Synthesis, characterization, antiproliferative, and antimicrobial activity of osmium(II) half-sandwich complexes

Journal of Coordination Chemistry ISSN: 0095-8972 (Print) 1029-0389 (Online) Journal homepage: http://www.tandfonline.com/loi/gcoo20 Synthesis, characterization, antiproliferative, and antimicrobial activity of osmium(II) half-sandwich complexes Joel M. Gichumbi, Holger B. Friedrich, Bernard Omondi, Kovashnee Naicker, Moganavelli Singh & Hafizah Y. Chenia To cite this article: Joel M. Gichumbi, Holger B. Friedrich, Bernard Omondi, Kovashnee Naicker, Moganavelli Singh & Hafizah Y. Chenia (2018) Synthesis, characterization, antiproliferative, and antimicrobial activity of osmium(II) half-sandwich complexes, Journal of Coordination Chemistry, 71:2, 342-354, DOI: 10.1080/00958972.2018.1434164 To link to this article: https://doi.org/10.1080/00958972.2018.1434164 Accepted author version posted online: 30 Jan 2018. Published online: 15 Feb 2018. Submit your article to this journal Article views: 79 View related articles View Crossmark data Full Terms & Conditions of access and use can be found at http://www.tandfonline.com/action/journalInformation?journalCode=gcoo20 Journal of Coordination Chemistry, 2018 Vol. 71, no. 2, 342–354 https://doi.org/10.1080/00958972.2018.1434164 Synthesis, characterization, antiproliferative, and antimicrobial activity of osmium(II) half-sandwich complexes Joel M. Gichumbia, Holger B. Friedricha, Bernard Omondia, Kovashnee Naickerb, Moganavelli Singhb and Hafizah Y. Cheniab aschool of Chemistry and Physics, university of KwaZulu-natal, durban, south africa; bschool of life sciences, university of KwaZulu-natal, durban, south africa ABSTRACT ARTICLE HISTORY New osmium(II)-arene complexes [(η6-C H )OsCl(C H N-2-CH=N- received 1 september 2017 6 6 5 4 C H X)](PF ) (X = p-flouro (1), p-chloro (2), p-methyl (3)) were accepted 20 January 2018 6 5 6 synthesized by reaction of the corresponding bidentate N,N′- KEYWORDS ligands with the osmium-arene precursor [(η6-C H )Os(μ-Cl)Cl] in 6 6 2 os(ii) arene complexes; a 2:1 ratio. These complexes were characterized by UV–Vis, IR, 1H, n,n′-Bidentate ligands; 13C NMR spectroscopy, and elemental analysis and, for compound mtt assay; Cytotoxicity; 1, a single crystal X-ray structure was also obtained. The complexes antimicrobial susceptibility were investigated for antiproliferative activity in vitro against MCF-7 (human breast adenocarcinoma), Caco-2 (human epithelial colorectal adenocarcinoma), and HepG2 (human hepatocellular carcinoma) tumor cell lines. To test their selectivity, the non-tumor HEK293 (human embryonic kidney) cell line was used. The compounds were selective toward the tumor cell lines when compared to the known anticancer drug 5-fluorouracil which displayed low selectivity. The compounds were substantially more active against Caco-2 than 5-fluorouracil. They also showed moderate activity against the other two tumor cell lines. In addition, the antimicrobial activity of complex 2 was explored against a panel of antimicrobial-susceptible and -resistant Gram-negative and Gram-positive bacteria. Complex 2 showed anti-mycobacterial activity against Mycobacterium smegmatis and bactericidal activity against drug-resistant Enterococcus faecalis and methicillin-resistant Staphylococcus aureus ATCC 43300. CONTACT holger B. friedrich friedric@ukzn.ac.za © 2018 informa uK limited, trading as taylor & francis Group JOURNAL OF COORDINATION CHEMISTRY 343 1. Introduction Cancer causes many deaths in both developing and developed countries. There is a steady increase in the number of people diagnosed with cancer and statistics show that every year one in four deaths in Europe and the USA are due to cancer [1]. Chemotherapy is the main therapeutic approach for the treatment of localized and metastasized cancers. The clinical success of platinum-based anticancer drugs has made them indispensable in cancer chemo- therapy. However, platinum-based drugs have serious side-effects, which include severe toxicity, drug-resistance, and lack of selectivity [2, 3]. Thus, it is important to develop novel metal-based anticancer compounds with increased selectivity and therapeutic index, thus overcoming the negative effects and resistance phenomena which are amongst the major limitations for the curative treatment of cancer [1]. To overcome the limitations of platinum-based therapies, compounds containing e.g. titanium, gold, and ruthenium have been synthesized. Some of these have already entered clinical trials [4]. Osmium has received less attention, when compared to ruthenium and iron, among the organometallic complexes within the group 8 transition metals. Thus, less is known about the chemistry of osmium as compared to ruthenium and iron. Especially, studies on biologically active osmium complexes are rare [5]. This