Osmium Arene Germyl, Stannyl, Germanate, and Stannate Complexes as Anticancer Agents.

Herein, we describe the synthesis, full spectroscopic characterization, DFT (density functional theory) calculations, and single-crystal X-ray diffraction analyses of a series of osmium arene σ-germyl, germanate, σ-stannyl, and stannate complexes, along with their cytotoxic (anticancer) investigations. The known dimer complexes [OsCl2(η6-C6H6)]2 (1) and [OsCl2(η6-p-cymene)]2 (2) were reacted with PPh3 to form the known mononuclear complex [OsCl2(η6-p-cymene)(PPh3)] (3) and the new complex [OsCl2(η6-C6H6)(PPh3)] (6); complex 3 was reacted with GeCl2·(dioxane) and SnCl2 to afford, by insertion into the Os-Cl bond, the neutral σ-germyl and stannyl complexes [OsCl(η6-p-cymene)(PPh3)(GeCl3)] (7) and [OsCl(η6-p-cymene)(PPh3)(SnCl3)] (11), respectively, as a mixture of enantiomers. Similarly, the reaction of complex 6 with GeCl2·(dioxane) afforded [OsCl(η6-C6H6)(PPh3)(GeCl3)] (9). Complex 2, upon reaction with 1,1-bis(diphenylphosphino)methane (dppm), formed a mixture of [OsCl2(η6-p-cymene)(κ1-dppm)] (4) and [Os(η6-p-cymene)(κ2-dppm)Cl]+Cl- (5) when prepared in acetonitrile and a mixture of 4 and the dinuclear complex [[OsCl2(η6-p-cymene)]2(μ-dppm)] (0) when prepared in dichloromethane. By utilizing either isolated 4 or a mixture of 4 and 5, the synthesis of κ2-dppm germanate and stannate salts, [OsCl(η6-p-cymene)(κ2-dppm)]+GeCl3 - (8) and [OsCl(η6-p-cymene)(κ2-dppm)]+SnCl3 - (10), were accomplished via halide-abstracting reactions with GeCl2·(dioxane) or SnCl2, respectively. All resulting complexes were characterized by means of multinuclear NMR, FT-IR, ESI-MS, and UV/Vis spectroscopy. X-ray diffraction analyses of 4, 8, 9, 10, and 11 were performed and are reported. DFT studies (B3LYP, basis set LANL2DZ for Os, and def2-TZVPP for Sn, Ge, Cl, P, C, and H) were performed on complex 9 and the benzene analogue of complex 11, 11-benzene, to evaluate the structural changes and the effects on the frontier molecular orbitals arising from the substitution of Ge for Sn. Finally, complexes 3 and 7-11 were investigated for potential anticancer activities considering cell cytotoxicity and apoptosis assays against Dalton's lymphoma (DL) and Ehrlich ascites carcinoma (EAC) malignant cancer cell lines. The complexes were also tested against healthy peripheral blood mononuclear cells (PBMCs). All cell lines were also treated with the reference drug cisplatin to draw a comparison with the results obtained from the reported complexes. The study was further corroborated with in silico molecular interaction simulations and a pharmacokinetic study.

Introduction

Despite the conclass="Chemical">tinuous decliclass="Chemical">ne iclass="Chemical">n class="Chemical">n class="Disease">cancer death rates in recent years, cancer remains one of the leading causes of death worldwide.[1] Chemotherapy remains the principal form of cancer treatment,[2] and in recent years, much research has focused on the development of targeted therapies that bypass the off-target toxic side effects of chemotherapy.[3] However, such therapies are often reliant on targetable pheno- and genotypes. As such, chemotherapy remains a reliable form of treatment for tumors that do not display such specific targets.[4]

Traditionally, class="Chemical">platinum-class="Chemical">n class="Chemical">based drugs occupy a prominent role in anticancer treatment[5] but causes severe side effects.[6] Additionally, an acquired resistance to platinum-based agents displayed by many tumors further limits their use in clinical practice.[7] These drawbacks have led to the pursuit of nonplatinum metal complexes that display superior anticancer activity while potentially minimizing the negative side effects associated with platinum agents.[8] Among these, ruthenium,[9] iridium,[10] gold,[11] and osmium[12] compounds have been established as potential candidates in the past 20 years, with ruthenium complexes being the most well studied and showing promise in a number of clinical trials.

We have reported a series of neutral class="Chemical">ruthenium-class="Chemical">n class="Chemical">based germyl ([Ru]–GeCl3) and stannyl ([Ru]–SnCl3) complexes that displayed less than optimal cytotoxic activity on numerous cell lines in vitro (Scheme ).[13,15] In contrast, ruthenium germanate and stannate salt complexes displayed potent and promising cytotoxic activity on two breast cancer cell lines (MCF-7 and MDA-MB-231) and good selectivity, suggesting that the ionic nature of the complexes influences their biological effects.[14]

Scheme 1

Several Previously Reported Ruthenium-Based Germyl, Stannyl, Germanate, and Stannate Complexes,[13a,14,15]

class="Chemical">Osmium, as a heavier aclass="Chemical">nalogue of class="Chemical">n class="Chemical">ruthenium, might be expected to be a suitable metal for anticancer applications; however, when compared to ruthenium, research into osmium-based anticancer agents is less pronounced. Osmium is the least abundant metal on Earth, which might limit its application with regard to widespread use in clinical applications. Nevertheless, in contrast to ruthenium, osmium complexes exhibit some different characteristics: slow ligand exchange kinetics,[16] predilection for higher oxidation states, elevated levels of inertness,[2b] and stronger π-backdonation when in low oxidation states.[17] Many osmium complexes have been synthesized and tested as potential anticancer agents,[12a,18] which in comparison to analogous ruthenium compounds suggest that osmium as a metal center may be beneficial.[12e–12g] Other osmium arene complexes have shown high cytotoxicity, although their mechanism implied DNA unwinding rather than bending.[19] Depending on the structure and the oxidation state of the osmium metal in numerous complexes, varied mechanisms of action have been observed. For example, synthesized octahedral complexes a and b (Scheme ) both induced cellular apoptosis: b by inducing endoplasmic reticulum stress and enlargement as well as upregulation of protein p53, whereas a by interrupting the replication cycle at the G2/M stage.[12a] Most importantly, a exhibited activity against cancer stem cells that are responsible for cancer recurrence.[12a] Another osmium complex, FY26 (Scheme ), inhibits tumor growth by increasing the production of reactive oxygen species (ROS) inside the targeted cells.[20]

Scheme 2

Selected Examples of Osmium-Based Anticancer Agents,[12a,20]

Several recent reviews have been devoted to the potential of class="Chemical">osmium aclass="Chemical">nticlass="Chemical">n class="Disease">cancer agents.[2b,17,19b] We have also recently reviewed all existing osmium complexes mutually bearing π-bound arene groups and a phosphane coligand in a biological context.[12a] Inspired by our earlier work on ruthenium-based stannyl/stannate and germyl/germanate complexes, we envisaged an extension of that work to osmium. Surprisingly, osmium arene complexes bearing σ-bound germyl or stannyl ligands are exceptionally rare in the literature: only two reports by Wen et al.[21] and Albertin et al.[22] exist. No osmium-based germanate or stannate complexes have been reported in the literature to this date to the best of our knowledge. The anticancer or biological activity investigations of any of these four classes of compounds or elucidations of their bonding nature by theoretical methods have not previously been evaluated. Moreover, given our previous numerous studies on ruthenium arene stannyl and stannate complexes, as well as their germanium analogues,[13a,14,15] the extension to osmium is expected to yield different biological activities given the rather inert nature of Os(II) vs Ru(II) (see above). This hence motivated us to explore the analogous osmium chemistry with a view of biological anticancer applications, not as a mere extension from Ru, but expecting different in vitro biological activities compared to analogous ruthenium complexes.

Herein, we describe the synthesis and full spectclass="Chemical">roscopic characterizatioclass="Chemical">n of rare examples of class="Chemical">n class="Chemical">osmium arene-based germyl and stannyl complexes and the first germanate and stannate complexes of osmium. We also report the in vitro activities of the complexes against Dalton’s lymphoma (DL) and Ehrlich ascites carcinoma (EAC) malignant cancer cell lines as well as healthy peripheral blood mononuclear cells (PBMCs) using MTT and apoptosis assays. Furthermore, an in silico molecular interaction study and pharmacokinetic tests were also performed to support the results.

Materials and Methods

General Considerations

All reactions were conducted under a class="Chemical">nitrogen atmclass="Chemical">n class="Gene">osphere using standard glovebox and Schlenk techniques, unless otherwise stated. Chemicals used were obtained from commercial sources and used as received. OsCl3·xH2O was purchased from Strem Chemicals. 1,3-Cyclohexadiene, α-terpinene (85%), triphenylphosphine (PPh3), and bis(diphenylphosphino)methane (97%) (dppm) were purchased from Sigma-Aldrich. Reagent-grade diethyl ether (stab. BHT) and dichloromethane (stab. amylene) were purchased from Biosolve Chimie SARL. [OsCl2(η6-C6H6)]2 (1), [OsCl2(η6-p-cymene)]2 (2), and [OsCl2(η6-p-cymene)(PPh3)] (3) were synthesized according to literature methods.[23] All solvents were purged with nitrogen prior to use. The NMR data were acquired on a 300 MHz Ultrashield Magnet System (Bruker). The NMR samples were prepared in deuterated chloroform (CDCl3), unless otherwise stated. Chemical shifts of 31P{1H} NMR peaks were reported relative to phosphoric acid (85%), and 1H and 13C{1H} NMR resonance signals were measured relative to tetramethylsilane (TMS) and were reported in parts per million. Abbreviations: s = singlet; d = doublet; t = triplet; dd = doublet of doublets; dt = doublet of triplets; pst = pseudo triplet; m = multiplet; sept = septet. NMR data were processed using TopSpin 4.0.8 software or MestReNova. Infrared spectroscopy was performed on a Shimadzu MIRacle 10 spectrometer (Shimadzu single reflexion ATR accessory). The analysis was performed with the following settings: 128 scans; resolution, 2; Happ–Genzel apodization. Abbreviations: v = variable; w = weak; m = medium; s = strong; br = broad. The IR spectra were processed using IR solution software. The UV/Vis absorption spectra were recorded with a Shimadzu UV-1800 spectrophotometer using 1 cm path length quartz cuvettes with a reference cell containing dichloromethane, and the maximum absorption peak(s) were reported. Melting points were measured in glass capillary tubes on a Stuart BioCote SMP10 machine and performed in triplicate. An average of the measurements was reported in degrees Celsius and was uncorrected. A direct injection method was utilized for EI-MS measurements performed on a GCMS-QP2010 Ultra (Shimadzu). For the high-resolution ESI-MS measurements, the samples were sent either to Technische Universität Berlin where the data were acquired with an Orbitrap LTQ XL mass spectrometer (Thermo Scientific) or to Maastricht MultiModal Molecular Imaging Institute (M4I) where the results were acquired on a Bruker SolariX XR FT-ICR-MS. The signals for both EI-MS and ESI-MS were thoroughly checked and compared to the theoretical isotope patterns predicted by online software enviPat Web 2.4.[24] The peaks with the highest intensity from the corresponding isotope pattern were reported. Crystals suitable for X-ray diffraction analysis were analyzed at TU Graz, Austria, on a BRUKER-AXS SMART APEX CCD diffractometer using graphite-monochromated Mo Kα radiation (0.71073 Å).

