Ruthenium- and osmium-arene complexes of 8-substituted indolo[3,2-c]quinolines: Synthesis, X-ray diffraction structures, spectroscopic properties, and antiproliferative activity.

Six novel ruthenium(II)- and osmium(II)-arene complexes with indoloquinoline modified ligands containing methyl and halo substituents in position 8 of the molecule backbone have been synthesised and comprehensively characterised by spectroscopic methods (1H, 13C NMR, UV-Vis), ESI mass spectrometry and X-ray crystallography. Binding of indoloquinolines to a metal-arene scaffold makes the products soluble enough in biological media to allow for assaying their antiproliferative activity. The complexes were tested in three human cancer cell lines, namely A549 (non-small cell lung cancer), SW480 (colon carcinoma) and CH1 (ovarian carcinoma), yielding IC50 values in the 10-6-10-7 M concentration range after continuous exposure for 96 h. Compounds with halo substituents in position 8 are more effective cytotoxic agents in vitro than the previously reported species halogenated in position 2 of the indoloquinoline backbone. High antiproliferative activity of both series of substances may be due at least in part to their potential to act as DNA intercalators.

Introduction

The fight against cancer has made considerable progress by the introduction of targeted therapies in recent years. This treatment modality takes advantage of certain features of malignant tumours to selectively inhibit their growth, ideally associated with low side effects for patients. The numerous concepts that are currently being explored to achieve tumour targeting in bioinorganic medicinal chemistry include ‘activation by reduction’ in hypoxic media, as well as ‘activation by ring opening’ in the solid tumour environment with lowered pH value [1-4]. Activation by reduction is believed to be the critical step in converting a prodrug into its active form [5]. Well-known examples supporting this hypothesis are satraplatin, a PtIV compound that reached a clinical phase III study [6], NAMI-A [7], as well as KP1019 [8], the first ruthenium(III) coordination compounds in clinical studies. Another way to gain selectivity for malignant cells over healthy tissue is targeting enzymes or receptors that are overexpressed in certain tumour types, e.g. thioredoxin reductase [9], ribonucleotide reductase [10,11], DNA topoisomerase [12] or glutathione S-transferase [13]. Another example are ferrocifen derivatives [14], which are based on hydroxytamoxifen, an oestrogen receptor antagonist used in hormone-positive breast cancer therapy [15]. In ferrocifen, one of the phenyl rings is replaced by a ferrocenyl unit, combining the hormone-antagonistic ligand with a metal–organic redox active moiety. Similar attempts combining the benefits of organometallic core with biologically active ligands were undertaken with indolobenzazepines, also referred to as paullones. The paullones were originally predicted to possess cyclin dependent kinase (CDK)-inhibitory properties by a COMPARE analysis [16]. CDKs together with their corresponding cyclins act as cell cycle triggers, controlling cell division [17]. By interference with this highly balanced regulatory system, cell proliferation can be controlled. In vitro models confirmed the CDK-inhibitory properties of the paullones [18], and up to date a broad range of paullone derivatives has been evaluated for biological activity [19,20]. For some paullones, other intracellular targets such as glycogen synthase kinase 3β (GSK3β) and mitochondrial malate dehydrogenase (mMDH) could be identified [21].

Indoloquinolines also attracted interest during the last few years [22-26] due to the development of convenient preparation routes [27]. In contrast to paullones with a folded seven-membered azepine ring, indoloquinolines are flat heteroaromatic ring systems, in which the paullone azepine ring was replaced by a six-membered pyridine ring. We anticipated that this transformation will alter significantly the physico-chemical and biological properties compared to the reference (paullone) compounds.

In order to overcome their limited solubility in biocompatible media, paullones were complexed to metal ions. Ga(III) [28], Ru(II) [29] and Cu(II) [30] coordination compounds, as well as a series of Ru(II)- and Os(II)-arene complexes of modified paullone ligands [31-33] are well-documented in the literature. Interestingly, CDK inhibition by metal-based paullones does not necessarily parallel their in vitro antiproliferative activity, making other intracellular targets likely to be involved in their mechanism of action [34].

Novel SAR studies showed that some ruthenium- and osmium- arene complexes of indoloquinolines are by a factor of 10 more active than corresponding paullone complexes in human cancer cell lines. It is worth noting, however, that the indoloquinoline-based complexes with a bidentate ethylenediamine binding site are less stable than their paullone counterparts, dissociating in aqueous media with release of the ligand [34]. Remarkably, other ethylenediamine based ruthenium-arene complexes do not show propensity for dissociation under similar conditions [35-37]. To increase the thermodynamic stability and kinetic inertness of the complexes, sp2-hybridised N-donor atoms were introduced by condensation of an indoloquinoline azine with 2-formyl- or 2-acetylpyridine [38]. This modification led to complexes with increased stability in biocompatible media, while retaining the in vitro antiproliferative activity. Further studies on modified indoloquinolines containing different substituents in position 2 of the molecular backbone showed that electron-withdrawing substituents are unfavourable for cytotoxicity, whereas an electron-donating methyl group has no influence on antiproliferative activity. The effect of substituents in position 8 of the indoloquinoline backbone was studied on copper(II) complexes which were found highly cytotoxic with IC50 values in the nanomolar concentration range [39]. Synthesis of those ligands is depicted in Scheme 1.