may be because osmium is considered a relatively inert third-row transition metal when compared to ruthenium, which is a second row transition metal. However, osmium offers properties that are distinct from ruthenium. These include stronger π-back donation from lower oxidation states, slower ligand exchange kinetics, the preference for higher oxidation states, and stronger spin-orbit coupling. Hence, osmium compounds may be seen as potential alternatives to rutheni- um-containing anticancer agents, since they are relatively inert and have sufficient stability under physiological conditions [6]. Osmium-arene complexes are very important in osmium chemistry [7]. They have recently attracted attention for a wide range of applications, including as building blocks for supra- molecular structures [8], a range of metathesis reactions, and they have also been shown to activate C–H bonds [9]. They also are useful catalysts for the cyclopropanation of olefins [10, 11] and have found applications as catalysts in transfer hydrogenation reactions [12–14]. In addition, the in vivo and in vitro cytotoxicity of η6-arene-osmium(II) complexes has recently triggered research activity targeting the design of organometallic agents [15, 16]. Various authors have reported the synthesis of osmium complexes containing bidentate heterocyclic N,N′-ligands, such as complexes of phenanthroline, bipyridine, and phenyla- zo-pyridine, which have been reported to show anticancer activity [15, 16]. Sadler and cow- orkers studied the binding mode adopted by piconilamide derivatives of Os(II) and Ru(II) half-sandwich compounds which showed contrasting cancer cell cytotoxicity between the Os(II) and Ru(II) compounds. They also reported that the compounds hydrolyzed rapidly and that they exhibited significant activity against cancer cell lines [9]. They further reported the study of osmium(II) half-sandwich chloro compounds and the influence of chelated ligands on hydrolysis, guanine binding, and cytotoxicity. The chelating ligands were found to con- tribute importantly to increasing aqueous stability [17]. A wide range of electronic and steric modifications are possible for complexes of transition metal compounds. This offers the potential rational design of therapeutic agents, which could have novel action mechanisms. Some chemical reactions and the anticancer activities of the organometallic Os(II) complexes [Os(η6-arene)X(AB)](PF ) (arene = biphenyl or 6 344 J. M. GICHUMBI ET AL. p-cymene; AB = N,N′-chelated phenyl-imino pyridine or phenyl azopyridine derivatives and X = Cl or I) have been reported. These complexes were potently active toward cancer cells and they underwent aquation, bound to the nucleobase guanine, and also oxidized the coenzyme nicotine adenine dinucleotide (NADH) [15]. The application of arene complexes of Ru(II) and Os(II) with indolo[3,2-c]-quinolines and paullones was studied by Filak and coworkers [18–21]. They tested the complexes for antiproliferative activity in vitro on CH1 (ovarian) human cancer cell lines, A549 (non-small cell lung), and SW480 (colon). The com- pounds showed IC values between submicromolar or low micromolar ranges. 50 Medicinal applications of osmium are significantly less common, when compared to ruthenium and iron in the same group. This is because osmium is generally not believed to be biologically relevant and in some forms, it can be very toxic. However, some osmium compounds have been applied in the treatment of rheumatoid arthritis in animals and trypanocidal activity against T.b. brucei has been reported [22]. Due to the growing interest in cytotoxic applications of half-sandwich Os(II)-arene com- plexes, we became interested in the preparation and characterization of new half-sandwich complexes of the type [(η6-C H )OsCl(C H N-2-CH=N-C H X)](PF ) (X = p-flouro (1), p-chloro 6 6 5 4 6 5 6 (2), p-methyl (3)). The complexes were investigated for their antiproliferative activity in vitro on Caco-2 (human epithelial colorectal adenocarcinoma), HepG2 (human hepatocellular carcinoma), and MCF-7 (human breast adenocarcinoma) tumor cell lines. To test their selec- tivity, the non-tumor HEK293 (human embryonic kidney) cell lines were used and the known anticancer drug 5-flourouracil, which is commonly used to treat these cancers, was used as control. Additionally, antimicrobial susceptibility tests were done against antimicrobial-re- sistant and antimicrobial-susceptible Gram-negative and Gram-positive bacteria to deter- mine their antibacterial potential. 