Synthesis of a Separable Mixture of [OsCl2(η6-p-cymene)(κ1-dppm)] (4) and [OsCl(η6-p-cymene)(κ2-dppm)]+Cl– (5)

A total of 185 mg (0.481 mmol, 2 equiv) of class="Chemical">dppm was reacted with 190.2 mg (0.241 mmol) of 2 iclass="Chemical">n 30 mL of class="Chemical">n class="Chemical">acetonitrile. The solution was refluxed under nitrogen for 1.5 h. The resulting orange solution was filtered and washed with acetonitrile; the solvent was then evaporated in vacuo. The resulting translucent yellow oil was washed with diethyl ether (3 × 10 mL), which was subsequently evaporated using a rotary evaporator. At this point, 31P{1H} NMR spectroscopy revealed the presence of two products: the desired 4 and another complex [Os(η6-p-cymene)(κ2-dppm)Cl]+Cl– (5). Thereafter, fractional recrystallization was performed using dichloromethane and diethyl ether. The supernatant was collected via decantation. The solvent of this supernatant was then evaporated to dryness, yielding an orange solid. The solvents remaining in the original reaction vessel were removed in vacuo, affording a yellow solid. The supernatant fraction was determined to be 4. Air-stable. 120.8 mg (0.155 mmol) of orange solid (32%). Melting point: 163 °C + dec. 1H NMR (300 MHz, CDCl3, 298 K): δ 7.85–7.81, 7.33–7.30 and 7.15–7.04 (20H, all m, phenyls, κ1-dppm), 5.44 (2H, d, 3JH–H = 5.4 Hz, C2,6 or 3,5H, η6-p-cymene), 5.32 (2H, d, 3JH–H = 5.6 Hz, C2,6 or 3,5H, η6-p-cymene), 3.55 (2H, dd, 2JP = 8.7 Hz, 2JP = 1.8 Hz, PCH2P), 3.48 (0.07 4·OEt2), 2.34 (1H, sept, 3JH–H = 6.9 Hz, CH(CH3)2), 1.98 (3H, s, CH3), 1.21 (0.07 4·OEt2), 0.83 (6H, d, 3JH–H = 6.9 Hz, CH(CH3)2). 31P{1H} NMR (121.4 MHz, CDCl3, 298 K): δ −16.1 (d, 2JP–P = 34.9 Hz, OsP), −28.8 (d, 2JP–P = 34.9 Hz, CH2P). FTIR (cm–1): 3055 (w), 2968 (w), 2875 (w), 1481 (vw), 1434 (m), 1374 (w), 1316 (w), 1269 (vw), 1204 (w), 1185 (w), 1155 (w), 1128 (w), 1097 (m), 1027 (w), 999 (w), 900 (w), 874 (vw), 800 (w), 769 (vm), 757 (m), 745 (s), 730 (s), 708 (s), 690 (s), 659 (m). UV/Vis (dichloromethane): λmax = 351 nm. The precipitate of the original reaction vessel was determined to be [OsCl(η6-p-cymene)(κ2-dppm)]+Cl– (5). Air-stable. 154.0 mg (0.198 mmol) of yellow solid (41%). Melting point: 176 °C + dec. 1H NMR (300 MHz, CDCl3, 298 K): δ 7.61–7.46 (20H, m, phenyls, κ2-dppm), 6.47 (1H, dt, 2JH–H = 15.3 Hz, 2JH–P = 9.8 Hz, PCHA or BP), 6.29 (4H, d, 3JH–H = 37.3 Hz, C3,5 and 2,6H, η6-p-cymene), 3.48 (0.33 5·OEt2), 4.70 (1H, dt, 2JH–H = 15.1 Hz, 2JH–P = 12.8 Hz, PCHA or BP), 2.41 (1H, sept, 3JH–H = 6.9 Hz, CH(CH3)2), 1.67 (3H, s, CH3), 1.21 (0.33 5·OEt2), 1.09 (6H, d, 3JH–H = 6.9 Hz, CH(CH3)2). 13C{1H} NMR (75.4 MHz, CDCl3, 298 K): δ 132.2 (pst, J = 5.0 Hz, κ2-dppm), 131.8 (pst, J = 5.0 Hz, κ2-dppm), 129.6 (pst, J = 5.5 Hz, κ2-dppm), 128.8 (pst, J = 5.5 Hz, κ2-dppm), 128.1 (s, C4H,κ2-dppm), 114.9 (s, C1 or 4, η6-p-cymene), 95.2 (s, C1 or 4, η6-p-cymene), 84.3 (s, C2,6 or 3,5H, η6-p-cymene), 82.2 (t, 2JC–P = 3.7 Hz, C2,6 or 3,5H, η6-p-cymene), 65.9 (5·OEt2), 46.1 (t, 1JC–P = 32.3 Hz, PCH2P), 30.5 (s, CH(CH3)2), 22.1 (s, CH(CH3)2), 17.2 (s, CCH3), 15.3 (5·OEt2). 31P{1H} NMR (121.4 MHz, CDCl3, 298 K): δ −40.0 (s).

Synthesis of a Mixture of [OsCl2(η6-p-cymene)(κ1-dppm)] (4) and [OsCl(η6-p-cymene)(κ2-dppm)]+Cl– (5)

The reaction was performed in 30 mL of degassed class="Chemical">acetonitrile. A total of 146.5 mg (0.185 mmol, 1 equiv) of 2 was dissolved aclass="Chemical">nd refluxed with 143.4 mg (0.373 mmol, 2 equiv) of class="Chemical">n class="Chemical">dppm for 1.5 h under nitrogen. After the reaction was complete, the solution was filtered and the solvent was removed in vacuo. The crude product was then washed with diethyl ether (3 × 10 mL), the washings were discarded, and the residue was dried in vacuo, yielding 228.7 mg (0.293 mmol) of a mixture of 4 and 5. This mixture was subsequently used for the reactions with germanium(II) chloride·dioxane and with tin(II) chloride to yield 8 and 10, respectively. Air-stable. All the spectral information was identical to those retrieved from the separated products 4 and 5 (see above).

Synthesis of a Mixture of [OsCl2(η6-p-cymene)(κ1-dppm)] (4) and [[OsCl2(η6-p-cymene)]2(μ-bis(diphenylphosphino)methane)] (0)

First, 121.2 mg (0.153 mmol, 1 equiv) of dimer 2 was dissolved in 30 mL of degassed class="Chemical">dichloromethane. Theclass="Chemical">n, 118.0 mg (0.307 mmol, 2 equiv) of class="Chemical">n class="Chemical">dppm was added to the flask and the solution was refluxed under nitrogen for 1.5 h. The resultant golden solution was filtered and the solvent was removed in vacuo. The obtained oily product was then washed with diethyl ether (3 × 10 mL), and the residue was dried in vacuo. A total of 133.9 mg of a mixture of two products was isolated: 4: 1H NMR (300 MHz, CDCl3, 298 K): δ 7.87–7.80 (4H, m, C2,6H of OsP(C6H5)2 or CH2P(C6H5)2, dppm), 7.33–7.28 (6H, m, C3–5H of OsP(C6H5)2 or CH2P(C6H5)2, dppm), 7.14–7.10 (10H, m, OsP(C6H5)2 or CH2P(C6H5)2, dppm),5.44 (2H, d, 3JH–H = 5.6 Hz, C2,6 or 3,5H, η6-p-cymene), 5.32 (2H, d, 3JH–H = 5.6 Hz, C2,6 or 3,5H, η6-p-cymene), 3.55 (2H, dd, 2JP = 8.9 Hz, 2JP = 1.8 Hz, PCH2P), 2.41 (1H, sept, 3JH–H = 6.9 Hz, CH(CH3)2), 1.98 (3H, s, CCH3), 0.83 (6H, d, 3JH–H = 6.9 Hz, CH(CH3)2). 13C{1H} NMR (75.4 MHz, CDCl3, 298 K): i 137.3 (dd, 1JC – P = 14.1 Hz, 3JC – P = 5.7 Hz, C1PA or B, dppm), 132.5 (dd, 1JC – P = 8.8 Hz, 3JC – P = 2.3 Hz, C1PA or B, dppm), 131.7 (s, C4H, PA or B, dppm), 131.4 (s, C4H, PA or B, dppm), 129.7 (d, YJC–P = 2.0 Hz, C3,5 or 2,6H, dppm), 127.5–126.7 (m, C2,6 or 3,5H, dppm (overlapping with signals from 0)), 97.6 (s, C1 or 4, η6-p-cymene), 85.2 (s, C1 or 4, η6-p-cymene), 81.0 (d, 2JC–P = 4.0 Hz, C2,6 or 3,5H, η6-p-cymene), 77.1 (d, 2JC–P = 5.2 Hz, C2,6 or 3,5H, η6-p-cymene), 28.7 (s, CH(CH3)2), 20.6 (s, CH(CH3)2), 17.9 (t, 1JC–P = 31.0 Hz, PCH2P), 16.2 (s, CCH3). 31P{1H} NMR (121.4 MHz, CDCl3, 298 K): δ −16.1 (d, 2JP–P = 34.9 Hz, OsP), −28.8 (d, 2JP–P = 34.9 Hz, CH2P). Dinuclear species (0): 1H NMR (300 MHz, CDCl3, 298 K): δ 7.57–7.51 (12H, m, C3–5H, dppm), 7.20–7.14 (8H, m, C2,6H, dppm), 7.03 (4H, d, 3JH–H = 5.5 Hz, C2,6 or 3,5H, η6-p-cymene), 5.13 (4H, d, 3JH–H = 5.5 Hz, C2,6 or 3,5H, η6-p-cymene), 4.68 (2H, t, 2JH–P = 7.6 Hz, PCH2P), 2.32 (2H, sept, 3JH–H = 6.9 Hz, CH(CH3)2), 2.05 (6H, s, CCH3), 0.98 (12H, d, 3JH–H = 6.9 Hz, CH(CH3)2). 13C{1H} NMR (75.4 MHz, CDCl3, 298 K): δ 133.4–133.1 (m, C2,6 and 3,5H, dppm), 131.2 (d, 2JC–P = 2.2 Hz, C2,6 or 3,5H, η6-p-cymene), 130.5 (d, 2JC–P = 2.1 Hz, C2,6 or 3,5H, η6-p-cymene), 129.3 (s, C4H, dppm), 127.4–126.7 (m, C1, dppm (overlapping with signals from 4)), 126.0 (t, 1JC–P = 5.0 Hz, PCH2P, dppm), 99.1 (s, C1 or 4, η6-p-cymene), 86.3 (s, C1 or 4, η6-p-cymene), 28.9 (s, CH(CH3)2), 21.3 (s, CH(CH3)2), 16.9 (s, CCH3). 31P{1H} NMR (121.4 MHz, CDCl3, 298 K): δ −21.8 (s).