Scheme 1

Synthesis of the indoloquinoline modified ligands [38,39]. Reagents and conditions: (i) BH3·THF, THF, Ar, r.t., 24–72 h; (ii) glacial HOAc, reflux, 3–4 h; (iii) POCl3, Ar, reflux, 26 h; (iv) N2H4·H2O, Ar, 100 °C, 24 h; (v) 2-acetylpyridine, EtOH, Ar, 65 °C, 18 h.

Herein we report on the synthesis of six novel ruthenium- and osmium-arene complexes with indoloquinoline-based ligands (1a,b–3a,b) containing substituents with different electronic properties in position 8 of the indoloquinoline backbone (Scheme 2). Their antiproliferative activity in three human cancer cell lines, namely A549 (non-small cell lung cancer), SW480 (colon carcinoma) and CH1 (ovarian carcinoma) has been studied and compared to that of chemically related complexes (4a,b–6a,b and others).

Scheme 2

Ruthenium- and osmium-arene complexes with substituted indoloquinoline ligands. (See Ref. [38] for details about 4a,b–6a,b.)

Experimental

Chemicals

Ethanol and THF were dried using standard procedures. α-Terpinene, 2-amino-5-chlorobenzonitrile were purchased from Acros Fisher, ruthenium trichloride and osmium tetroxide from Johnson Matthey, KBr from Merck, while 2-acetylpyridine, hydrazine dihydrochloride, hydrazine hydrate, phosphorus oxychloride, isatin, glacial acetic acid, borane in THF, 2-aminobenzonitrile, 2-amino-5-methylbenzonitrile, sodium perborate tetrahydrate were from Sigma–Aldrich. All these chemicals were used as received.

Synthesis

The ligands HL were synthesized by following the literature protocols [38,39]. Ruthenium- and osmium-arene starting compounds [M(p-cymene)(Cl)(μ-Cl)]2, where M = RuII and OsII, were prepared as described previously [40,41]. For preparation of [Os(p-cymene)(Cl)(μ-Cl)]2 OsO4 was reduced first to H2[OsCl6] by N2H4·2HCl in conc. HCl [42], and then reacted with α-terpinene.

General procedure A for the complexation of -arene scaffold: The corresponding ligand HL in a Schlenk tube was flushed with argon and suspended in dry ethanol. The corresponding metal-arene dimer was dissolved in chloroform and added to the ethanolic ligand suspension. The reaction mixture was stirred at room temperature under argon atmosphere (in the case of the Os complexes, light protection was also applied). The reaction mixture was filtered through a GF3 filter paper, and slowly added to diethyl ether previously dried over sodium sulfate. The precipitate formed was separated by filtration and dried in vacuo at 50 °C.

[Ru(p-cymene)(HL1)Cl]Cl, 1a

General procedure A: N-(8-Methyl-5,11-dihydroindolo [3,2-c]—quinolin-6-ylidene)-N’-(1-pyridin-2-yl-ethylidene)azine (HL, 120 mg, 0.33 mmol), bis((η-cymene)(chlorido)(μ-chlorido)ruthenium(II)) (101 mg, 0.16 mmol), EtOH abs. (4 mL), CHCl3 (0.3 mL), diethyl ether dried over Na2SO4 (100 mL), stirring for 22.5 h. To remove traces of the unreacted ruthenium dimer the red precipitate was dissolved in a minimal amount of EtOH and filtered through a GF3 filter paper. After addition of CHCl3 (1 mL), the filtrate was added dropwise to diethyl ether previously dried over Na2SO4 (100 mL). The resulting precipitate was filtered off and dried in vacuo at 50 °C. Yield 154 mg, 69%. Anal. Calc. for C33H33Cl2N5Ru·1.5H2O (Mr 698.65): C, 56.73; H, 5.19; N, 10.02. Found: C, 56.64; H, 5.01; N, 9.94%. ESI-MS (methanol), positive: m/z 636 [M−Cl]+.

1H NMR (500 MHz, DMSO-d): 12.99 (s, 1H, H11), 10.36 (s, 1H, H5), 9.60 (d, 1H, 3J = 6 Hz, H17), 8.36 (d, 1H, 3J = 8 Hz, H1), 8.31–8.26 (m, 1H, H19), 8.23–8.19 (m, 2H, H7 + H20), 7.83–7.79 (m, 1H, H18), 7.63 (d, 1H, 3J = 8.3 Hz, H10), 7.61–7.57 (m, 2H, H3 + H4), 7.42 (ddd, 1H, 3J = 8 Hz, 3J = 6 Hz, 4J = 2 Hz, H2), 7.31 (dd, 1H, 3J = 8 Hz, 4J = 1 Hz, H9), 6.03 (d, 1H, 3J = 6 Hz, Hcy1), 5.80 (d, 1H, 3J = 6 Hz, Hcy2), 5.71 (d, 1H, 3J = 6 Hz, Hcy1’), 5.30 (d, 1H, 3J = 6 Hz, Hcy2’), 2.78–2.72 (m, 1H, Hcy3), 2.71 (s, 3H, H21), 2.54 (s, 3H, H8a), 2.13 (s, 3H, Hcy5), 1.08 (d, 3H, 3J = 7 Hz, Hcy4’), 1.04 (d, 3H, 3J = 7 Hz, Hcy4) ppm.