2. Experimental 2.1. General procedures The organometallic complexes were prepared under nitrogen with Schlenk line techniques. Ether and hexane were used immediately after drying over sodium/benzophenone, while MeCN and acetone were dried over anhydrous CaCl . All other reagents were used without 2 purification. IR spectra were obtained with an ATR Perkin-Elmer Spectrum 100 spectrometer. Melting points were obtained using an Ernst Leitz Wetzlar hot-stage microscope, while 1H and 13C NMR spectra were recorded in DMSO-d (Sigma Aldrich) using a Bruker 400 MHz 6 spectrometer. Electronic spectra were obtained from acetonitrile solutions using a Perkin- Elmer Lambda 35 UV–vis spectrometer. A ThermoScientific Flash 2000 instrument was used for elemental analyses. Mass spectra (Waters Micromass LCT Premier TOF-MS) were obtained by direct infusion. The precursor [(η6-C H )Os(μ-Cl)Cl] was synthesized via a reported 6 6 2 method [5], as were the pyridine–imine ligands [23]. 2.2. In vitro anticancer activity 2.2.1. Materials and method Eagle’s Minimum Essential Medium (EMEM) with L-glutamine (4.5 g L−1), trypsin–versene mixture, and antibiotic were obtained from Lonza BioWhittaker, while phosphate buffered JOURNAL OF COORDINATION CHEMISTRY 345 saline (PBS) tablets, MTT reagent [3-(4,5-dimethyl-2-thiazolyl)-2,5-diphenyl-2H-tetrazolium bromide] and DMSO were obtained from Merck. Fetal bovine serum (FBS) was obtained from HyClone. All tissue culture consumables were purchased from Corning. The other reagents used were analytical grade and also only ultrapure deionized 18 MΩ water (Milli-Q50) was used. 2.2.2. Cell culture Originally, the HEK293 (human embryonic kidney) cells were purchased from the Anti-viral Gene Therapy Unit at the University of the Witwatersrand Medical School. The Caco-2 (human colorectal adenocarcinoma) and the human hepatocellular carcinoma (HepG2) were obtained from Highveld Biologicals (Pty) Ltd. Human breast adenocarcinoma (MCF7) cells were purchased from the ATCC, Manassas, VA, USA. The MCF-7 (human breast adenocarci- noma), HEK293 (human embryonic kidney), Caco-2 (human epithelial colorectal adenocar- cinoma), and HepG2 (human hepatocellular carcinoma) cell lines were kept at 37 °C under 5% CO in culture flasks that contained 5 mL of complete medium [EMEM supplemented 2 with 10% (v/ ) γ-irradiated FBS and antibiotic (0.25 μg mL−1 amphotericin B, 100 μg mL−1 v penicillin, and 100 μg mL−1 streptomycin)]. 2.2.3. Cytotoxicity test 2.2.3.1. MTT assay. The cells were grown to semi-confluency in EMEM which was supplemented with antibiotics 10% and fetal bovine serum in tissue culture flasks (25 cm3). They were seeded in a 96-well plate which contained 100 μL of medium at a density of 1.8 × 103 cells per well. Thereafter, these cells were incubated for 24 h at 37 °C under 5% CO . 2 Subsequently the medium was removed and fresh medium (100 μL) was added. Compounds 1, 2, and 3, at concentrations of 25, 50, 75, 100, 125, 150 μg mL−1, were subsequently added to the cells in triplicate. They were then incubated at 37 °C (48 h). Cells that did not receive these compounds were used as the positive controls (100% viability). The MTT assay method used was adapted from the reported protocol [24]. Thus, reduction of MTT to formazan with the succinate–tetrazolium reductase system was used to measure the metabolic activity of the cells. After 48 h incubation, fresh medium (100 μL) and 100 μL MTT (5 mg mL−1 in PBS) was used to replace the spent medium. The cells were then incubated at 37 °C for a further 4 h. Then, the medium and MTT were removed, and the formazan salt dissolved in 200 μL of DMSO was added to each well. The resultant purple solution’s absorb- ance was determined using a microplate reader (Mindray 96A, Vacutec) at wavelengths of 570 nm (detection λ) and 630 nm (reference λ for nonspecific signals). The cell viability (%) was correlated directly to the absorbance. The calculation method for the comparison to the untreated control was: [(OD Treated−OD Treated)∕(OD Control−OD Control)]×100. 570 630 570 630 All tests were carried out in triplicate. Calculations of the (IC ) concentrations (where 50% 50 cell death occurred) were carried out with Microsoft Excel 2010™. 2.2.4. Assessment of antibacterial activity using the disk diffusion assay The antibacterial activities of the three synthesized compounds and the ligand were deter- mined via disk diffusion [25]. Stock solutions of the complex or ligand (20 mg) were made up in DMSO (1 mL). The sterile blank disks (6 mm) were obtained from MAST (UK) and were 346 J. M. GICHUMBI ET AL. respectively impregnated with 10 μL (0.2 mg), 20 μL (0.4 mg), and 40 μL (0.8 mg) of these stock solutions. Then they were left to dry (1 h). Four Gram-negative (Klebsiella pneumoniae ATCC 700603, Escherichia coli ATCC 35218, Pseudomonas aeruginosa ATCC 27853, and E. coli ATCC 25922) and six Gram-positive (Bacillus subtilis ATCC 6633, Enterococcus faecalis ATCC 51299, Mycobacterium smegmatis mc2 155, Staphylococcus aureus ATCC 29213, S. aureus ATCC 43300, and Staphylococcus saprophyticus ATCC 35552) microbial strains, which were grown on TSA agar