Synthesis of [OsCl(η6-p-cymene)(PPh3)(GeCl3)] (7)

One hundred milliliters of class="Chemical">dichloromethane was dried by passaclass="Chemical">n class="Chemical">ge through a plug of alumina in a Schlenk flask. Dichloromethane was then degassed prior to other manipulations. A total of 130.0 mg (0.198 mmol) of 3 was added to the Schlenk flask together with 51.0 mg (0.220 mmol, 1.1 equiv) of germanium(II) chloride·dioxane. The mixture was stirred at room temperature under a positive pressure of nitrogen for 1 h. The solvent was then removed in vacuo on the Schlenk line. Properties: sparingly soluble in CDCl3 and DMSO and unstable in DMSO. 132.9 mg (0.166 mmol) of yellow solid (84%). Melting point: 245 °C + dec. 1H NMR (300 MHz, CDCl3, 298 K): δ 7.68–7.65 and 7.42–7.40 (15H, both m, PPh3), 6.19 (1H, d, 3JH–H = 5.7 Hz, η6-C6H4), 5.84 (1H, d, 3JH–H = 5.6 Hz, η6-C6H4), 5.51 (1H, d, 3JH–H = 5.4 Hz, η6-C6H4), 5.02 (1H, dd, 3JH–H = 5.6 Hz,4JH–H = 1.4 Hz, η6-C6H4), 3.48 (0.2 7·OEt2), 2.54 (1H, sept, 3JH–H = 6.9 Hz, CH(CH3)2), 1.79 (3H, s, CCH3), 1.26 (3H, d, 3JH–H = 6.9 Hz, CH(CAH3CH3)), 1.22 (3H, d, 3JH–H = 7.0, CH(CH3CBH3)), 1.21 (0.2 7·OEt2).13C{1H} NMR (75.4 MHz, CDCl3, 298 K): δ133.4 (d, YJC–P = 9.9 Hz, C3,5 or 2,6H, PPh3), 132.6 (d, 1JC–P = 55.0 Hz, C1, PPh3), 129.7 (d, 4JC–P = 2.3 Hz, C4H,PPh3), 127.2 (d, YJC–P = 10.5 Hz, C2,6 or 3,5H, PPh3), 113.7 (d, 2JC–P = 1.7 Hz, C2 or 3 or 5 or 6H, η6-p-cymene), 93.2 (s, C1 or 4, η6-p-cymene), 84.9 (d, 2JC–P = 2.8 Hz, C2 or 3 or 5 or 6H, η6-p-cymene), 84.1 (d, 2JC–P = 2.2 Hz, C2 or 3 or 5 or 6H, η6-p-cymene), 80.2 (s, C1 or 4, η6-p-cymene), 79.7 (d, 2JC–P = 7.2 Hz, C2 or 3 or 5 or 6H, η6-p-cymene), 28.6 (s, CH(CH3)2), 21.3 (s, CH(CAH3CH3)), 20.9 (s, CH(CH3CBH3)), 16.0 (s, CCH3). 31P{1H} NMR (121.4 MHz, CDCl3, 298 K): δ −13.73 (s). FTIR (cm–1): 3064 (w), 2962 (w), 2869 (w), 1485 (w), 1433 (m), 1375 (w), 1261 (m), 1161 (w), 1090 (br s), 1017 (br m), 925 (w), 904 (w), 855 (vm), 797 (s), 747 (s), 695 (s). UV/Vis (dichloromethane): λmax = 361 nm. EIMS (70 eV): 802.0 ([OsCl(η6-p-cymene)(PPh3)GeCl3]+, <1%), 766.0 ([Os(η6-p-cymene)(PPh3)GeCl3]+, 7.9%), 623.1 ([OsCl(η6-p-cymene)(PPh3)]+, 3.5%), 585.1 ([Os(η6-p-cymene)(PPh3) – H]+, 3.4%), 262.1 ([PPh3]+, 100%), 152.1 ([C10H14 + H2O]+, 19%). ESI-MS: m/z calcd. for [M + Na]+, 824.9480; expt., 824.9465.

Synthesis of [OsCl(η6-p-cymene)(κ2dppm)]+GeCl3– (8)

Procedure 1: A total of 100.2 mg of the nonseparated mixture of 4 and 5 (0.129 mmol, 1 equiv) was dissolved in class="Chemical">dichloromethane (100 mL) which was previously dried aclass="Chemical">nd degassed, followed by the additioclass="Chemical">n of 34.8 mg (0.150 mmol, 1.2 equiv) of class="Chemical">n class="Chemical">germanium(II) chloride·dioxane. The mixture was stirred under a positive pressure of nitrogen for 1.5 h at room temperature. Upon completion, a bright yellow-green solution was filtered and the solvent was removed in vacuo. The resultant brown oil was washed with diethyl ether (3 × 10 mL), and the remaining yellow solid was dried in vacuo. A total of 98.0 mg (0.106 mmol) of “lemon chiffon” yellow solid was obtained (yield: 82%). Properties: sparingly soluble in CDCl3 and DMSO, somewhat stable in DMSO, air-stable. Melting point: 246 °C + dec. 1H NMR (300 MHz, CDCl3, 298 K): δ 7.62–7.28 (20H, m, phenyls, dppm), 6.26 (1H, m, PCHAHP, dppm), 6.20 (2H, d, 3JH–H = 5.9 Hz, C2,6 or 3,5H, η6-p-cymene), 6.14 (2H, d, 3JH–H = 5.8 Hz, C2,6 or 3,5H, η6-p-cymene), 4.69 (1H, dt, 2JX = 15.1 Hz, 2JX = 12.7 Hz, PCHHBP, dppm), 3.48 (0.09 8·OEt2), 2.45 (1H, sept, 3JH–H = 6.9 Hz, CH(CH3)2), 1.67 (3H, s, CCH3), 1.21 (0.09 8·OEt2), 1.11 (6H, d, 3JH–H = 6.9 Hz, CH(CH3)2). 31P{1H} NMR (121.4 MHz, CDCl3, 298 K): δ −39.9 (s). FTIR (cm–1): 3060 (w), 2963 (w), 2920 (w), 2869 (w), 1740 (w), 1484 (w), 1435 (m), 1260 (m), 1089 (br s), 1058 (br m), 1018 (br s), 869 (w), 801 (s), 739 (m), 729 (m), 683 (br s), 667 (m). UV/Vis λmax (dichloromethane): 359 nm. ESI-MS: m/z calcd. for [M]+, 745.1596; expt., 745.1575. Procedure 2: One hundred milliliters of dichloromethane was dried and degassed prior to addition of the reactants. A total of 88.0 mg (0.113 mmol, 1 equiv) of 4 was added to a Schlenk flask together with 39.8 mg (0.172 mmol, 1.5 equiv) of germanium(II) chloride·dioxane. The mixture was stirred at room temperature under a positive pressure of nitrogen for 1 h. The solvent was then evaporated in vacuo on the Schlenk line. Air-stable. A total of 83.1 mg (0.090 mmol) of yellow solid was obtained (80%). All the spectra were identical as for compound 8 from Procedure 1.