13C NMR (125 MHz, DMSO-d): 166.18 (Cq, C14), 156.36 (CH, C17), 155.32 (Cq, C15), 147.54 (Cq, C6), 140.28 (CH, C19), 139.42 (Cq, C11a), 137.20 (Cq, C10a), 136.69 (Cq, C4a), 130.77 (Cq, C8), 130.21 (CH, C3), 127.34 (CH, C18), 126.59 (CH, C20), 126.54 (CH, C9), 124.15 (Cq, C6b), 123.41 (CH, C2), 123.09 (CH, C1), 121.91 (CH, C7), 117.12 (CH, C4), 113.57 (Cq, C11b), 112.23 (CH, C10), 104.17 (Cq, Ccy1a), 103.78 (Cq, C6a), 103.35 (Cq, Ccy2a), 87.66 (CH, Ccy1’), 86.45 (CH, Ccy1), 86.14 (CH, Ccy2), 84.70 (CH, Ccy2’), 31.05 (CH, Ccy3), 22.25 (CH, Ccy4), 22.04 (CH, C8a), 21.78 (CH, Ccy4’), 18.94 (CH, Ccy5), 15.99 (CH, C21) ppm.

[Os(p-cymene)(HL1)Cl]Cl, 1b

General procedure A: N-(8-Methyl-5,11-dihydroindolo[3,2-c]quinolin-6-ylidene)-N’-(1-pyridin-2-yl-ethylidene)azine (HL, 150 mg, 0.41 mmol), bis((η-cymene)(chlorido)(μ-chlorido)osmium(II)) (167 mg, 0.22 mmol), EtOH abs. (4 mL), CHCl3 (1.1 mL), diethyl ether dried over Na2SO4 (100 mL), stirring for 18 h. Yield 207 mg, 66%. Anal. Calc. for C33H33Cl2N5Os·1.5H2O (Mr 787.81): C, 50.31; H, 4.61; N, 8.89. Found: C, 50.40; H, 4.45; N, 8.87%. ESI-MS (methanol), positive: m/z 726 [M−Cl]+.

1H NMR (500 MHz, DMSO-d): 13.03 (s, 1H, H11), 9.96 (s, 1H, H5), 9.54 (d, 1H, 3J = 6 Hz, H17), 8.36 (d, 1H, 3J = 8 Hz, H1), 8.32 (d, 1H, 3J = 8 Hz, H20), 8.29–8.25 (m, 1H, H19), 8.15 (s, 1H, H7), 7.82–7.78 (m, 1H, H18), 7.63 (d, 1H, 3J = 8 Hz, H10), 7.60–7.55 (m, 1H, H3), 7.45 (d, 1H, 3J = 8 Hz, H4), 7.42–7.38 (m, 1H, H2), 7.30 (dd, 1H, 3J = 8 Hz, 4J = 1 Hz, H9), 6.32 (d, 1H, 3J = 6 Hz, Hcy1), 6.02 (d, 1H, 3J = 6 Hz, Hcy2), 5.97 (d, 1H, 3J = 6 Hz, Hcy1‘), 5.43 (d, 1H, 3J = 6 Hz, Hcy2‘), 2.78 (s, 3H, H21), 2.69–2.60 (m, 1H, Hcy3), 2.53 (s, 3H, H8a), 2.18 (s, 3H, Hcy5), 1.03 (d, 3H, 3J = 7 Hz, Hcy4‘), 1.00 (d, 3H, 3J = 7 Hz, Hcy4) ppm.

13C NMR (125 MHz, DMSO-d): 167.18 (Cq, C14), 156.30 (Cq, C15), 156.28 (CH, C17), 147.66 (Cq, C6), 140.40 (CH, C19), 139.35 (Cq, C11a), 137.17 (Cq, C10a), 136.47 (Cq, C4a), 130.76 (Cq, C8), 130.20 (CH, C3), 128.21 (CH, C18), 126.74 (CH, C20), 126.51 (CH, C9), 124.11 (Cq, C6b), 123.38 (CH, C2), 123.13 (CH, C1), 121.83 (CH, C7), 117.04 (CH, C4), 113.50 (Cq, C11b), 112.24 (CH, C10), 103.42 (Cq, C6a), 96.47 (Cq, Ccy2a), 95.29 (Cq, Ccy1a), 79.44 (CH, Ccy1’), 78.23 (CH, Ccy1), 76.70 (CH, Ccy2), 75.16 (CH, Ccy2’), 31.21 (CH, Ccy3), 22.55 (CH, Ccy4), 22.02 (CH, C8a), 21.89 (CH, Ccy4’), 18.84 (CH, Ccy5), 15.74 (CH, C21) ppm.

[Ru(p-cymene)(HL2)Cl]Cl, 2a

General procedure A: N-(8-Chloro-11H-indolo[3,2-c]quinolin-6-yl)-N’-(1-pyridine-2-yl-ethylidene)azine (HL, 100 mg, 0.26 mmol) and bis((η-cymene)(chlorido)(μ-chlorido)ruthenium(II)) (80 mg, 0.13 mmol). Yield 103 mg, 58%. Anal. Calc. for C32H30Cl3N5Ru·1.5H2O (Mr 719.07): C, 53.45; H, 4.63; N, 9.74. Found: C, 53.50; H, 4.41; N, 9.69%. ESI-MS (methanol), positive: m/z 656 [M−Cl]+.