plates overnight, were subsequently suspended in sterile distilled water. Thereafter, the turbidity of the cell suspensions was adjusted to be equivalent to that of a 0.5 McFarland standard. Mueller–Hinton (MH) agar plates were inoculated with these sus- pensions by streaking swabs over the entire agar surface, then the complexes were applied to the prepared disks. The plates were incubated at 37 °C for 18 h. M. smegmatis mc2 155 plates were incubated at 37 °C for 48 h. Tests were carried out in duplicate and the results averaged. Tetracycline (TE30) and Ampicillin (AMP 10) disks (Oxoid, UK) were used as standard antibacterial agent controls. Disks impregnated with DMSO were used as the negative con- trols. The zone diameter criteria used to assign resistance or susceptibility to the tested compounds were: Resistant (R) ≤ 10 mm, Intermediate (I) = 11–14 mm and Susceptible (S) ≥ 15 mm. The criteria used for assigning resistance or susceptibility to AMP10 were: (R) ≤ 13 mm, (I) = 14-16 mm, and (S) ≥ 17 mm, and the criteria for TE30 were: (R) ≤ 14 mm, (I) = 15–18 mm, and (S) ≥ 19 mm [24]. 2.3. Preparation and characterization of complexes 1–3 To a solution of [(η6-C H )Os(μ-Cl)Cl] (100 mg, 0.15 mmol) in methanol (30 mL) at 40 °C, the 6 6 2 pyridine–imine ligand (0.32 mmol) in methanol (10 mL) was added dropwise. The color of the solution changed from yellow to red immediately. The mixture was stirred at 40 °C for 2 h. Then the volume was reduced to ~10 mL by evaporation of methanol on a rotary evap- orator. Thereafter, NH PF (0.33 mmol) dissolved in ethanol (10 mL) was added and the solu- 4 6 tion stood at 0 °C overnight. The precipitate which had formed was collected by filtration, washed with cold ethanol and diethyl ether, and then dried under vacuum. 1: Red solid. Yield 70%. m.p. 204 °C 1H NMR, δ (ppm): 9.59 (d, J = 5.48 Hz, 1H, py), (dec). HH 9.28 (s, 1H, CH=N), 8.39 (d, J = 7.32 Hz, 1H, py), 8.269 (m, 1H, py), 7.97 (m, 2H, py), 7.85 (m, HH J = 8.6 Hz, 1H, py), 6.09 (s, 6H, C H ). 13C NMR, δ (ppm): 169.40 (CH=N), 156.10 (py), 155.8 HH 6 6 (py), 148.04 (py), 140.43 (py), 130.19 (py), 129.95 (py), 125.31 (py), 125.22 (Ar), 116.46 (Ar), 116.23 (Ar), 78.99 (C H ). IR (KBr, cm−1): 1612.5 v(–CH=N), 827.1 v(P-F). Anal. Calcd (%): C, 6 6 33.31; H, 2.33; N, 4.32. Found: C, 33.70; H, 2.47; N, 4.42. MS (ESI, M/Z): 505.0533 [C H ClN F Os]+. 18 15 2 1 Crystals of compound 1 were grown by layering a fourfold volume of hexane on a solution of compound 1 in dry acetone and standing it for 2 days in the dark at room temperature. Data reduction was done using SAINT+ [26], while the structure was solved by direct methods with SHELXS [27] and refined using SHELXL [26]. The crystal and structure refinement data for compound 1 are shown in Table 1. 2: Red solid. Yield 84%. m.p. 200 °C . 1H NMR, δ (ppm): 9.60 (d, J = 5.32 Hz, 1H, py), (dec.) HH 9.30 (s, 1H, CH=N), 8.40 (d, J = 7.56 Hz, 1H, py), 8.30 (m, 1H, py), 7.86 (m, 1H, Py), 7.77 (d, HH J = 8.56 Hz, 2H, Ar), 7.69 (d, J = 8.76 Hz, 2H, Ar), 6.11 (s, 6H, C H ). 13C NMR, δ (ppm): 170.28 HH HH 6 6 (CH=N), 156.63 (py), 156.31, 150.87 (py), 140.94 (py), 134.66 (py), 130.56 (py), 130.01 (Ar), 125.41 (Ar), 79.58 (C H ). IR (KBr, cm−1): 1611.8 v(–CH=N), 822.5 v(P-F). Anal. Calcd (%): C, 6 6 32.49; H, 2.27; N, 4.21. Found: C, 32.46; H, 2.17; N, 3.82. MS (ESI, M/Z): 521.0237 [C H Cl N Os]+. 18 15 2 2 JOURNAL OF COORDINATION CHEMISTRY 347 Table 1. Crystal structure and refinement data for complex 1. Formula C H ClFNOsP 18 15 7 2 formula weight 648.94 Crystal system triclinic space group P1̄ unit cell dimensions a = 7.1129(2) Å α = 95.2550(10)° b = 9.9492(3) Å β = 95.9100(10)° c = 14.1312(4) Å γ = 104.6480(10)° V (Å3) 95,517(5) Z 2 ρ (mg m−3) 2.256 Calcd T (K) 173(2) μ (mm−1) 3.235 λ (Å) 0.71073 F(0 0 0) 616 Crystal size (mm3) 0.220 × 0.190 × 0.130 θ to θ (°) 1.460 to 25.499 min max no. of reflns. collected 18,978 no. of indep. reflns. 3549 [R(int)] = 0.0199] Completeness to θ 99.9% absorbed correction semi-empirical from equivalents max. and min. transmission 0.404 and 0.235 refinement method full-matrix least-squares on F2 data / restraints / parameters 3549 / 0 / 271 Goodness-of-fit on F2 1.172 R and wR [I >2 σ(I)] 0.0147 and 0.0398 1 2 R and wR (all data) 0.0148 and 0.0398 1 2 largest diff. peak and hole (e.Å−3) 1.068 and −0.933 3: Red solid, yield 80%, m.p. 200 °C 1H NMR, δ (ppm): 9.58 (d, J = 5.48 Hz, 1H, py), (dec.). HH 9.25 (s, 1H, CH=N), 8.37 (d, J = 7.44 Hz, 1H, py), 8.28 (m, 1H, py), 8.24 (m, 1H, Py), 7.84 (m, HH 1H, Ar), 7.63 (d, J = 8.32 Hz, 1H, Ar), 7.41 (d, J = 8.16 Hz, 1H, Ar), 6.07 (s, 6H, C H ), 2.49 (m, HH HH 6 6 3H, CH ). 13C NMR, δ (ppm): 168.34 (CH=N), 155.99 (py), 149.34 (py), 140.35 (py), 139.69 (py), 3 129.93 (py), 129.83 (Ar), 129.74 (Ar), 122.81 (Ar), 78.88 (C H ), 20.75 (CH ). IR (KBr, cm−1): 1616.3 6 6 3 v(–CH=N), 819.9 v(P-F). Anal. Calcd (%): C, 35.88; H 2.81; N, 4.34. Found: C, 35.50; H, 2.79; N 4.30. MS (ESI, M/Z): 501.0781 [C H ClN Os]+. 