (CoCp) complexes with N,N-heteroaromatic bidentate ligands of general formula [CoCp(PPh₃)(NN)][(CF₃SO₃)₂], where NN = 2,2′-bipyridine (bipy; complex 1); 4,4′-dimethyl-2,2′-bipyridine (Me₂bipy, complex 2); 1,10′-phenanthroline (phen; complex 3); or 5-amino-1,10′-phenanthroline (NH₂phen, complex 4), was synthesized according to Scheme 1. The four new complexes were prepared from [CoCp(CO)I₂] by σ-chelation of the N,N-heteroaromatic ligands after iodide abstraction with silver trifluoro methanesulfonate, followed by σ-coordination of the triphenylphosphane coligand. The reactions were carried out in situ, in acetonitrile at 0 °C for 10 and 5 min, respectively, achieving the desired products after recrystallization by slow diffusion of diethyl ether in dichloromethane or methanol at room temperature. The CoCp complexes 1–4 were obtained in moderate-high yields (54 to 78%), in the same order of magnitude as the analogous RuCp complexes with triphenylphosphane and bidentate N,N-heteroaromatic ligands.^29,30

The solid-state FT-IR spectra of complexes 1–4 showed the characteristic ν_C–H and ν_C=C stretching bands of the aromatic rings of the cyclopentadienyl and N,N-heteroaromatic ligands in the ranges of 3120–3064 and 1583–1433 cm^–1, respectively. All complexes also showed bands at 1261–1257 cm^–1 characteristic of the CF₃SO₃^– counterions and at 1624–1604 cm^–1 assigned to ν_C=N vibrations of the ligands. Complex 2 also presented the ν_C–H and δ_CH3 vibrations of the methyl groups of Me₂bipy in the ranges of 3120–3082 and 1487–1436 cm^–1, respectively. For complex 4, the vibrations of the amine group of the NH₂Phen ligand were observed at values 3250 cm^–1 (ν_N–H) and 1645–1598 cm^–1 (δ_N–H).

The new complexes were also characterized by ¹H, APT-¹³C{H}, and ³¹P{H} NMR experiments complemented by 2D experiments (COSY, HMQC, and HMBC) in acetone-d₆ at room temperature; see Scheme 1 for the atom labeling of the coordinated N,N-heteroaromatic ligands. The ¹H NMR spectra of the free ligands were also recorded under the same experimental conditions for comparison. As a general trend, a deshielding was observed on the ¹H NMR signals of the N,N-heteroaromatic ligands upon coordination to the metal center in all complexes, a feature particularly perceived in the ortho protons (Δ_δ H₁/H₁₀ = 0.83–1.04 ppm), which is a strong evidence of a successfully σ-coordination. The assignment of the H₁/H₁₀ protons is consistent with the relatively low values of their vicinal coupling constants (³J_HH = 4.07–5.00 Hz), which are in good agreement with the literature values for ³J_HH of protons adjacent to electronegative atoms such as nitrogen.⁶⁴ Moreover, deshielding of the ¹H NMR signals of the methyl and amino groups from the ligands of complexes 2 and 4 was also observed upon coordination (Δ_δ H₆ = 0.12 ppm, and Δ_δ H₁₃ = 0.88 ppm, respectively), being found within their typical range values. The ¹H resonances of the η⁵-cyclopentadienyl (Cp) ring of all the compounds were also found within the expected range for dicationic half-sandwich cobalt(III) complexes (δ H_Cp = 6.56–6.72 ppm).^65,66 As expected, upon coordination, a deshielding of the Cp ring signal was observed compared to the neutral precursor [CoCp(CO)I₂] (Δ_δ H_Cp = 0.61–0.77 ppm), mostly due to the change in the overall charge of the complexes. A π-back-donation effect may also play a role. Indeed, the magnitude of this deshielding is attenuated within the new cobalt complexes following the order 2 > 1 > 4 > 3, reflecting the σ donor nature of the N,N-hetereroaromatic ligands to the metal center.^67,68 A general deshielding of the ¹H resonances of the triphenylphosphane coligand was observed as expected upon its coordination in all four cobalt complexes, particularly for the para protons (Δ_δ H_para = 0.16 to 0.30 ppm). Interestingly, for complexes 3 and 4, with the phenantroline coligands, a slight shielding of the ortho protons (Δ_δ H_ortho = −0.12 ppm) was also found, suggesting the involvement of the P center in the overall extended π back-donation system. The ³¹P NMR spectra are in good agreement with these findings. An accentuated deshielding of the single sharp signal, attributed to the triphenylphosphane, was observed upon coordination (Δ_δ PPh₃ = 46.69–47.25 ppm), coherent with the strong σ-donor character of this coligand and the overall dicationic charge character of the cobalt complexes. The APT-¹³C{H} NMR spectra are consistent with the results discussed above. However, it is interesting to note that, generally the ¹³C resonances of the complexes generally do not follow the same order as the respective ¹H signals. Nonetheless, a similar pattern upon coordination as in the ¹H NMR spectra was observed. Figure S1 exemplifies the general behavior observed in complexes 1–4 and depicts the ¹H and ³¹P{H} NMR spectra of [CoCp(PPh₃)(bipy)][(CF₃SO₃)₂] (1) compared to the spectra of its precursor complex and free ligands.