1H NMR (500 MHz, DMSO-d): 13.42 (s, 1H, H11), 10.42 (s, 1H, H5), 9.61 (d, 1H, 3J = 5 Hz, H17), 8.40 (d, 1H, 3J = 8 Hz, H1), 8.34 (d, 1H, 4J = 2 Hz, H7), 8.32–8.27 (m, 1H, H19), 8.22 (d, 1H, 3J = 8 Hz, H20), 7.84–7.79 (m, 1H, H18), 7.77 (d, 1H, 3J = 9 Hz, H10), 7.65–7.61 (m, 1H, H3), 7.59 (d, 1H, 3J = 8 Hz, H4), 7.49 (dd, 1H, 3J = 9 Hz, 4J = 2 Hz, H9), 7.46–7.41 (m, 1H, H2), 6.06 (d, 1H, 3J = 6 Hz, Hcy1), 5.82 (d, 1H, 3J = 6 Hz, Hcy2), 5.76 (d, 1H, 3J = 6 Hz, Hcy1‘), 5.33 (d, 1H, 3J = 6 Hz, Hcy2‘), 2.77–2.67 (m, 1H, Hcy3 + s, 3H, H21), 2.12 (s, 3H, Hcy5), 1.07 (d, 3H, 3J = 7 Hz, Hcy4‘), 1.04 (d, 3H, 3J = 7 Hz, Hcy4) ppm.

13C NMR (125 MHz, DMSO-d): 166.75 (Cq, C14), 156.39 (CH, C17), 155.18 (Cq, C15), 147.23 (Cq, C6), 140.58 (Cq, C11a), 140.33 (CH, C19), 137.45 (Cq, C10a), 137.03 (Cq, C4a), 130.79 (CH, C3), 127.51 (CH, C18), 126.76 (CH, C20), 126.31 (Cq, C8), 125.02 (Cq, C6b), 124.95 (CH, C9), 123.55 (CH, C2), 123.42 (CH, C1), 121.13 (CH, C7), 117.19 (CH, C4), 114.19 (CH, C10), 113.28 (Cq, C11b), 104.17 (Cq, Ccy1a), 103.59 (Cq, C6a), 103.37 (Cq, Ccy2a), 87.45 (CH, Ccy1’), 86.69 (CH, Ccy1), 86.11 (CH, Ccy2), 84.81 (CH, Ccy2’), 31.06 (CH, Ccy3), 22.33 (CH, Ccy4), 21.70 (CH, Ccy4’), 18.95 (CH, Ccy5), 16.02 (CH, C21) ppm.

[Os(p-cymene)(L2)Cl]Cl, 2b

N-(8-Chloro-11H-indolo[3,2-c]quinolin-6-yl)-N’-(1-pyridine-2-yl-ethylidene)azine (150 mg, 0.19 mmol) and bis((η-cymene)(chlorido)(μ-chlorido)osmium(II)) in a 25 mL Schlenk tube were suspendend in dry ethanol (3 mL) under Ar. The mixture was stirred for 24 h at room temperature. The precipitate formed was collected under suction, washed with small amounts of ethanol and dried in vacuo at 50 °C. Yield 268 mg, 88%. Anal. Calc. for C32H30Cl3N5Os·2.25H2O (Mr 821.74): C, 46.77; H, 4.23; N, 8.52. Found: C, 46.58; H, 3.83; N, 8.40%. ESI-MS (methanol), positive: m/z 746 [M−Cl]+.

1H NMR (500 MHz, DMSO-d): 13.36 (s, 1H, H11), 10.03 (s, 1H, H5), 9.54 (d, 1H, 3J = 6 Hz, H17), 8.37 (d, 1H, 3J = 8 Hz, H1), 8.34 (d, 1H, 3J = 8 Hz, H20), 8.30 (d, 1H, 4J = 2 Hz, H7), 8.29–8.25 (m, 1H, H19), 7.83–7.79 (m, 1H, H18), 7.76 (d, 1H, 3J = 9 Hz, H10), 7.63–7.59 (m, 1H, H3), 7.49 (dd, 1H, 3J = 9 Hz, 4J = 2 Hz, H9), 7.46 (d, 1H, 3J = 8 Hz, H4), 7.44–7.40 (m, 1H, H2), 6.34 (d, 1H, 3J = 6 Hz, Hcy1), 6.04–6.01 (m, 2H, Hcy1‘ + Hcy2), 5.47 (d, 1H, 3J = 6 Hz, Hcy2‘), 2.78 (s, 3H, H21), 2.67–2.59 (m, 1H, Hcy3), 2.18 (s, 3H, Hcy5), 1.02 (d, 3H, 3J = 7 Hz, Hcy4‘), 1.00 (d, 3H, 3J = 7 Hz, Hcy4) ppm.

13C NMR (125 MHz, DMSO-d): 167.77 (Cq, C14), 156.29 (CH, C17), 156.17 (Cq, C15), 147.35 (Cq, C6), 140.51 (Cq, C11a), 140.44 (CH, C19), 137.40 (Cq, C10a), 136.83 (Cq, C4a), 130.79 (CH, C3), 128.38 (CH, C18), 126.92 (CH, C20), 126.32 (Cq, C8), 125.00 (Cq, C6b), 124.95 (CH, C9), 123.54 (CH, C2), 123.37 (CH, C1), 121.07 (CH, C7), 117.14 (CH, C4), 114.19 (CH, C10), 113.18 (Cq, C11b), 103.28 (Cq, C6a), 96.47 (Cq, Ccy2a), 95.34 (Cq, Ccy1a),79.26 (CH, Ccy1’), 78.46 (CH, Ccy1), 76.71 (CH, Ccy2), 75.31 (CH, Ccy2’), 31.23 (CH, Ccy3), 22.62 (CH, Ccy4), 21.82 (CH, Ccy4’), 18.85 (CH, Ccy5), 15.77 (CH, C21) ppm.