19 18 2 3. Results and discussion 3.1. Preparation and characterization of the Os(II) complexes Complexes 1–3 were synthesized by stirring the osmium dimer [(η6-C H )Os(μ-Cl)Cl)] with 6 6 2 the corresponding N,N′-chelating ligand in methanol at 40 °C. Ammonium hexaflourophos- phate was used to exchange the chloride salts to give air stable compounds (Scheme 1). The formulation of complexes 1–3 was confirmed by 1H and 13C NMR spectroscopy, where the imine proton and carbon shifts of the complexes and the ligands were compared. Thus, the proton signal of the imine (CH=N) moves downfield to δ = 8.90–9.10 ppm for the complex from the region of 8.57-8.59 ppm of the uncoordinated ligand (Table 2), due to the imine proton being deshielded, since a lone pair of electrons is donated from the nitrogen to the osmium [28–30]. The 13C NMR spectra show the imine carbon shifting from the δ region 160.10–164.15 ppm for the uncoordinated ligands to the region 168.32–169.10 ppm for the coordinated ligand in the complexes. Similar shifts were observed for related ruthenium compounds [30]. 348 J. M. GICHUMBI ET AL. Scheme 1. synthesis of [(η6-benzene)osCl(Chn-2-Ch=n-ChX)](Pf) (X = f (1), Cl (2), methyl (3)). 5 4 6 5 6 Table 2. 1h, 13C nmr, and ir data for the imine groups of the free ligands and complexes 1–3. 1H NMR (ppm) 13C NMR (ppm) IR (cm−1) Ligand Complex Complex Ligand Complex Ligand Complex v(C=N) v(C=N) pyridine pyridine 1 8.57 9.28 161.4 169.4 1624.5 1612.5 2 8.58 9.30 160.6 169.9 1623.3 1611.8 3 8.49 9.25 163.2 168.3 1625.6 1616.3 A strong absorption band between 1609 and 1616 cm−1 is seen in the IR spectra of the complexes. This is due to the symmetrical vibration of the v(C=N) bonds. These peaks have shifted to lower wavenumbers than those of the free pyridine–imine ligands, which further supports that the ligand has coordinated to osmium (Table 2). Similar trends on complex formation have been well documented [28–30]. Furthermore, a sharp peak is seen for the compounds in the region 840-940 cm−1. This is due to the PF − counter ion [28–31]. 6 High-resolution mass spectra further confirmed the formation of the mononuclear osmium complexes. Thus, the spectra showed the expected molecular ion of the N,N′- bidentate complexes [(η6-arene)OsCl(C H N-2-CH=N-R)]+, with the characteristic multiple 5 4 peaks that are expected for the isotopes of osmium. The UV–vis data of complexes 1–3 (Figure 1) obtained in acetonitrile showed n–π* tran- sition absorption bands in the region 240–250 nm and π–π* transition bands in the region 280–330 nm. Also, the compounds showed peaks in the region 420–429 nm. These are due to a metal-to-ligand (dπ–π*) charge transfer transition to the empty π* orbital from the filled 5d orbital of the osmium [30], thus further confirming complex formation. 3.2. X-ray crystallography The single crystal X-ray diffraction structure of complex 1 is shown in Figure 2, while Table 3 shows important bond angles and lengths. Compound 1 crystallized as red blocks, with one molecule of the cation [(η6-benzene)Os(4-(fluoro-phenyl)-pyridin-2-yl-methylene amine)]+ and one anion PF −, in the asymmetric unit. In the structure, osmium is bonded to 6 the pyridyl-imine ligand via the imine and pyridine N atoms and to a chlorine atom as a third “leg” of the base of what can be described as a piano stool. The carbon atoms of the arene ring serve as the apex or seat of the piano stool. The geometry formed in this instance is JOURNAL OF COORDINATION CHEMISTRY 349 0.45 0.4 0.35 0.3 0.25 0.2 0.15 0.1 0.05 0 210 260 310 360 410 460 510 560 generally referred to as “pseudo-octahedral geometry of a three-legged piano stool” structure. The Os–Cl distance is 2.399(7) Å while the Os–N distances are 2.082(2) and 2.076(2) Å; these are comparable to distances found in similar compounds [16, 32]. Interestingly they are even comparable to our previous published Ru(II) complexes in which the Ru–Cl bond distances are just slightly longer and ranged between 2.4009(6)–2.4064(5) Å. The Os–N bond distances in 1, however, seem consistent with Ru–N bond distances in which one is short while the other is slightly longer [33, 34]. The N–Os–N bond angle is slightly more acute ecnabrosbA 1 2 3 Wavelength(nm) Figure 1. uV–Vis absorption spectra of the mononuclear complexes 1–3 in acetonitrile at 298 K. Figure 2. the molecular structure of complex 1 along with the atomic numbering scheme. the displacement ellipsoids are drawn at a 50% probability level. 350 J. M. GICHUMBI ET AL. compared to the two N–Os–Cl bond angles and are 75.84, 82.47(6), and 85.30(7)°, respec- tively, as observed in the Ru analogs and other similar compounds [16, 32–34]. 