Synthesis of [OsCl(η6-C6H6)(PPh3)(GeCl3)] (9)

Complex 9 was synthesized following the same procedure as complexes 7. A total of 142.3 mg (0.237 mmol) of 6 was added to 30 mL of dried and class="Chemical">nitrogen-purclass="Chemical">n class="Chemical">ged dichloromethane. A total of 60.6 mg (0.262 mmol, 1.1 equiv) of germanium(II) chloride·dioxane was then added rapidly to the mixture. The solution was stirred at room temperature under a positive pressure of nitrogen for 1 h. The solvent was then evaporated in vacuo, affording the product. Air-stable. A total of 127.7 mg (0.171 mmol) of yellow solid was obtained (72%). Melting point: 265 °C + dec. 1H NMR (300 MHz, CDCl3, 298 K): δ 7.64–7.57 and 7.45–7.42 (15H, both m, PPh3), 5.71 (6H, s, η6-C6H6), 3.48 (0.5 9·OEt2), 1.21 (0.5 9·OEt2). 13C{1H} NMR (75.4 MHz, CDCl3, 298 K): δ 134.2 (d, XJC–P = 9.9 Hz, C2,6 or 3,5H, PPh3), 133.0 (d, 1JC–P = 56.6 Hz, C1, PPh3), 131.1 (d, 4JC–P = 2.7 Hz, C4H, PPh3), 128.5 (d, XJC–P = 11.0, C2,6 or 3,5H, PPh3), 85.2 (d, 2JC–P = 2.7 Hz, η6-C6H6), 65.9 (9·OEt2), 15.3 (9·OEt2). 31P{1H} NMR (121.4 MHz, CDCl3, 298 K): δ −12.16 (s). FTIR (cm–1): 3082 (w), 2978 (w), 2867 (w), 1482 (vw), 1433 (m), 1383 (w), 1251 (w), 1184 (w), 1100 (m), 1088 (m), 999 (w), 870 (w), 833 (m), 746 (s), 698 (s), 691 (s), 619 (w). UV/Vis (dichloromethane): λmax = 320 nm. ESI-MS: m/z calcd. for [M]+, 745.8962; expt., 745.1587; m/z calcd. for [M + 2ACN + H]+, 828.9565; expt., 829.1587.

Synthesis of [OsCl(η6-p-cymene)(κ2-dppm)]+SnCl3– (10)

One hundred milliliters of class="Chemical">dichloromethane was dried over aclass="Chemical">n class="Chemical">n class="Chemical">alumina column and degassed with nitrogen. After dissolving 150.3 mg of the mixture of intermediates 4 and 5 (0.193 mmol, 1 equiv) in dichloromethane, 40.3 mg (0.213 mmol, 1.1 equiv) of SnCl2 was added to the bright yellow solution. The mixture was thereafter refluxed under nitrogen for 3.5 h at 60 °C. Upon reaction completion, the murky golden solution was filtered and the filtrate was evaporated to dryness. The resulting brown oil was washed with diethyl ether (3 × 10 mL) and was subsequently dried in vacuo. A total of 152.7 mg (0.158 mmol) of a bright, shiny yellow crystalline solid was formed (yield: 82%). Properties: sparingly soluble in CDCl3, readily soluble and stable in DMSO, and air-stable. Melting point: 223 °C + dec. 1H NMR (300 MHz, CDCl3, 298 K): δ 7.64–7.38 (20H, m, dppm), 6.27 (1H, m, PCHAHP, dppm), 6.22 (2H, d, 3JH–H = 6.0 Hz, C2,6 or 3,5H, η6-p-cymene), 6.17 (2H, d, 3JH–H = 6.1 Hz, C2,6 or 3,5H, η6-p-cymene), 4.69 (1H, dt, 2JX = 15.1 Hz, 2JX = 12.6 Hz, PCHHBP, dppm), 3.48 (0.05 10·OEt2), 2.46 (1H, sept, 3JH–H = 6.6 Hz, CH(CH3)2), 1.68 (3H, s, CCH3), 1.21 (0.05 10·OEt2), 1.12 (6H, d, 3JH–H = 6.6 Hz, CH(CH3)2). 13C{1H} NMR (75.4 MHz, DMSO, 298 K): δ 132.3–129.0 (m, C1–6P, dppm), 112.7 (s, C1 or 4, η6-p-cymene), 96.3 (s, C1 or 4, η6-p-cymene), 84.8 (t, 2JC–P = 2.5 Hz, C2,6 or 3,5H, η6-p-cymene), 81.5 (t, 2JC–P = 3.3 Hz, C2,6 or 3,5H, η6-p-cymene), 43.5 (t, 1JC–P = 33.0 Hz, PCH2P, dppm), 29.9 (s, CH(CH3)2), 21.9 (s, CH(CH3)2), 16.6 (s, CCH3). 31P{1H} NMR (121.4 MHz, CDCl3, 298 K): δ −39.8 (s). FTIR (cm–1): 3062 (w), 2963 (w), 2922 (w), 2867 (w), 1685 (vw), 1559 (vw), 1483 (w), 1435 (m), 1374 (w), 1309 (w), 1260 (w), 1187 (w), 1156 (w), 1101 (m), 1057 (vw), 1025 (vw), 999 (vw), 869 (w), 800 (w), 775 (vw), 743 (vm), 728 (s), 712 (m), 703 (vm), 692 (s), 668 (m), 635 (w). UV/Vis λmax (dichloromethane): 296 nm (major), 269 nm (minor). ESI-MS: m/z calcd. for [M]+, 745.1596; expt., 745.2011.

Synthesis of [OsCl(η6-p-cymene)(PPh3)(SnCl3)] (11)

In a Schlenk flask, 50 mL of dry class="Chemical">dichloromethane was degassed prior to other maclass="Chemical">nipulatioclass="Chemical">ns. A total of 71.2 mg (0.108 mmol) of 3 was added to the Schleclass="Chemical">nk flask toclass="Chemical">n class="Chemical">gether with 31.1 mg (0.164 mmol, 1.5 equiv) of tin(II) chloride. The mixture was refluxed under a positive pressure of nitrogen for 17.5 h. The reaction yielded a bright, yellow liquid, which was subsequently filtered. The filtrate was evaporated on a rotary evaporator and washed with diethyl ether. Diethyl ether washings were discarded and the remaining residue was dried in vacuo on a Schlenk line. Properties: sparingly soluble in CDCl3 and soluble in DMSO. 73.2 mg (0.086 mmol) of bright orange crystals (80%). Melting point: 239 °C + dec. 1H NMR (300 MHz, CDCl3, 298 K): δ 7.68–7.61 and 7.46–7.41 (15H, m, PPh3), 6.20 (1H, d, 3JH–H = 5.7 Hz, η6-C6H4), 6.15 (1H, d, 3JH–H = 5.7 Hz, η6-C6H4), 5.42 (1H, d, 3JH–H = 5.7 Hz, η6-C6H4), 5.16 (1H, d, 3JH–H = 5.7 Hz, η6-C6H4), 3.48 (0.025 11·OEt2), 2.25 (1H, sept, 3JH–H = 7.0 Hz, CH(CH3)2), 1.94 (3H, s, CCH3), 1.21 (3H, d, 3JH–H = 6.9 Hz, CH(CH3CAH3)), 1.14 (3H, d, 3JH–H = 6.9 Hz, CH(CH3CBH3)). 13C{1H} NMR (75.4 MHz, DMSO, 298 K): δ 134.5 (d, 1JC–P = 56.0 Hz, C1, PPh3), 133.9 (d, XJC–P = 9.9 Hz, C2,6 or 3,5H, PPh3), 131.1 (d, 4JC–P = 2.2 Hz, C4H, PPh3), 128.7 (d, XJC–P = 10.5 Hz, C2,6 or 3,5H, PPh3), 109.1 (s, C1 or 4, η6-p-cymene), 97.1 (s, C1 or 4, η6-p-cymene), 83.1 (d, 2JC–P = 3.3 Hz, C2 or 3 or 5 or 6H, η6-p-cymene), 82.2 (d, 2JC–P = 2.2 Hz, C2 or 3 or 5 or 6H, η6-p-cymene), 81.0 (d, 2JC–P = 1.6 Hz, C2 or 3 or 5 or 6H, η6-p-cymene), 80.5 (d, 2JC–P = 5.0 Hz, C2 or 3 or 5 or 6H, η6-p-cymene), 29.5 (s, CH(CH3)2), 22.6 (s, CH(CAH3CH3)), 22.5 (s, CH(CH3CBH3), 17.6 (s, CCH3). 31P{1H} NMR (121.4 MHz, CDCl3, 298 K): δ −14.5 (s). FTIR (cm–1): 3803 (w), 3676 (w), 2163 (w), 1734 (w), 1684 (w), 1654 (w), 1559 (w), 1507 (w), 1457 (w), 1436 (m), 1374 (w), 1160 (w), 1091 (m), 1050 (w), 1030 (w), 999 (w), 870 (w), 799 (w), 746 (m), 694 (s). UV/Vis λmax (dichloromethane): 350 nm (major), 311 nm (minor). ESI-MS: m/z calcd. for [M – SnCl3]+, 623.1310; expt., 623.1658.

Density Functional Theory Calculations

The ground-state stclass="Chemical">ructures of all complexes were determiclass="Chemical">ned by declass="Chemical">nsity fuclass="Chemical">nctioclass="Chemical">nal theory (DFT) methods, with the resolutioclass="Chemical">n of ideclass="Chemical">ntity approximatioclass="Chemical">n[25] usiclass="Chemical">ng the TURBOMOLE[26] program packaclass="Chemical">n class="Chemical">ge and the user interface TmoleX 4.2.[27] The basis set def2-SV(P)[28] was used to pre-optimize the structures. The final optimized structures and orbital occupancies were then calculated using the def2-TZVPP[28] basis set for Sn, Ge, P, Cl, C, and H atoms, with Hay&Wadt ECP-46 for Sn (ECP, effective core potential).[29] The basis set LANL2DZ with Hay&Wadt ECP-60 was used for Os.[29] The hybrid functional B3LYP[30] was used for all calculations. The boundary surface representations of the HOMO and LUMO frontier orbitals are represented with an isosurface of 0.03. Color codes: Sn, purple; Ge, cyan; P, orange; Cl, green; C, olive; H, white; Os, dark blue.

Cancer Cell Lines

Short-term (24 h) in vitro anticlass="Disease">cancer activities of the syclass="Chemical">nthesized compouclass="Chemical">nds were studied usiclass="Chemical">ng Daltoclass="Chemical">n’s class="Chemical">n class="Disease">ascites lymphoma (DL) and Ehrlich ascites carcinoma (EAC) malignant cell lines. The cancer cells (DL) were cultured in RPMI-1640 supplemented with 10% FBS, gentamycin (20 mg/mL), streptomycin (100 mg/mL), and penicillin (100 IU) in a CO2 incubator at 37 °C with 5% CO2. The exponentially growing cells (80% confluency) were subcultured and used for testing the cytotoxicity and apoptosis potential in the said experiments. All the reagent solutions were autoclaved followed by sterilization by filtration through 0.22 μm membranes.