[Ru(p-cymene)(L3)Cl]Cl, 3a

General procedure A: N-(8-Bromo-11H-indolo[3,2-c]quinolin-6-yl)-N’-(1-pyridine-2-yl-ethylidene)azine (120 mg, 0.28 mmol), bis((η-cymene)(chlorido)(μ-chlorido)ruthenium(II)) (81 mg, 0.13 mmol), ethanol abs. (3 mL), CHCl3 (0.5 mL), diethyl ether dried over Na2SO4 (180 mL), stirring for 23 h. Yield 124 mg (61%). Anal. Calc. for C32H30BrCl2N5Ru·1.5H2O (Mr 763.52): C, 50.34; H, 4.36; N, 9.17. Found: C, 50.36; H, 3.97; N, 9.02%. ESI-MS (methanol), positive: m/z 702 [M−Cl]+.

1H NMR (500 MHz, DMSO-d): 13.38 (s, 1H, H11), 10.41 (s, 1H, H5), 9.61 (d, 1H, 3J = 5 Hz, H17), 8.50 (br s, 1H, H7), 8.38 (d, 1H, 3J = 8 Hz, H1), 8.32–8.26 (m, 1H, H19), 8.22 (d, 1H, 3J = 8 Hz, H20), 7.84–7.79 (m, 1H, H18), 7.72 (d, 1H, 3J = 9 Hz, H10), 7.66–7.57 (m, 3H, H3 + H4 + H9), 7.46–7.41 (m, 1H, H2), 6.05 (d, 1H, 3J = 6 Hz, Hcy1), 5.83 (d, 1H, 3J = 6 Hz, Hcy2), 5.75 (d, 1H, 3J = 6 Hz, Hcy1‘), 5.34 (d, 1H, 3J = 6 Hz, Hcy2‘), 2.77–2.66 (m, 1H, Hcy3 + s, 3H, H21), 2.13 (s, 3H, Hcy5), 1.07 (d, 3H, 3J = 7 Hz, Hcy4‘), 1.04 (d, 3H, 3J = 7 Hz, Hcy4) ppm.

13C NMR (125 MHz, DMSO-d): 166.78 (Cq, C14), 156.38 (CH, C17), 155.18 (Cq, C15), 147.20 (Cq, C6), 140.39 (Cq, C11a), 140.32 (CH, C19), 137.71 (Cq, C10a), 137.04 (Cq, C4a), 130.81 (CH, C3), 127.52 (2 CH, C9 + C18), 126.77 (CH, C20), 125.64 (Cq, C6b), 124.17 (CH, C7), 123.56 (CH, C2), 123.39 (CH, C1), 117.20 (CH, C4), 114.61 (CH, C10), 114.27 (Cq, C8), 113.23 (Cq, C11b), 104.19 (Cq, Ccy1a), 103.47 (Cq, C6a), 103.37 (Cq, Ccy2a), 87.57 (CH, Ccy1’), 86.55 (CH, Ccy1), 86.11 (CH, Ccy2), 84.82 (CH, Ccy2’), 31.06 (CH, Ccy3), 22.28 (CH, Ccy4), 21.75 (CH, Ccy4’), 18.94 (CH, Ccy5), 16.02 (CH, C21) ppm.

[Os(p-cymene)(L3)Cl]Cl, 3b

General procedure A: N-(8-Bromo-11H-indolo[3,2-c]quinolin-6-yl)-N’-(1-pyridine-2-yl-ethylidene)azine (150 mg, 0.35 mmol), bis((η-cymene)(chlorido)(μ-chlorido)osmium(II)) (137 mg, 0.17 mmol), ethanol abs. (3 mL), CHCl3 (0.5 mL), diethyl ether dried over Na2SO4 (250 mL), stirring for 23 h. The precipitate was dissolved in ethanol (70 mL) and solution filtered. The solvent was removed under reduced pressure to ca. 5 mL and chloroform (1 mL) was added. Then the mixture was added dropwise to diethyl ether (300 mL) previously dried over Na2SO4, and the precipitate formed was collected under suction in vacuo at 50 °C. Further purification was achieved by crystallisation: dissolution of crude product (100 mg) in ethanol (100 mL), evaporation of the solvent to 40 mL and slow diffusion of diethyl ether into the solution (under Ar atmosphere). Yield: 71 mg (71%, based on 100 mg raw product taken for the recrystallisation, overall yield 25%). Anal. Calc. for C32H30BrCl2N5Os·0.75H2O (Mr 839.17): C, 45.80; H, 3.78; N, 8.35. Found: C, 45.48; H, 3.39; N, 8.28%. ESI-MS (methanol), positive: m/z 790 [M−Cl]+.