3.3. Stability in aqueous media Hydrolysis (aquation) of complexes is potentially part of the mechanism of the activation of halo-osmium arene compounds when they interact with biological targets such as DNA. Therefore, the aqueous behavior of complexes 1–3 was investigated in a DMSO-d /D O 6 2 solution and was monitored by 1H NMR spectroscopy. All the compounds showed signs of aquation equilibrium after about 30 min to 1 h. The hydrolysis of the complexes implies that the mechanism for anticancer activity of these complexes involves activation by hydrolysis [33]. Figure 3 shows the proton NMR spectra of 2 as a function of time in DMSO-d /D O, as 6 2 example. This is slower than what is observed in similar Ru complexes [33, 34]. 3.4. Cytotoxicity tests by MTT assay The cytotoxicity of the [(η6-C H )OsCl(C H N-2-CH=N-C H X)](PF ) complexes 1–3 was deter- 6 6 5 4 6 5 6 mined using the MTT assay against Caco-2, MCF-7, and HepG2 cancer lines and the non-can- cer cell line HEK293. The anti-cancer drug 5-fluorouracil, commonly used for the treatment of these cancers, was used as a control (Table 4). In addition the free ligand 4-(chlorophe- nyl)-pyridin-2-yl-methylene amine was tested and was found to be inactive. Table 3. selected bond distances (Å) and angles (°) for complex 1. Bond distances os(1)–n(1) 2.076(2) os(1)–n(2) 2.082(2) os(1)–Cl(1) 2.399(7) n(1)–C(1) 1.339(4) n(1)–C(5) 1.355(4) n(2)–C(6) 1.286(4) n(2)–C(7) 1.443(4) Bond angles n(1)–os(1)–n(2) 75.84(9) n(1)–os(1)–Cl(1) 85.30(7) n(2)–os(1)–Cl(1) 82.47(6) Figure 3. 1h nmr spectra of complex 2 in dmso-d/do. 6 2 JOURNAL OF COORDINATION CHEMISTRY 351 Table 4. In vitro cytotoxic effect of the osmium compounds 1–3. IC (μM) a 50 Compound HEK293 Caco2 MCF7 HepG2 1 213.7 ± 3.8 46.6 ± 0.5 77.1 ± 6.2 101.7 ± 2.3 2 221.2 ± 5.3 61.2 ± 4.4 49.6 ± 6.2 34.3 ± 8.4 3 205.1 ± 1.4 46.3 ± 11.1 76.6 ± 0.9 66.7 ± 5.7 5–fu b 47.4 ± 0.8 73.9 ± 4.5 30.5 ± 2.9 50.2 ± 1.2 athe iC values correspond to the concentrations of the respective complexes needed to cause 50% net cells mortality. 50 b5–fu or 5-fluorouracil was used as a reference drug. In order to explore the selectivity of the compounds towards tumor cells, as opposed to non-cancer cell lines, the compounds were screened for their antiproliferative effects using the human embryonic kidney cell (HEK293), a model for healthy cell lines. All compounds were found to be selective towards the tumor cells, as opposed to the control, 5-fluorouracil, which was found to be nonselective between the non-cancer and cancer cell lines. Also, all the complexes were found to be more active against Caco-2 (human epithelial colorectal adenocarcinoma) than the anticancer drug 5-flourouracil, with IC values ranging from 50 46.3 ± 11.1 to 61.2 ± 4.4 μM. Compounds 1–3 were found to have moderate activity against MCF-7 (human breast adenocarcinoma) with IC values ranging from 49.6 ± 6.2 to 50 76.6 ± 0.9 μM. Complex 2 showed higher activity than the anticancer drug for HepG2 (human hepatocellular carcinoma), complex 1 was inactive, and complex 3 showed moderate activity (Table 4). A comparison of the activity of these complexes to other osmium compounds is difficult due to the different cancer cell lines used by various researchers. Researchers have reported arene complexes of osmium of the general formula [(η6-biphenyl)Os(ethylenediamine)Cl] (PF ) [5], and [(η6-arene)Os(AB)X]+ (AB = azopyridine derivatives or phenyl imonopyridine, 6 X=Cl or I, arene = biphenyl or p-cymene) [16, 35] which showed good activity. In addition, arene compounds of Os(II) with different indolo[3,2-d]benzazepines (paullones), indolo[3,2-c] quinolones [19, 36, 37], and picolinate derivatives were reported to show good activity [35]. However, the cationic Os(II) complexes reported by Govender et al. were reported to be inactive [38]. Comparing these osmium complexes to related ruthenium complexes, the osmium com- plexes were found to be more selective between tumor and non-tumor cells as opposed to most of the ruthenium compounds [33, 34]. This may be attributed to the different ligands used with the Ru complexes. The ligands in the osmium complexes are closer to the coor- dination sphere of osmium than was observed for the Ru complexes. This is likely because of the shorter bond lengths in the osmium complexes and less strained bond angles, possibly resulting in a more stable active compound. In addition the osmium complexes were more active against Caco2 and HepG2 as compared to most of the related ruthenium compounds [33, 34]. This shows that the metal in the half-sandwich complexes has an effect on the cytotoxicity properties. 