The electronic absorption spectra of the four CoCp complexes (1–4) were acquired in dichloromethane solutions (1 × 10^–5–1 × 10^–3 M) at room temperature. For comparison, the electronic spectra of the precursor [CoCp(CO)I₂] and free ligands were recorded under the same experimental conditions. The corresponding values of absorption maximum are collected in Table S1 of the Supporting Information. Figure 1 depicts the general behavior for this set of complexes, showing the electronic spectra of [CoCp(PPh₃)(bipy)][(CF₃SO₃)₂] (1), its precursor complex, and the free ligand.

Electronic spectra of [CoCp(PPh₃)(bipy)][(CF₃SO₃)₂] (1, ––), its precursor [CoCp(CO)I₂] (**- - -**), and free bipy ligand (^···) in dichloromethane.

All complexes show intense absorption bands in the UV region at 235–370 nm, characteristic of π → π* intraligand electronic transitions. Additionally, a band in the visible region (465–467 nm) was also observed in the electronic spectra of all complexes except for 4, which was found blue-shifted, appearing at 339 nm with ε (molar extinction coefficient) of 3.15 × 10^–3 M^–1·cm^–1, with the same order of magnitude as the ones placed in the visible area (464–467 nm). This band is tentatively assigned as a metal-to-ligand d_Co to π*_Cp charge transitions (MLCT). Nonetheless, other contributions cannot be ruled out. Further clarification can be found below in the DFT calculations section. For complexes 1 and 2, coordination of the bipyridyl-derived ligands led to an enhancement of the UV bands at 236–312 nm, assigned to the π → π* transitions occurring in the ligand. For complexes 3 and 4, containing the phenanthroline-derived ligands, no significant alterations were observed in the intensity of the UV bands upon coordination of the N,N-heteroaromatic ligand. Overall, the new cobalt complexes show similar electronic spectra to their analogous complexes containing other metals, such as ruthenium, although, in these cases, the origin of the MLCT involved the d_Ru orbitals to π*_NNligand.^29,30 The existence of MLCT bands denoting π interaction of the metal with coordinate ligands whether η⁵-Cp or NN can constitute a piece of information concerning the further design of new compounds since they allow us to fine-tune the electronic flow inside the molecule. In the present cobalt derivatives, the electronic richness at Cp might be inestimable for its functionalization, thus opening a new area of synthesis.

The percentual elemental analyses of all the complexes were also in accordance with the proposed formulations.

Stability in Aqueous Solution and Estimation of Lipophilicity

Determining the stability of organometallic complexes in aqueous solution before in vitro biological studies is a crucial requirement for any substance with a potential biomedical application. Therefore, the stability of the CoCp complexes in aqueous solution (95% D₂O/5% DMSO-d₆) was evaluated over 24 h at room temperature by ¹H and ³¹P{H} NMR spectroscopy. A small percentage of DMSO was needed to facilitate the dissolution of the complexes as they were only partially soluble in water at the tested concentrations (2.5–3.5 mM). The collected spectra do not display any significant variation over time regarding the number, chemical shift, integration, or multiplicity of the ¹H and ³¹P resonances. Figure 2 represents the evaluation of the stability of complex 1 over 24 h. All CoCp complexes are stable in aqueous medium for at least 24 h, and thus suitable for further biological studies.

Evaluation of the stability of complex 1 in 90% D₂O/10% DMSO-d₆ solution (2.5 mM) over 24 h, by ¹H NMR (left) and ³¹P{H} NMR (right).