MTT-Based Cell Viability Assay

The in vitro growth inhiclass="Chemical">bitory effect of syclass="Chemical">nthesized compouclass="Chemical">nds (3–11) iclass="Chemical">n DL aclass="Chemical">nd EAC class="Chemical">n class="Disease">cancer cell lines was evaluated by the colorimetric MTT assay.[31] The in vitro short-term cytotoxicity (24 h) test was carried out at different concentrations (0 and 0.01 to 100 or 400 μM) prepared in PBS (PH: 7.4) in a 96-well cell culture plate (Thermo Scientific, cat. no. 265301). The above dose range was selected considering the permissive cytotoxicity (5–10%) in a normal cell line (PBMCs). The cytotoxicity results of potent complexes were compared with the positive reference drug cisplatin. To evaluate the toxic effect of the aforesaid compounds on normal cells, nontumorigenic peripheral blood mononuclear cells (PBMCs) were used under the same experimental conditions, compound dosing regimen, and culture times as mentioned above. The detail about the methodology is already described previously.[32] The IC50 profiles of all the potent compounds on DL, EAC, and PBMC cell lines were determined from dose–response curves using the nonlinear curve fit. For the IC50 plot, the x axis was considered as different doses (log concentration) and the y axis as the percentage of cell death (cytotoxicity).[33] For IC50 calculation in DL and EAC cell lines, higher concentration ranges from 0.01 to 100 μM were tested, whereas in normal cells (PBMCs), the tested concentration ranged from 0.01 to 400 μM. The nonlinear curve fit function used for IC50 calculations for the different compounds is as follows:where A1 refers to the bottom asymptote, A2 refers to the top asymptote, LOGx0 refers to the center, and p refers to the hill slope. The full dose–response curves are available in the Supporting Information.

Apoptosis Study

Apoptclass="Chemical">osis, or programmed cell death, is a class="Chemical">n class="Chemical">genetically regulated inherent property of metabolically active cells that eliminates severely damaged cells from the body. The tendency to avoid apoptosis leading to a drug resistance phenotype is a characteristic of most malignant cancers. In the present study, compound-mediated apoptotic cell death was determined by the acridine orange and ethidium bromide (AO/EB) dual staining method.[34] The principle of the assay is based on the fact that acridine orange is taken up by both viable and nonviable cells due to membrane permeability and emits green fluorescence, while ethidium bromide is taken up by apoptotic cells due to membrane damage and emits red fluorescence after intercalation with DNA within the nucleus.[35] Briefly, the cancer cells (1 × 106) were treated with different concentrations (0.01, 0.1, 0.5, 1, 5, and 10 μM) of synthesized compounds for 24 h in a 96-well plate (tissue culture grade, flat-bottom, sterile; Thermo Fisher Scientific, Waltham, Massachusetts, USA). After treatment, cells were washed twice with PBS (PH: 7.4) followed by staining with acridine orange and ethidium bromide (100 μg/mL each in PBS, pH 7.4) for 5 min in a dark cold room. The control and treated cells were examined in three replicates using a fluorescence microscope and photographed (Medlab Solutions Lx400 FLR Fluorescence Microscope). About 1000 cells were counted per group, and the percentage of apoptotic nuclei (orange/red) was calculated based on differential staining patterns.

Molecular Docking Simulation

The intermolecular interactions between the common template of the synthesized compound and anti-apoptotic tarclass="Chemical">get proteiclass="Chemical">ns (class="Chemical">n class="Gene">BCL-2 (PDB: 2O22), BCL-XL (PDB: 1R2D), and MCL-1 (PDB: 6QGD) were studied using Molegro Virtual Docker (Trial MVD 2010.4.0) software for Windows. The docking run was carried out using a GRID of 15 Å in radius and 0.30 in resolution with a number of runs of 10, maximum interactions of 1500, a maximum population size of 50, maximum steps of 300, a neighbor distance factor of 1.00, and a maximum number of poses returned of 5.[36] After molecular docking, the protein–ligand complex was further analyzed and visualized by Chimera software (https://www.cgl.ucsf.edu/chimera/) and BIOVIA Discovery Studio Visualization (https://www.3dsbiovia.com/products/collaborative-science/biovia-discovery-studio/). On the basis of the docking score, the obtained potent compounds were further subjected to pharmacokinetic and druglikeness property calculation using the SwissADME online tool (http://www.swissadme.ch/index.php).[37]

Statistical Analysis

The experimental results are expressed as mean values ± S.D. All measurements were replicated three times. The data were analyzed by an analysis of variance (An class="Chemical">NOVA) (*P ≤ 0.05).

Results and Discussions

Synthesis and Spectroscopic Properties of the Neutral Compounds 7, 9, and 11

The precursors [class="Chemical">OsCl2(η6-Cclass="Chemical">n class="Chemical">6H6)]2 (1) and [OsCl2(η6-p-cymene)]2 (2) were synthesized according to literature procedures via redox reactions using OsCl3·xH2O with 1,3-cyclohexadiene and α-terpinene, respectively.[23a,23b] After cleaving the dimers 1 and 2 with PPh3 (see the Supporting Information), the resulting isolated piano-stool intermediates 3 and 6 were reacted with a slight excess of GeCl2·(dioxane), yielding the desired σ-germyl complexes, via a facile GeCl2 insertion into the Os–Cl bonds, rac-7 and rac-9, respectively. Similarly, complex 3 was reacted with an excess of SnCl2 to produce the stannyl complex rac-11 (Scheme ). Osmium arene complexes bearing a σ-bound germyl or stannyl group are surprisingly rare in the literature, with only two reports by Wen and co-workers[21] and Castro and co-workers.[22] Detailed biological investigations of these complexes and elucidation of bonding by density functional theory (DFT) methods have not been evaluated and are reported for these complexes.

Scheme 3

Synthesis of the Neutral Compounds rac-7, rac-9, and rac-11

The class="Chemical">1H aclass="Chemical">nd class="Chemical">n class="Chemical">13C{1H} NMR spectra were especially instructive in revealing the structures of the final p-cymene-containing complexes 7 and 11 due to the emerging chirality around the osmium centers upon insertion of Ge or Sn. In fact, complexes 7, 9, and 11 exist as a racemic mixture of enantiomers rac-7, rac-9, and rac-11, but we denote only one of the mirror image stereoisomers for simplicity. The aromatic region of the p-cymene ligand in complexes 7 and 9 is represented by four sets of doublet resonances due to asymmetry at the Os center. The iPr moiety also has two sets of resonance signals corresponding to CH(CH3aCH3b) in the 1H NMR spectrum. This contrasts with the achiral starting material 3, where the aromatic protons are depicted by two sets of doublets and the one methyl resonance signal for iPr, which are equivalent and therefore display a single doublet resonance signal in the 1H NMR spectrum corresponding to CH(CH3)2. Similar observations are apparent in the 13C{1H} NMR spectra of 7 and 11, and the p-cymene ring carbon atoms are all inequivalent, exhibiting six sets of resonance signals, most of which are doublets due to 2JC–P coupling to the P atom of the phosphine ligand bound to the Os center. The carbons from the methyl groups from iPr are also split into two resonance signals, again in contrast with the 13C{1H} NMR spectrum of the achiral complex 3. The 31P{1H} NMR spectra of 7, 9, and 11 exhibit sharp singlet resonances and only show slight shifts compared to the starting materials 3 and 6.

Synthesis and Spectroscopic Properties of the Ionic Compounds 8 and 10

The precursor complex [class="Chemical">OsCl2(η6-class="Chemical">n class="Chemical">p-cymene)(κ1-dppm)] (4) was prepared in two different solvents: dimer 2 was reacted with an excess of dppm (1,1-bis(diphenylphosphino)methane) in (1) acetonitrile (a coordinating solvent) or (2) dichloromethane (a noncoordinating solvent). Strikingly, the nature of the solvent dictates the path that the reaction follows. The reaction in acetonitrile yielded a mixture of the desired 4 and a bidentate κ2-dppm species (5) (Scheme ), while the reaction in dichloromethane afforded a mixture of complex 4 and a dinuclear complex ([[OsCl2(η6-p-cymene)]2(μ-dppm)], (0) (Scheme ).

Scheme 4

Solvent-Dependent Pathway to [OsCl2(η6-p-cymene)(κ1-dppm)] (4)

It is likely that the mechanism of the reaction in class="Chemical">acetonitrile iclass="Chemical">nvolves coordiclass="Chemical">natioclass="Chemical">n thereof to the class="Chemical">n class="Chemical">osmium center of the formed complex 4, resulting in 4–NCCH3+Cl–, which then undergoes transformation to 5 with CH3CN elimination.[38]31P{1H} NMR spectroscopy of the crude reaction mixture in this case showed the clean formation of three resonance signals: two doublets at δ = −16.1 and −28.8 ppm corresponding to complex 4 and a singlet resonance signal for the salt complex 5 at δ = −40.0 ppm. Strikingly, the peaks of the ruthenium analogue of 4, [RuCl2(η6-p-cymene)(κ1-dppm)], can be found at δ = 25.6 and −28.1 ppm.[39] Fortunately, complexes 4 and 5 could be separated by fractional recrystallization using dichloromethane and diethyl ether. In an attempt to surmount the problem of two emerging products, dichloromethane was used instead as a solvent, hoping that the reaction would proceed cleanly to 4 from precursor 2. However, 31P{1H} NMR again indicated two sets of signals with two doublets (δ = −16.1 and −28.8 ppm) and a singlet at δ = −21.8 ppm, corresponding to 4 and a new product, [[OsCl2(η6-p-cymene)]2(μ-dppm)] (0), respectively. The peaks attributed to the suspected compound 0 were in good agreement with the 1H NMR spectrum of the corresponding known ruthenium dimer [[RuCl2(η6-p-cymene)]2(μ-dppm)].[39]

With the isolated 4 in hand, the class="Chemical">germanate salt 8 could be isolated upoclass="Chemical">n the reactioclass="Chemical">n of complex 4 with Lewis acidic class="Chemical">n class="Chemical">GeCl2·(dioxane) (Scheme ).