1H NMR (500 MHz, DMSO-d): 13.26 (s, 1H, H11), 10.02 (s, 1H, H5), 9.53 (d, 1H, 3J = 6 Hz, H17), 8.46 (d, 1H, 4J = 2 Hz, H7), 8.36–8.32 (m, 2H, H1 + H20), 8.30–8.25 (m, 1H, H19), 7.83–7.79 (m, 1H, H18), 7.71 (d, 1H, 3J = 9 Hz, H10), 7.64–7.59 (m, 2H, H3 + H9), 7.46 (d, 1H, 3J = 8 Hz, H4), 7.45–7.40 (m, 1H, H2), 6.34 (d, 1H, 3J = 6 Hz, Hcy1), 6.04 (d, 1H, 3J = 6 Hz, Hcy2), 6.01 (d, 1H, 3J = 6 Hz, Hcy1‘), 5.48 (d, 1H, 3J = 6 Hz, Hcy2‘), 2.78 (s, 3H, H21), 2.68–2.61 (m, 1H, Hcy3), 2.19 (s, 3H, Hcy5), 1.04 (d, 3H, 3J = 7 Hz, Hcy4‘), 1.01 (d, 3H, 3J = 7 Hz, Hcy4) ppm.

13C NMR (125 MHz, DMSO-d): 167.81 (Cq, C14), 156.28 (CH, C17), 156.18 (Cq, C15), 147.31 (Cq, C6), 140.44 (CH, C19), 140.32 (Cq, C11a), 137.66 (Cq, C10a), 136.85 (Cq, C4a), 130.81 (CH, C3), 128.38 (CH, C18), 127.53 (CH, C9), 126.93 (CH, C20), 125.63 (Cq, C6b), 124.12 (CH, C7), 123.55 (CH, C2), 123.31 (CH, C1), 117.17 (CH, C4), 114.60 (CH, C10), 114.29 (Cq, C8), 113.14 (Cq, C11b), 103.18 (Cq, C6a), 96.49 (Cq, Ccy2a), 95.36 (Cq, Ccy1a), 79.42 (CH, Ccy1’), 78.32 (CH, Ccy1), 76.73 (CH, Ccy2), 75.30 (CH, Ccy2’), 31.23 (CH, Ccy3), 22.57 (CH, Ccy4), 21.88 (CH, Ccy4’), 18.84 (CH, Ccy5), 15.80 (CH, C21) ppm.

Characterisation in solution

One-dimensional 1H and 13C NMR and two-dimensional 1H–1H COSY, 1H–1H TOCSY, 1H–1H ROESY or 1H–1H NOESY, 1H–13C HSQC and 1H–13C HMBC NMR spectra were recorded on two Bruker Avance III spectrometers at 500.32 or 500.10 (1H), and 125.82 or 125.76 (13C) MHz, respectively, by using as a solvent DMSO-d or CD3OD at room temperature and standard pulse programs. 1H and 13C shifts are quoted relative to the solvent residual signals. For the atom numbering scheme used for the NMR assignments, see Chart S1. UV–Vis spectra were recorded with a Perkin Elmer Lambda 650 spectrophotometer equipped with a six cell changer and a Peltier element for temperature control or an Agilent 8453 spectrophotometer. Electrospray ionisation mass spectrometry (ESI-MS) was carried out with a Bruker Esquire 3000 instrument; the samples were dissolved in methanol. All elemental analyses were performed at the Microanalytical Laboratory of the University of Vienna with a Perkin Elmer 2400 CHN Elemental Analyzer. Analytical HPLC analysis was performed on a Dionex Summit system controlled by Dionex Chromelion 6.60 software. The experimental conditions were as follows: a reversed phase silica-based C18 gel as stationary phase (Zorbax SB-Aq, 4.6 × 250 mm, 5 μm pore size), acetonitrile/15 mM aqueous formic acid as mobile phase with gradient elution (5–80% acetonitrile), flow rate 1.00 mL/min, concentration of the investigated complexes: 0.1 mM; 30 μL as injection volume; 25 °C as column temperature; UV–Vis detection set at 225, 254, 280 and 300 nm.

Crystallographic structure determination

X-ray diffraction measurements were performed on a Bruker X8 APEXII CCD diffractometer. Single crystals were positioned at 35, 35, 40, 35, 35 and 40 mm from the detector, and 890, 1086, 1292, 1615, 1278 and 1832 frames were measured, each for 60, 60, 50, 80, 30 and 30 s over 1° scan width for HL, HL·CH3OH, HL·2C2H5OH 1b·1.38H2O·0.25(C2H5)2O, 3a·CH3OH and 3b·0.75H2O, respectively. The data were processed using Saint software [43]. Crystal data, data collection parameters, and structure refinement details are given in Table 1. The structures were solved by direct methods and refined by full-matrix least-squares techniques. Non-H atoms were refined with anisotropic displacement parameters. H atoms were inserted in calculated positions and refined with a riding model. The following computer programs and hardware were used: structure solution, Shelxs-97 and refinement, Shelxl-97 [44]; molecular diagrams, Ortep [45] computer, Intel CoreDuo.

Table 1

Crystal data and details of data collection for HL, HL·CH3OH, HL·2C2H5OH, 1b·1.38H2O·0.25(C2H5)2O, 3a·CH3OH and 3b·0.75H2O.