3.5. Antibacterial susceptibility tests Complex 2 was taken as a representative, due to its high anticancer activity, and its corre- sponding pyridyl-imine ligand (4-(chloophenyl)-pyridin-2-yl-methylene amine) was also 352 J. M. GICHUMBI ET AL. chosen, and both were investigated for potential antimicrobial activities against six Gram- positive and four Gram-negative bacteria. In addition, the activity of the ligand and its com- plex 2 against these bacteria were compared to tetracycline (TE30) and ampicillin (AMP10), which are known antibacterial drugs (Table 5). The ligand did not demonstrate any antimi- crobial activity against the panel of bacterial strains tested. Complex 2 was most effective against the Gram-positive bacteria and showed activity against all the bacterial strains tested. Thus, complex 2 showed anti-mycobacterial activity against the microbial strain M. smeg- matis, which belongs to the same genus as M. tuberculosis, the bacterium which causes tuberculosis. In addition, compound 2 showed bactericidal activity against the drug-resistant E. faecalis and methicillin-resistant S. aureus ATCC 43300. The complex was less effective against the Gram-negative bacteria tested. Complex 2 showed activity against E. coli ATCC 25922 at high concentrations. Of interest is the activity observed against the β-lactam-resistant strain E. coli ATCC 35218 (TEM-containing strain) which is usually drug-resistant. In addition, it was interesting to observe some promising activity, although at 0.8 mg, against the sulfhydryl variable (SHV)-containing K. pneumoniae ATCC 700603 strain (extended-spectrum β-lactamase-expressing), which is known to cause community-acquired bacterial pneumonia. It is also an important hospital-acquired patho- gen that causes high morbidity and mortality [39]. Here, the activity of complex 2 was better than that demonstrated by ampicillin. Furthermore, complex 2 showed activity at 0.8 mg against the highly resistant and therapy-recalcitrant P. aeruginosa ATCC 27853 strain. Table 5. results of the antibacterial susceptibility test data for complex 2, with the inhibition zones reported to the nearest mm. Gram-positive bacteria B. subtilis E. faecalis S. aureus S. aureus S. saprophyticus M. smegmatis ATCC 6653 ATCC 51299 ATCC 29213 ATCC 43300 ATCC 35552 mc2 155 Compound (20 mg/mL) 10 20 40 10 20 40 10 20 40 10 20 40 10 20 40 10 20 40 Ligand 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 na na na (r) (r) (r) (r) (r) (r) (r) (r) (r) (r) (r) (r) (r) (r) (r) 2 15 17 20 0 15 18 7 13 18 10 15 20 0 0 12 14 16 21 dmso 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 (r) (r) (r) (r) (r) (r) (r) (r) (r) (r) (r) (r) (r) (r) (r) (r) (r) (r) tetracycline 36 (s) 0 (r) 28 (s) 36 (s) 26 (s) na (te30) – 30 μg ampicillin 40 (s) 25 (s) 25 (s) 20 (s) 11 (r) na (amP10) – 10 μg Gram-negative bacteria K. pneumonia ATCC P. aeruginosa ATCC E. coli ATCC 25922 E. coli ATCC 35218 700603 27853 Compound 10 20 40 10 20 40 10 20 40 10 20 40 Ligand 0 (r) 0 (r) 0 (r) 0 (r) 0 (r) 0 (r) 0 (r) 0 (r) 0 (r) 0 (r) 0 (r) 0 (r) 2 0 0 12 0 7 17 0 5 15 0 0 7 dmso 0 (r) 0 (r) 0 (r) 0 (r) 0 (r) 0 (r) 0 (r) 0 (r) 0 (r) 0 (r) 0 (r) 0 (r) tetracycline 27 (s) 23 (s) 12 (r) 15 (i) (te30) – 30 μg ampicillin 20 (s) 0 (r) 0 (r) 0 (r) (amP10) – 10 μg JOURNAL OF COORDINATION CHEMISTRY 353 4. Conclusion A series of new cationic mononuclear iminopyridyl complexes (1–3) were prepared from the osmium precursor [(η6-benzene)Os(μ-Cl)Cl] and the corresponding pyridine–imine lig- 2 ands. These compounds were fully characterized using different spectroscopic and analytical techniques. The X-ray structure of complex 1 was solved and showed a pseudo-octahedral geometry around the Os atom and this, as do data from the spectral studies on the com- plexes, shows bidentate coordination of the ligands to the metal. All the compounds were found to be selective towards tumor cells as opposed to the control 5-fluorouracil, which was found to be significantly less selective between the non-cancer and cancer cell lines. Furthermore, the complexes were found to be more active against Caco-2 (human epithelial colorectal adenocarcinoma) than the anticancer drug 5-fluorouracil with IC values ranging 50 from 46.3 ± 11.1 to 61.2 ± 4.4 μM. Compounds 1–3 were found to have moderate activity against MCF-7 (human breast adenocarcinoma) with IC values ranging from 49.6 ± 6.2 to 50 76.6 ± 0.9 μM. Complex 2 further showed higher activity than the anticancer drug towards HepG2 (human hepatocellular carcinoma). Complex 2, as well as the pyridyl-imine ligand (4-(chrolo-phenyl)-pyridin-2-yl-methylene amine), were tested for antibacterial activity against six Gram-positive and four Gram-negative bacteria. Complex 2 showed promising antibacterial activity against both Gram-positive and Gram-negative bacteria. Supplementary material CCDC 1570620 contains the supplementary crystallographic data for this paper. These data can be obtained free of charge via http://www.ccdc.cam.ac.uk/conts/retrieving.html (or from the Cambridge Crystallographic Data Center, 12 Union Road, Cambridge CB2 1EZ, UK; Fax:+44 1223 