Scheme 5

Synthesis of the Ionic Complex 8 from the Isolated Intermediate 4

Subsequently, complex 8 could also be selectively synthesized using the mixture of complexes 4 and 5 as a starclass="Chemical">ting material upoclass="Chemical">n reactioclass="Chemical">n with class="Chemical">n class="Chemical">GeCl2·(dioxane), simplifying the synthesis considerably (Scheme ). Additionally, in close analogy, SnCl2 was reacted with the mixture of 4 and 5, affording the stannate complex 10 in high yields (Scheme ).

Scheme 6

Synthesis of the Ionic Complexes 8 and 10 from a Mixture of 4 and 5

In the class="Chemical">1H class="Chemical">n class="Chemical">NMR spectra of salt complexes 8 and 10, the chelated κ2 ligand exhibits two sets of resonance signals for the two diastereotopic protons on the dppm bridgehead carbon atom due to conformational locking. For both 8 and 10, these signals appeared as doublets of triplets due to 2JH–H coupling to the other diastereotopic proton and 2JH–P to both P atoms. Interestingly, the chemical shifts of the two protons are several ppm apart in the 1H NMR spectra, which was confirmed by H,H COSY experiments. The 31P{1H} NMR spectra of both salt complexes exhibited sharp singlet resonance signals, which slightly shifted downfield upon the reaction with GeCl2/SnCl2 (δ = −39.9 and −39.8 ppm, respectively). Finally, the cations of both salts were identified with high precision by ESI-MS.

All the final reactions could easily be monitored, as a clear chanclass="Chemical">ge iclass="Chemical">n color of the solutioclass="Chemical">ns could be class="Chemical">noticed upoclass="Chemical">n additioclass="Chemical">n of class="Chemical">n class="Chemical">GeCl2 or SnCl2. Final complexes (7–11) were synthesized successfully with high to excellent yields (72.2–83.8%), and their color ranged from light yellow to dark orange. UV/Vis measurements of compounds 8 and 9 indicated only one maximum peak at 358 and 320 nm, respectively. Intriguingly, two maximum absorptions were observed for compounds 7, 10, and 11.

FTIR spectclass="Chemical">roscopy was useful iclass="Chemical">n corroboraclass="Chemical">n class="Chemical">ting the presence of a para-substituted p-cymene ring and mono-substituted phenyl rings of PPh3 or dppm ligands. All the vibrations from C–H stretching and bending of alkanes and aromatics were additionally visible. Furthermore, the trichlorogermyl (in 7 and 9) and trichlorogermanate (in 8) groups’ bonds are hydrolytically robust in contact with air since the wavenumber for hydrolyzed Ge–OH vibration (3571 cm–1)[40] was absent from the spectra, nor were any hydrolysis products detected in the NMR spectra.

DMSO-d6 Stability Studies

Finally, compounds 7, 8, and 10 were tested for staclass="Chemical">bility over time iclass="Chemical">n dimethyl sulfoxide (class="Chemical">n class="Chemical">DMSO-d6).[41] The testing was performed by recording the 31P{1H} and 1H NMR spectra of the complexes dissolved in DMSO-d6 at various time intervals over a period of 6 h. Complex 10 is stable in DMSO-d6 for as long as 6 h. This was demonstrated by only one original peak in 31P{1H} NMR, which does not change over time. Immediately after dissolving compound 8 in DMSO-d6, some decomposition of the complex occurs, which is demonstrated by a small peak at δ = −23.5 ppm (see the Supporting Information, Figure S29). It appears that it reaches its maximum height after an hour. It is worth mentioning that no visible changes are observed in 1H NMR after 6 h (see the Supporting Information, Figure S30). The germyl complex 7 is somewhat less stable in DMSO-d6: again, two additional resonance signals arose in the 31P NMR spectrum after only 30 min at δ = −10.2 (doublet) and −26.2 (singlet) ppm, suggesting the formation of other adducts upon dissolution in DMSO-d6, possibly DMSO adducts (see the Supporting Information, Figure S31). Many studies on the instability of cisplatin and its derivatives (oxaliplatin and carboplatin) in the DMSO solvent have been published.[42] The ruthenium analogue of 7 was also noted to be DMSO-unstable, where the suggested adduct structures consisted of [(η6-arene)Ru(PR3)(DMSO)Cl]+GeCl3– and [(η6-arene)Ru(PR3)(DMSO)GeCl3]+Cl–[13a] and were linked to the loss of cytotoxic action previously (Scheme ).

Scheme 7

Potential Structures of the DMSO Adducts of 7

Coordination via the S atom of DMSO is also possible.

Coordination via the S atom of n class="Chemical">DMSO is also pclass="Chemical">n class="Chemical">ossible.

Single-Crystal X-ray Diffraction Investigations

Crystals suitable for X-ray diffraction analysis of 4, 8, 9, 10, and 11 were grown from class="Chemical">dichloromethane or, iclass="Chemical">n some cases, class="Chemical">n class="Chemical">dichloromethane/diethyl ether (see the Supporting Information for complete data).

The precursor compound 4 crystallizes in the monoclinic crystal system with the space group class="Chemical">P2(1)/class="Chemical">n aclass="Chemical">nd adopts a characteristic to such stclass="Chemical">n class="Chemical">ructures, “piano-stool” geometry, with all the angles being less than 90° (Figure ), except for the CH2 bridge of dppm (116.8° (2), Table ). It clearly reveals a pendant phosphorus atom, in accord with our spectral findings.

Figure 1

ORTEP representations of complex 4 at the 30% probability level. H atoms are omitted for clarity.

Table 1

Selected Bond Lengths and Angles of the Intermediate 4

selected bond lengths [Å]
Os(1)–P(1)	2.3562(12)
Os(1)–Cl(1)	2.4208(11)
Os(1)–Cl(2)	2.4265(10)
P(1)–C(20)	1.841(4)
P(2)–C(20)	1.854(4)

ORTEP representations of complex 4 at the 30% proban class="Chemical">bility level. class="Chemical">n class="Disease">H atoms are omitted for clarity.

The class="Chemical">germanate complex 8 crystallizes iclass="Chemical">n a moclass="Chemical">nocliclass="Chemical">nic system as well, with the Cc space group. The boclass="Chemical">nds aclass="Chemical">nd aclass="Chemical">ngles are observed to be similar as iclass="Chemical">n complex 10: the loclass="Chemical">nclass="Chemical">n class="Chemical">gest bond of the cation was between osmium and chlorine, while all the angles are less than 90° (Figure ), except for the CH2 dppm bridge and the angles of the anion (Tables and 4). The germanate anion is separated from the cation and is thus a separated ion pair. It exhibits trigonal pyramidal geometry in close similarity to its Ru analogue reported by us earlier.[14]

Figure 2

ORTEP representations of complex 8 at the 30% probability level. H atoms are omitted for clarity. The minor component of the disordered GeCl3– is also removed for clarity.

Table 2

Selected Bond Lengths and Angles of Complex 8

selected bond lengths [Å]
Os(1)–P(1)	2.3193(19)
Os(1)–P(2)	2.3336(16)
Os(1)–Cl(7)	2.3990(14)
Ge(1)–Cl(1)	2.294(5)

Table 4

Selected Bond Lengths and Angles of Complex 10

selected bond lengths [Å]
Os(1)–P(1)	2.3347(12)
Os(1)–P(2)	2.3162(13)
Os(1)–Cl(1)	2.4001(11)
Sn(1)–Cl(2)	2.4379(18)

ORTEP representations of complex 8 at the 30% probaclass="Chemical">bility level. class="Chemical">n class="Disease">H atoms are omitted for clarity. The minor component of the disordered GeCl3– is also removed for clarity.

Complex 9 also crystallizes in the monoclinic crystal system with the space group class="Chemical">P2(1)/class="Chemical">n. The solid-state stclass="Chemical">n class="Chemical">ructure reveals an η6 coordination of benzene to the osmium center and a distorted tetrahedral geometry around the Os center (Figure ). The Os–Ge bond length of 2.450 Å lies close to the sum of the single-bond covalent radii of Os and Ge (2.40 Å) and can thus be considered a single bond. Selected bond lengths and bond angles are reported in Table .

Figure 3

ORTEP representations of complex rac-9 at the 30% probability level. H atoms are omitted for clarity.

Table 3

Selected Bond Lengths and Angles of Complex 9

selected bond lengths [Å]
Os(1)–P(1)	2.354(3)
Os(1)–Cl(4)	2.412(3)
Os(1)–Ge(2)	2.450(12)

ORTEP representations of complex rac-9 at the 30% proban class="Chemical">bility level. class="Chemical">n class="Disease">H atoms are omitted for clarity.

Complex 10 was found to be in a monoclinic crystal system with the Cc space group. Similar to the precursor 4, all the class="Chemical">osmium-related aclass="Chemical">ngles are less thaclass="Chemical">n 90° (Figure ). As with the class="Chemical">n class="Chemical">germanium analogue, a large distance separates the trigonal pyramidal tin center in the anion from the osmium center and is thus a separated ion-pair salt.

Figure 4

ORTEP representations of complex 10 at the 30% probability level. H atoms are omitted for clarity. The minor component of the disordered SnCl3– is also removed for clarity.

ORTEP representations of complex 10 at the 30% probaclass="Chemical">bility level. class="Chemical">n class="Disease">H atoms are omitted for clarity. The minor component of the disordered SnCl3– is also removed for clarity.

The orthorhomclass="Chemical">bic crystal system with the space group Pca2(1) was adopted by the class="Chemical">neutral class="Chemical">n class="Chemical">stannyl compound 11. Similar to 9, the complex has a distorted tetrahedral geometry (Figure ) with the Os–Sn bond length of 2.5934(9) Å being slightly shorter than the sum of covalent single-bond radii of osmium and tin (2.69 Å) but can be considered a single bond based on these data (Table ).