Lipophilicity/hydrophilicity is among the most important physicochemical properties in biomedical applications, as it considerably influences its cytotoxicity/bioactivity, general toxicity, cell membrane/tissue permeability, availability to interact with drug targets, biodistribution, and excretion, as well as drug formulation.⁶⁹ Thus, evaluating the lipophilicity of a drug candidate during the initial phase of the drug development process is a crucial step. Herein, the n-octanol/water partition coefficients of the CoCp complexes were estimated by the shake-flask method at room temperature.⁵⁴ All the complexes are slightly hydrophilic, nonetheless, complexes containing bipyridine ligands (logP1 = −0.67 ± 0.02; logP2 = −0.56 ± 0.02) are less lipophilic than those containing the phenanthroline ligands (logP3 = −0.28 ± 0.05; logP4 = −0.22 ± 0.03).

DFT Calculations

To further elucidate the structural features, owing to the absence of X-ray structures and help in the interpretation of the electronic absorption spectra, DFT and TD-DFT calculations were performed on complexes 1–4. We first started by optimizing their geometries (Figure 3). Notice that due to the pro-chiral nature of the CoCp(PPh₃) fragment and the single substitution of the NH2phen ligand in 4, both diastereoisomers were considered (denoted 4′ and 4″). All complexes feature the classic piano-stool geometry and despite the difference in the NN ligand, no substantial structural changes are observed. Nonetheless, when the NN ligand changes from the bipy-based (1–2) to phen-based (3–4), a slight contraction of the Co–P bond is observed (from 2.30 to 2.28 Å) with a concomitant slight elongation of the Co–N bonds (1.93–1.94 Å). This observation is in good agreement with the shielding observed on the H_ortho protons of the PPh₃ attributed to π back-donation d_Co-to π*_Cp.

DFT optimized structures of complexes 1–4. For 4, the possible pair of diastereoisomers was considered (denoted as 4′ and 4″). Relevant distances are given in Angstrom.

Subsequently, TD-DFT calculations were performed in dichloromethane for all complexes to help in the interpretation of the UV–vis data. Figure 4 shows a superposition of the experimental spectra with the calculated excitations, while the values for the relevant excitations, highlighted with an asterisk, along with their oscillator strengths and composition are listed in Table S2. The superposition of the experimental spectra with the calculated excitations is exceptional for complexes 1–3, while for complex 4, the agreement is much worse. Despite complex 4 showing a different behavior experimentally, the TD-DFT calculations predict very similar excitation patterns for all complexes, comprising low-energy excitations at >450 nm, which should account for the experimental bands observed in the visible region (465–467 nm) for 1–3, and two main excitations at higher energies. The excitations possess significant contributions from several orbitals (Table S2), therefore, a natural transition orbital (NTO) analysis⁷⁰ was performed to assist in the assignment of the transitions. Figure 5 illustrates the NTO pairs for the calculated low-energy transitions for complexes 1–4. The excitation possesses a large MLCT character, mainly from d_Co to the Cp ligand, though a non-negligible contribution of d–d character is also observed. The π system of the NN ligand does not participate in these low-energy excitations in all complexes. As stated, the TD-DFT calculations predict the existence of an MLCT band for complex 4, despite this band seeming absent in the experimental spectrum. A possible explanation is that the UV–vis spectrum for complex 4 is blue-shifted, hence, the 339 nm band should correspond to the MLCT transition calculated at 459 nm (see Table S2). Indeed, as mentioned above upon the discussion of the experimental UV–vis spectra of these new CoCp compounds, a band at 339 nm was found for compound 4 with a ε value compatible with a blue-shifted MLCT band (ε = 3.15 × 10^–3 M^–1·cm^–1 in the same order of magnitude as the MLCT bands observed for 1, 2 and 3 compounds). The NTO pairs for the remainder of relevant high-energy excitations are shown in Figures S2–S5 and confirm the LLCT/ π → π* nature of the transitions.

TD-DFT excitations plotted against the UV–visible absorption spectra for complexes 1–4. The excitations reported in Table S2 are highlighted (*). Both calculated and experimental spectra were obtained in dichloromethane.

Representative natural transition orbital pairs along with the respective coefficients for the TD-DFT calculated low-energy excitations of complexes 1-4. For 4, only 4″ is represented for simplicity.