Complex	HL1	HL4·CH3OH	HL5·2C2H5OH	1b·1.38H2O·0.25(C2H5)2O	3a·CH3OH	3b·0.75H2O
Empirical formula	C23H19N5	C24H23N5O	C26H28ClN5O2	C34H38.25Cl2N5O1.63Os	C33H34BrCl2N5ORu	C32H31.5BrCl2N5O0.75Os
Forward	365.43	397.47	477.98	804.05	768.53	839.13
Space group	P21/n	P21/n	P21/n	P1¯	P1¯	P1¯
Unit cell dimensions
a (Å)	17.0279(16)	11.4565(7)	7.4390(6)	11.2980(9)	9.3254(5)	12.3508(4)
b (Å)	11.1855(11)	13.1693(7)	15.6393(14)	14.4940(12)	13.5216(8)	13.3564(5)
c (Å)	12.0305(11)	13.3620(8)	21.2178(18)	21.3506(18)	13.5306(8)	20.5840(8)
α (°)				99.901(5)	67.646(4)	98.018(2)
β (°)	129.439(4)	95.346(3)	99.794(5)	92.072(4)	89.989(4)	95.312(2)
γ (°)				98.838(5)	85.886(3)	112.859(3)
V (Å3)	1769.6(3)	2007.2(2)	2432.5(4)	3396.2(5)	1573.17(16)	3057.97(19)
Z	4	4	4	4	2	4
λ (Å)	0.71073	0.71073	0.71073	0.71073	0.71073	0.71073
ρcalcd. (g cm−3)	1.372	1.315	1.305	1.573	1.622	1.823
Crystal size (mm3)	0.17 × 0.15 × 0.14	0.08 × 0.06 × 0.03	0.20 × 0.13 × 0.03	0.25 × 0.15 × 0.09	0.30 × 0.20 × 0.10	0.15 × 0.12 × 0.04
T (K)	100(2)	100(2)	100(2)	100(2)	100(2)	150(2)
μ (mm−1)	0.085	0.084	0.190	3.948	1.974	5.686
R1a	0.0551	0.0527	0.0465	0.0386	0.0368	0.0208
wR2b	0.1610	0.1427	0.1284	0.0988	0.0969	0.0506
Goodness-of-fit (GOF)c	1.004	1.029	1.017	1.057	1.036	1.018

R1 = Σ||Fo| − |Fc||/Σ|Fo|.

wR2 = {Σ[w(Fo2 − Fc2)2]/Σ[w(Fo2)2]}1/2.

GOF = {Σ[w(Fo2 − Fc2)2]/(n − p)}1/2, where n is the number of reflections and p is the total number of parameters refined.

Cell lines and cell culture conditions

For cytotoxicity determination, three different cell lines were used: A549, a human non-small cell lung cancer cell line, and SW480, a human colon carcinoma cell line (both kindly provided by Brigitte Marian, Institute of Cancer Research, Department of Medicine I, Medical University of Vienna, Austria) as well as CH1, a human ovarian carcinoma cell line (kindly provided by Lloyd R. Kelland, CRC Centre for Cancer Therapeutics, Institute of Cancer Research, Sutton, UK). Cells were grown as adherent monolayer cultures in 75 cm2 culture flasks (Iwaki/Asahi Technoglass) in Minimal Essential Medium supplemented with 10% heat-inactivated fetal bovine serum, 1 mM sodium pyruvate, 1% non essential amino acids (from 100× ready-to-use stock) and 4 mM l-glutamine but without antibiotics at 37 °C under a moist atmosphere containing 5% CO2 and 95% air. All cell culture media and reagents were purchased from Sigma–Aldrich.

Cytotoxicity assay

Cytotoxicity was determined by the colorimetric MTT assay (MTT = 3-(4,5-dimethyl-2-thiazolyl)-2,5-diphenyl-2H-tetrazolium bromide) as described previously [39]. Briefly, cells were harvested by trypsinisation and seeded into 96-well plates in volumes of 100 μL/well. Depending on the cell line, different cell densities were used to ensure exponential growth of the untreated controls during the experiment: 1.0 × 103 (CH1), 2.0 × 103 (SW480), 3.0 × 103 (A549). In the first 24 h the cells were allowed to settle and resume exponential growth. Then the test compounds were dissolved in DMSO, serially diluted in medium and added to the plates in volumes of 100 μL/well so that the DMSO content did not exceed 0.5%. After continuous exposure for 96 h (in the incubator at 37 °C and under 5% CO2), the medium was replaced with 100 μL/well RPMI 1640 medium (supplemented with 10% heat-inactivated fetal bovine serum and 4 mM l-glutamine) and MTT solution (MTT reagent in phosphate-buffered saline, 5 mg/mL) in a ratio of 7:1, and plates were incubated for further 4 h. Then the medium/MTT mixture was removed and the formed formazan product was dissolved in DMSO (150 μL/well). Optical densities at 550 nm were measured (reference wavelength 690 nm) with a microplate reader (BioTek ELx 808). The quantity of viable cells was expressed as a percentage of untreated controls, and 50% inhibitory concentrations (IC50) were calculated from the concentration-effect curves by interpolation. Every test was repeated in at least three independent experiments, each consisting of three replicates per concentration level.

Results and discussion

Synthesis and characterization of organic compounds and their metal complexes

In the case of the paullones, substitution in position 9 of the ligand backbone led to pronounced differences both in cytotoxic and in enzymatic inhibitory activity. Electron-withdrawing halo substituents had favourable effects on both free ligands and their metal complexes [18,31]. All this prompted us to investigate the effect of substitution in position 8 of the indoloquinoline backbone on antiproliferative activity of ruthenium- and osmium-arene complexes with the ligands HL–HL reported by us previously [38,39]. Briefly, the ligand backbone was assembled in a one-pot reaction from substituted aminobenzylamine and isatin in glacial acetic acid [27] (Scheme 1). The indoloquinolin-6-ones were further chlorinated with POCl3. Treatment of 6-chloro-indoloquinolines with N2H4·H2O gave rise to the indoloquinoline-6-azines. Finally, condensation reaction of 2-acetylpyridine with the corresponding azine afforded the chelating ligands HL–HL. For comparison two ligands with substituents in position 2, namely HL and HL have also been prepared and characterised (vide infra).