Figure 5

ORTEP representations of complex 11 at the 30% probability level. H atoms are omitted for clarity.

Table 5

Selected Bond Lengths and Angles of Complex 11

selected bond lengths [Å]
Os(1)–P(1)	2.344(4)
Os(1)–Cl(4)	2.412(3)
Os(1)–Sn(1)	2.5934(9)

ORTEP representations of complex 11 at the 30% proban class="Chemical">bility level. class="Chemical">n class="Disease">H atoms are omitted for clarity.

Quantum Chemical DFT Calculations

Density functional theory calculations (B3LYP, basis set LAclass="Chemical">NL2DZ for class="Chemical">n class="Chemical">Os, and def2-TZVPP for Sn, Ge, Cl, P, C, and H) were employed to gain insights into the electronic nature of the complexes synthesized. The obtained geometry-optimized structure of complex 9 is in good agreement with the experimentally determined X-ray crystal structure (Table ), with the exception of the phenyl rings of PPh3, which have a slightly different orientation in the optimized structure compared to the crystal structure (see the Supporting Information, Section 11, for complete DFT data).

Table 6

Selected Bond Lengths and Bond Angles, Both from the X-ray Crystal Structure and Geometry Optimization, of Complex 9

	selected bond lengths (Å)
bonds	X-ray crystal structure	DFT calculation
Os(1)–P(1)	2.354(3)	2.393
Os(1)–Cl(4)	2.412(3)	2.439
Os(1)–Ge(2)	2.450(12)	2.504

class="Chemical">Geometry optimizatioclass="Chemical">n of the class="Chemical">n class="Chemical">benzene analogue of complex 11, 11–benzene, was also performed to evaluate the structural changes that could arise from the substitution of Ge for Sn. It appeared that both geometries are fairly similar, the only noticeable difference being the bond lengths of Os–E, which is longer in 11–benzene, as expected, due to the longer covalent radius of Sn vs Ge (Table ).

Table 7

Selected Bond Lengths and Bond Angles from the Geometry Optimizations of Complexes 9 and the Theoretical Complex 11–Benzene

	selected bond lengths (Å)
bonds	9 (E = Ge)	11–benzene (E = Sn)
Os–P	2.393	2.378
Os–Cl	2.439	2.445
Os–E	2.504	2.654

The frontier molecular orclass="Chemical">bitals of class="Chemical">n class="Chemical">9 and 11–benzene were also calculated to evaluate what changes in orbital energies and compositions could arise from the substitution of Ge by Sn in such complexes (Figure ). The frontier molecular orbitals are virtually the same in composition in 9 and 11–benzene, with a LUMO majorly localized on the ECl3 group and spreading through the Os center and the arene ring, with smaller contribution of the Cl bound to Os. The HOMO however shows greater contribution from the Cl bound to the Os center and is also delocalized around E and the arene ring. The PPh3 moiety does not seem to play any role in the composition of the HOMO and LUMO frontier orbitals in both 9 and 11–benzene.

Figure 6

Boundary surface representations of HOMO (bottom) and LUMO (top) for complexes 9 (left) and 11–benzene (right) and relative energies (HOMO (bottom) and LUMO (top) in eV).

Boundary surface representations of HOMO (bottom) and LUMO (top) for complexes 9 (left) and 11–n class="Chemical">benzene (right) aclass="Chemical">nd relative eclass="Chemical">nergies (HOMO (bottom) aclass="Chemical">nd LUMO (top) iclass="Chemical">n eV).

The Wiberg bond indices were also calculated for both complexes class="Chemical">9 and 11–class="Chemical">n class="Chemical">benzene, revealing values close to unity for the Os–E bond (for E = Ge, WBI = 0.792, and for E = Sn, WBI = 0.921). This corroborates the findings of the X-ray analyses, in which, on comparison to covalent radii, the bonds could be considered as single bonds for both the germyl and stannyl complexes.

In Vitro Cell Viability

The evaluation of cell proliferation is often used in medicinal chemistry research as well as the class="Disease">cytotoxicity screeclass="Chemical">niclass="Chemical">ng of other class="Chemical">n class="Chemical">biologically active compounds. The MTT assay is based on the ability of the treated cells to convert the water-soluble MTT (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide) into an insoluble formazan crystal by NADPH-dependent cellular oxidoreductase enzymes. The end product, formazan crystal, is then solubilized and the concentration is determined by the optical density at 570 nm.[43] This method is also used as an important parameter to estimate the cellular energy capacity of the cells.[44] The cytotoxic potential of the synthesized compounds was studied against DL and EAC cell lines at different concentrations. A concentration-dependent decrease in cell viability (increase in cytotoxicity) was observed in both cell lines (Figures and 8). Compound 9 has significantly induced more cytotoxicity in DL cells followed by compounds 8, 10, 11, 7, and 3. Interestingly, an almost similar pattern of cytotoxicity was also observed in the case of the EAC cell line. On the other hand, a negligible cytotoxic effect was observed in normal cells (PBMCs) as compared to DL and EAC cell lines (Figure ), suggesting that the complexes exhibit some selectivity toward the cancer cells. The IC50 (half-maximal inhibitory concentration) is a crucial pharmacodynamic index of drug efficacy. The dose–response relationship must be defined before this value can be measured, which is normally done by fitting monotonic sigmoidal models. The IC50 of drug candidates can be determined by constructing a dose–response curve and examining the effect of different concentrations on biological response that reduced to 50%. The obtained IC50 values for the potent compounds on the PBMC cell lines are presented in Table , and all dose–response curves are available in the Supporting Information.

Figure 7

Compound-mediated percent cell cytotoxicity in DL cells after treatment with different dosages. Cisplatin was used as a positive (reference drug) control. Data are mean ± S.D., n = 3. CP, cisplatin; CMP, compound. Only the concentrations up to 10 μM are shown; see the Supporting Information for the other data.

Figure 8

Compound-mediated percent cell cytotoxicity in EAC cells after treatment with different dosages. Cisplatin was used as a positive (reference drug) control. Data are mean ± S.D., n = 3. CP, cisplatin; CMP, compound. Only the concentrations up to 10 μM are shown; see the Supporting Information for the other data.

Figure 9

Compound-mediated percent cell cytotoxicity in normal cells (PBMCs) after treatment with different concentrations. Cisplatin was used as a positive (reference drug) control. Data are mean ± S.D., n = 3. CP, cisplatin; CMP, compound. Only the concentrations up to 10 μM are shown; see the Supporting Information for the other data.

Table 8

IC50 Values of All Compounds on DL, EAC, and PBMC Cell Linesa

	IC50 (μM)
cells	cisplatin	3	7	8	9	10	11
DL cells	0.19 ± 0.07	112 ± 3.4	93 ± 2.8	8.3 ± 0.05	1.7 ± 0.06	15.8 ± 0.05	63.7 ± 1.3
SI PBMC/DL	79.15	1.33	4.28	19.42	82.88	9.18	6.72
EAC cells	0.37 ± 0.10	122.8 ± 2.8	103.8 ± 0.82	18.2 ± 0.05	13.7 ± 0.03	21.4 ± 0.57	87.3 ± 1.6
SI PBMC/EAC	40.64	1.22	3.83	8.86	10.28	6.78	4.90
PBMCs	15.04 ± 0.05	149.9 ± 1.2	397.6 ± 3.5	161.2 ± 2.8	140.9 ± 1.5	145.1 ± 1.8	428.1 ± 3.1

Dose–response curves were used for the calculation of IC50. SI refers to the selectivity index computed by IC50 (PBMC)/IC50 (DL or EAC).

Compound-mediated percent cell class="Disease">cytotoxicity iclass="Chemical">n DL cells after treatmeclass="Chemical">nt with differeclass="Chemical">nt dclass="Chemical">n class="Chemical">osages. Cisplatin was used as a positive (reference drug) control. Data are mean ± S.D., n = 3. CP, cisplatin; CMP, compound. Only the concentrations up to 10 μM are shown; see the Supporting Information for the other data.

Compound-mediated percent cell class="Disease">cytotoxicity iclass="Chemical">n EAC cells after treatmeclass="Chemical">nt with differeclass="Chemical">nt dclass="Chemical">n class="Chemical">osages. Cisplatin was used as a positive (reference drug) control. Data are mean ± S.D., n = 3. CP, cisplatin; CMP, compound. Only the concentrations up to 10 μM are shown; see the Supporting Information for the other data.

Compound-mediated percent cell class="Disease">cytotoxicity iclass="Chemical">n class="Chemical">normal cells (PBMCs) after treatmeclass="Chemical">nt with differeclass="Chemical">nt coclass="Chemical">nceclass="Chemical">ntratioclass="Chemical">ns. class="Chemical">n class="Chemical">Cisplatin was used as a positive (reference drug) control. Data are mean ± S.D., n = 3. CP, cisplatin; CMP, compound. Only the concentrations up to 10 μM are shown; see the Supporting Information for the other data.

Dn class="Chemical">ose–respoclass="Chemical">nse curves were used for the calculatioclass="Chemical">n of IC50. SI refers to the selectivity iclass="Chemical">ndex computed by IC50 (PBMC)/IC50 (DL or EAC).

It is striking to note that although all complexes exhiclass="Chemical">bit less class="Chemical">n class="Disease">cytotoxicity compared to the positive control, the best performing complexes in this series are the germanium-containing neutral complex 9 and the germanate salt (8). A previous study from our group revealed that ruthenium germyl complexes exhibit almost no cytotoxicity against the cancer cell line A2780 and the healthy cell line HEK-293.[13a] This was attributed to rapid exchange kinetics, where coordination of water to the Ru center occurred rapidly, for which there was spectroscopic evidence. In the present case, the neutral germyl complex 9 exhibits much more potent cytotoxicity, which is likely due to the more inert osmium center. Also, it is worth pointing out that complex 9, the best performing complex in this series, exhibits a higher selectivity index (SI) compared to cisplatin (82.9 on the DL cell line vs 79.2 for cisplatin), where SI = IC50 (PBMC)/IC50 (DL).