336033). Acknowledgements We thank the NRF, THRIP, and UKZN for financial support. JMG thanks Prof. E.N. Njoka for his support. Disclosure statement No potential conflict of interest was reported by the authors. Funding This work was supported by the NRF; THRIP [grant number TP 1208035643]; UKZN. References [1] E.A. Enyedy, G.M. Bognar, T. Kiss, M. Hanif, C.G. Hartinger. J. Organomet. Chem., 734, 38 (2013). [2] E. Sija, C.G. Hartinger, B.K. Keppler, T. Kiss, E.A. Enyedy. Polyhedron, 67, 51 (2014). [3] S. Mukhopadhyay, R.K. Gupta, R.P. Paitandi, N.K. Rana, G. Sharma, B. Koch, L.K. Rana, M.S. Hundal, D.S. Pandey. Organometallics, 34, 4491 (2015). [4] P. Govender, N.C. Antonels, J. Mattsson, A.K. Renfrew, P.J. Dyson, J.R. Moss, B. Therrien, G.S. Smith. J. Organomet. Chem., 694, 3470 (2009). [5] A.F.A. Peacock, A. Habtemariam, S.A. Moggach, A. Prescimone, S. Parsons, P.J. Sadler. Inorg. Chem., 46, 4049 (2007). [6] M. Hanif, M.V. Babak, C.G. Hartinger. Drug Discovery Today, 19, 1640 (2014). 354 J. M. GICHUMBI ET AL. [7] W.D. Harman, W.P. Schaefer, H. Taube. J. Am. Chem. Soc., 112, 2682 (1990). [8] K. Severin. Chem. Commun., 3859 (2006). [9] J.L. Boyer, M.L. Kuhlman, T.B. Rauchfuss. Acc. Chem. Res., 40, 233 (2007). [10] T. Reiner, M. Waibel, A.N. Marziale, D. Jantke, F.J. Kiefer, T.F. Fässler, J. Eppinger. J. Organomet. Chem., 695, 2667 (2010). [11] B. Therrien, G. Süss-Fink, P. Govindaswamy, A.K. Renfrew, P.J. Dyson. Angew. Chem. Int. Ed., 47, 3773 (2008). [12] J.W. Faller, A.R. Lavoie. Org. Lett., 3, 3703 (2001). [13] J.W. Faller, A.R. Lavoie. Organometallics, 20, 5245 (2001). [14] J.M. Gichumbi, H.B. Friedrich, B. Omondi. Eur. J. Inorg. Chem., 2017, 915 (2017). [15] Y. Fu, R. Soni, M.J. Romero, A.M. Pizarro, L. Salassa, G.J. Clarkson, J.M. Hearn, A. Habtemariam, M. Wills, P.J. Sadler. Chem. Eur. J., 19, 15199 (2013). [16] Y. Fu, M.J. Romero, A. Habtemariam, M.E. Snowden, L. Song, G.J. Clarkson, B. Qamar, A.M. Pizarro, P.R. Unwin, P.J. Sadler. Chem. Sci., 3, 2485 (2012). [17] A.F.A. Peacock, A. Habtemariam, S.A. Moggach, A. Prescimone, S. Parsons, P.J. Sadler. Inorg. Chem., 46, 4049 (2007). [18] L.K. Filak, S. Göschl, P. Heffeter, K. Ghannadzadeh Samper, A.E. Egger, M.A. Jakupec, B.K. Keppler, W. Berger, V.B. Arion. Organometallics, 32, 903 (2013). [19] L.K. Filak, G. Mühlgassner, F. Bacher, A. Roller, M. Galanski, M.A. Jakupec, B.K. Keppler, V.B. Arion. Organometallics, 30, 273 (2011). [20] L.K. Filak, S. Göschl, S. Hackl, M.A. Jakupec, V.B. Arion. Inorg. Chim. Acta, 393, 252 (2012). [21] G. Mühlgassner, C. Bartel, W.F. Schmid, M.A. Jakupec, V.B. Arion, B.K. Keppler. J. Inorg. Biochem., 116, 180 (2012). [22] P.M. Loiseau, D.G. Craciunescu, J.C. Doadrio-Villarejo, G. Certad-Fombona, P. Gayrol. Trop. Med. Parasitol., 43, 110 (1992). [23] S. Dehghanpour, N. Bouslimani, R. Welter, F. Mojahed. Polyhedron, 26, 154 (2007). [24] T. Mosmann. J. Immun. Methods, 65, 55 (1983). [25] CLSI. Performance Standard for Antimicrobial Susceptibility Testing; Seventeenth Information Supplement, CLSI document M100-S17, Clinical and Laboratory Standards Institute, Pennsylvania, ISBN 1-56238-625-5, p. 34 (2007). [26] Bruker-AXS. Bruker-AXS, Madison, Wisconsin, USA (2009). [27] G.M. Sheldrick. Acta Crystallogr., A64, 112 (2008). [28] J.M. Gichumbi, H.B. Friedrich, B. Omondi. J. Mol. Struct., 1113, 55 (2016). [29] J.M. Gichumbi, H.B. Friedrich, B. Omondi. J. Mol. Catal. A: Chem., 416, 29 (2016). [30] J.M. Gichumbi, H.B. Friedrich, B. Omondi. J. Organomet. Chem., 808, 87 (2016). [31] S. Orbisaglia, C. Di Nicola, F. Marchetti, C. Pettinari, R. Pettinari, L.M.D.R.S. Martins, E.C.B.A. Alegria, M.F.C. Guedes da Silva, B.G.M. Rocha, M.L. Kuznetsov, A.J.L. Pombeiro, B.W. Skelton, A.N. Sobolev, A.H. White. Chem. Eur. J., 20, 3689 (2014). [32] A.F.A. Peacock, A. Habtemariam, R. Fernández, V. Walland, F.P.A. Fabbiani, S. Parsons, R.E. Aird, D.I. Jodrell, P.J. Sadler. J. Am. Chem. Soc., 128, 1739 (2006). [33] J.M. Gichumbi, H.B. Friedrich, B. Omondi, M. Singh, K. Naicker, H.Y. Chenia. J. Coord. Chem., 69, 3531 (2016). [34] J.M. Gichumbi, B. Omondi, G. Lazarus, M. Singh, N. Shaikh, H.Y. Chenia, H.B. Friedrich. Z. Anorg. Allg. Chem., 643, 699 (2017). [35] S.H. Rijt, A.F.A. Peacock, P.J. Sadler. In Platinum and Other Heavy Metal Compounds in Cancer Chemotherapy, A. Bonetti, R. Leone, F.M. Muggia, S.B. Howell (Eds), Humana Press, Totowa, NJ, p. 73 (2009). [36] S.H. van Rijt, A.F.A. Peacock, R.D.L. Johnstone, S. Parsons, P.J. Sadler. Inorg. Chem., 48, 1753 (2009). [37] L.K. Filak, G. Mühlgassner, M.A. Jakupec, P. Heffeter, W. Berger, V.B. Arion, B.K. Keppler. J. Biol. Inorg. Chem., 15, 903 (2010). [38] P. Govender, F. Edafe, B.C.E. Makhubela, P.J. Dyson, B. Therrien, G.S. Smith. Inorg. Chim. Acta, 409, 112 (2014). [39] A. Doosti, M. Pourabbas, A. Arshi, M. Chehelgerdi, H. Kabiri. Osong Public Health Res. Persp., 6, 3 (2015).