Apoptosis Investigations

Apoptclass="Chemical">osis is a form of class="Chemical">n class="Chemical">genetically programmed cell death mechanism that regulates multicellular organisms’ growth by maintaining a balance between cell proliferation and cell death by eliminating physiologically redundant, physically damaged, and abnormal cells.[45] Resistance to apoptosis is a characteristic feature of all types of cancer; therefore, current research focusing on the genes and signals regulating apoptosis has played an important role in cancer research. Most chemotherapeutic drugs destroy tumor cells and restrain their normal cell proliferation rate primarily by inducing apoptosis.[46] In the present study, the acridine orange/ethidium bromide (AO/EB) dual staining method was used to evaluate compound-mediated apoptotic and viable cells in treated groups. Apoptosis-associated changes in treated cells caused an increase in membrane permeability of ethidium bromide and hence made cells appear red, whereas viable cells appeared green.[47] Treatment of DL and EAC cell lines with different compounds (3 and 7–11) led to the development of apoptotic features that included membrane blebbing, nuclear condensation, and apoptotic bodies (Figure ). Furthermore, findings from several other studies indicated that the majority of the synthesized compounds had anticancer properties as a result of the formation of stable DNA adducts.[48] It is worth noting that cancer cells have a higher oxidative status than normal cells, which means that they may undergo more oxidative DNA damage than the normal cells. The compounds 8 and 9 have shown a stronger apoptosis-inducing ability than the rest of the tested compounds in the DL cell line (Figure ). Almost similar results were also obtained in the EAC cell line (Figure ). Thus, based on the present cytotoxicity and apoptosis assays, it is evident that compounds 8 and 9 possess promising anticancer activity that needs further study to establish the molecular mode of action.

Figure 10

Morphological features of viable and apoptotic cells observed under a fluorescence microscope. Treated cells showing an apoptotic nucleus with membrane damage and chromatin condensation. Cisplatin was used as a reference drug.

Figure 11

Percentage of apoptotic cells after treatment with different compounds in the DL cell line. Data are mean ± S.D., n = 3, one-way ANOVA; *P ≤ 0.05 as compared to compounds 3, 7, 8, 10, and 11, whereas #P ≤ 0.05 as compared to compounds 3, 7, 10, and 11.

Figure 12

Percentage apoptotic cells after treatment with different compounds in the EAC cell line. Data are mean ± S.D., n = 3, one-way ANOVA; *P ≤ 0.05 as compared to compounds 3, 7, 8, 10, and 11, whereas #P ≤ 0.05 as compared to compounds 3, 7, 10, and 11.

Morphological features of viable and apoptotic cells observed under a fluorescence micclass="Chemical">roscope. Treated cells showiclass="Chemical">ng aclass="Chemical">n apoptotic class="Chemical">nucleus with membraclass="Chemical">ne damaclass="Chemical">n class="Chemical">ge and chromatin condensation. Cisplatin was used as a reference drug.

Percentaclass="Chemical">ge of apoptotic cells after treatmeclass="Chemical">nt with differeclass="Chemical">nt compouclass="Chemical">nds iclass="Chemical">n the DL cell liclass="Chemical">ne. Data are meaclass="Chemical">n ± S.D., class="Chemical">n = 3, oclass="Chemical">ne-way Aclass="Chemical">n class="Chemical">NOVA; *P ≤ 0.05 as compared to compounds 3, 7, 8, 10, and 11, whereas #P ≤ 0.05 as compared to compounds 3, 7, 10, and 11.

Percentaclass="Chemical">ge apoptotic cells after treatmeclass="Chemical">nt with differeclass="Chemical">nt compouclass="Chemical">nds iclass="Chemical">n the EAC cell liclass="Chemical">ne. Data are meaclass="Chemical">n ± S.D., class="Chemical">n = 3, oclass="Chemical">ne-way Aclass="Chemical">n class="Chemical">NOVA; *P ≤ 0.05 as compared to compounds 3, 7, 8, 10, and 11, whereas #P ≤ 0.05 as compared to compounds 3, 7, 10, and 11.

Molecular Docking Simulations

Molecular docking has become an important class="Chemical">bioiclass="Chemical">nformatics tool iclass="Chemical">n the mediciclass="Chemical">nal chemistry field to elucidate fuclass="Chemical">ndameclass="Chemical">ntal class="Chemical">n class="Chemical">biochemical processes by analyzing the interacting behavior of molecules in the active site of a receptor.[49] The results of molecular docking further corroborate the findings obtained by MTT and apoptosis assays. Anti-apoptotic proteins, mostly BCL-2, BCL-XL, MCL-1, and many others, are well known to interact with pro-apoptotic proteins to execute apoptosis events.[50] Therefore, molecular docking was performed with the above-mentioned anti-apoptotic target proteins to further dissect the possible molecular mode of action of synthesized compounds. The high expression of BCL-2, BCL-XL, and MCL-1 proteins is reported in multiple cancer types that ultimately induced malignant phenotypes by skipping apoptosis. Moreover, their inhibition by various chemotherapeutic agents is associated with the apoptotic induction, and hence, regression in tumor cell proliferation occurs. Our finding (Figures and 14) showed that complex 9 has high affinity with BCL-2 family proteins. The π–π stacked and π–alkyl interactions have been found to be greatly involved with the active site amino acids of all the receptors (BCL-2, BCL-XL, and MCL-1). Compound 9 showed one π–π stacked (Tyr199) and two π–alkyl (Ala97 and Val145) interactions with BLC-2, two π–π stacked (Tyr195) and two π–alkyl (Ala93 and Val141) interactions with BCL-XL, and one π–π stacked (Phe270) and four π–alkyl (Val253, Met231, Arg263, and Leu267) interactions with MCL-1. Thus, based on the docking results, it can be suggested that compound 9 possesses an apoptotic-inducing ability due to efficient interaction with anti-apoptotic target proteins. Interestingly, the docking score of complex 9 is found to be comparable with the reference ligands (PDB ID) of the respective receptors (Figure ). Although these in silico studies provide useful insights into the potential mechanism at play for the germyl complex 9 and mirror the findings of the MTT assay and apoptosis studies, additional experimental binding studies are required to confirm this. In addition, in our study, we used the neutral form of complex 9, whereas dissociation into ionic species is very likely under biological conditions (see Section ).

Figure 13

Docking structures of 9 with (a) BCL-2, (b) BCL-XL, and (c) MCL-1 receptors are shown. Chemical interactions are shown along with ligand atoms and interacting amino acids in the inhibitor binding sites of different receptors.

Figure 14

Binding affinities of 9 with BCL-2, BCL-XL, and MCL-1 receptors. Data are mean ± S.D., n = 5.

Docking stclass="Chemical">ructures of 9 with (a) class="Chemical">n class="Gene">BCL-2, (b) BCL-XL, and (c) MCL-1 receptors are shown. Chemical interactions are shown along with ligand atoms and interacting amino acids in the inhibitor binding sites of different receptors.

class="Chemical">Biclass="Chemical">ndiclass="Chemical">ng afficlass="Chemical">nities of 9 with class="Chemical">n class="Gene">BCL-2, BCL-XL, and MCL-1 receptors. Data are mean ± S.D., n = 5.

To be successful as a dclass="Chemical">rug caclass="Chemical">ndidate, a poteclass="Chemical">nt molecule must reach its tarclass="Chemical">n class="Chemical">get in an adequate concentration and remain there in a bioactive form long enough for the expected biologic events to occur. The drug development process involves assessment of absorption, distribution, metabolism, and excretion (ADME) at a stage when considered compounds are numerous, but access to the physical samples is limited.[37] In that context, we performed the pharmacokinetic and druglikeness study of the compounds. Pharmacokinetics is the study of chemical metabolism and the discovery of a chemical’s fate from the time it is delivered until it is totally removed from the body. All the tested compounds showed high GI absorption and no blood–brain barrier (BBB) permeant (Table ). Druglikeness is a qualitative concept used in drug design that describes how drug-like substances behave in terms of bioavailability. It is estimated from the molecular descriptors before the substance is even synthesized and tested for bioactivities. Druglikeness molecular properties of synthesized compounds were tested and found to pass Lipinski, Ghose, Veber, Egan, and Muegge experiments (Table ), considering the permissive limit. Pharmacokinetic and druglikeness properties of compounds 3 and 7–11 revealed that compounds 8–10 possess strong pharmacokinetic and druglikeness properties (Table ) that need to be further explored for better understanding of their molecular mode of action. This is also in accord with the in vitro apoptosis and docking study.

Table 9

Pharmacokinetic and Druglikeness Properties of Synthesized Compounds (3 and 7–11)

	compounds
pharmacokinetic properties	3	7	8	9	10	11
GI absorption	high	high	high	high	high	high
BBB permeant	no	no	no	no	no	no
P-gp substrate	yes	yes	yes	yes	yes	yes
CYP1A2 inhibitor	no	no	no	no	no	no
CYP2C19 inhibitor	no	no	no	no	no	no
CYP2D6 inhibitor	yes	no	no	no	no	yes
CYP3A4 inhibitor	no	yes	no	no	yes	yes
water solubility	no	yes	yes	yes	yes	no

Conclusions

The synthesis of class="Chemical">osmium-class="Chemical">n class="Chemical">based germyl or stannyl/germanate and stannate complexes characterized by FTIR, 1H, 13C{1H}, and 31P{1H} NMR, MP, UV/Vis, ESI-MS, and X-ray diffraction analysis was reported and subjected to anticancer studies. Complexes 3 and 7–11 showed moderate cytotoxic activity on two cell lines in vitro compared to cisplatin. Strikingly, the germyl complex 9 exhibited the most promising cytotoxic activity in this series of compounds in contrast to the ruthenium analogues, which were completely inactive, which was potentially due to enhanced stability toward aquation compared to the more labile Ru analogues. Compounds 8 and 9 significantly (P ≤ 0.05) induced cell cytotoxicity in DL and EAC cell lines by inducing apoptotic cell death with negligible cytotoxicity in healthy PBMCs. The molecular mode of action might involve anti-apoptotic proteins as confirmed by our docking studies, but additional biological investigations and experimental studies need to be conducted to confirm this, along with other studies on other biochemical targets.