Complexes 1a,b−3a,b (Scheme 2) were prepared from [M(η-p-cymene)Cl(μ-Cl)]2, where M = RuII or OsII, and indoloquinolines HL–HL in absolute ethanol in 25–88% yield by exploring the μ-chlorido-bridge splitting reaction. The ESI mass spectra of the complexes showed a single strong peak at m/z 636, 656, 702 for the ruthenium compounds 1a–3a, and m/z 726, 746, 790 for the osmium congeners 1b–3b, which in all cases can be attributed to the [M−Cl]+ ion.

The 1H and 13C NMR spectra are also consistent with the chemical formulae proposed for 1a,b–3a,b (Scheme 2). The number of resonances indicates C1 point symmetry for both investigated ligands and complexes. All complexes are racemates due to the presence of the stereogenic metal centre. They show a typical pattern of the four diastereotopic doublets of the aromatic p-cymene protons between 6.06 and 5.30, and between 6.34 and 5.43 ppm for the ruthenium(II) and osmium(II) complexes, respectively.

One feature of note for this class of substances is the upfield shift of the quaternary, aromatic carbon C6a to about 103 ppm in all complexes (ca. 105 ppm for HL–), caused by the three surrounding electron withdrawing nitrogen atoms N5, N11 and N12. (The atom numbering scheme is depicted in Chart S1).

Most of the resonances did not shift significantly after complexation. As expected, the most noteworthy changes concern the atoms in the close proximity to the metal centre. A downfield shift by 9–11 ppm was observed for C14 (166–168 ppm versus 156–157 ppm in the free ligands), the shift being about 1 ppm larger in the case of the osmium centre. C17 was shifted downfield by about 7 ppm (156 ppm versus 149 ppm), and C18–C20 showed upfield shifts (4–5 ppm), again with a 1 ppm greater change for the osmium complexes. Other shifts in resonances were less significant, e.g. 2–3 ppm upfield shift for C6, C6a and C21. The structures proposed for 1a,b−3a,b were further confirmed by X-ray crystallography.

X-ray crystallography

The results of the X-ray diffraction studies of HL, HL·CH3OH, HL·2C2H5OH, [(η6-p-cymene)Os(L1)Cl]Cl·1.38H2O·0.25Et2O (1b·1.38H2O·0.25Et2O), [(η6-p-cymene)Ru(L3)Cl]Cl·CH3OH (3a·CH3OH), [(η6-p-cymene)Os(L3)Cl]Cl·0.75H2O (3b·0.75H2O) are shown in Fig. 1–3 and Figure S1, respectively. All three complexes crystallised in the triclinic centrosymmetric space group , while the indoloquinoline derivatives in the monoclinic space group P21/n. The asymmetric units of metal complexes consist of two, one and two crystallographically independent complexes of 1b, 3a and 3b, respectively, and-co-crystallised solvent. The complexes have a typical “three-leg piano-stool” geometry of ruthenium(II) and osmium(II)-arene complexes, with an η6 π-bound p-cymene ring forming the seat and three other donor atoms (two nitrogens, N13 and N17, of indolo [3,2-c]quinoline and one chlorido ligand) as the legs of the stool.

The conformation adopted by the indoloquinoline ligand HL is very close to that of HL (in HL·2C2H5OH), while being quite different from that found for HL (in HL·CH3OH), an isomer of HL with methyl group in position 2 of the molecular backbone. This difference is clearly seen by comparing the torsion angle ΘN13−C14−C16−N17 of 3.85(17) and 4.95(19) in HL and HL, respectively, with that of −163.91(13)° in HL. A strong hydrogen bonding interaction is evident between atom N11 acting as proton donor and atom N17i (−x + 0.5, y−0.5, −z + 0.5) of the neighbouring molecule [N11⋯N17i 2.8593(15) Å, N11−H⋯N17i 173.0°] in HL. In HL the atom N11 is also involved in intermolecular H-bonding with oxygen atom of the co-crystallised methanol molecule acting as proton acceptor. The N11−H⋯O1i(x−1, y + 1, z + 1) bond is also strong [N11⋯O1i 2.7491(17) Å, N11−H⋯O1i 170.1°]. The methanol molecule also acts as a proton donor in H-bonding to atom N17ii(x + 0.5, −y + 0.5, z−0.5). The parameters O1⋯N17ii 2.7342(18) Å, O1−H⋯N17ii 169.2° also suggest strong interaction. The conformation adopted by HL is stabilised by intermolecular hydrogen bonding interactions with an ethanol molecule as shown in Figure S1.

Another feature of note is the presence of isolated pairs of molecules or complex cations with offset parallel arrangement stabilised by π-stacking interactions in all compounds studied by X-ray diffraction, but HL·2C2H5OH. The interplanar separation of the indoloquinoline backbones in the pairs is of 3.393, 3.483, 3.448, 3.396 and 3.744 Å in HL, HL·CH3OH, 1b·1.38H2O·0.25(C2H5)2O, 3a·CH3OH and 3b·0.75H2O, respectively (see Figures S2−S6). Figure S7 shows the packing diagram with hydrogen bonding interactions of 3a·CH3OH. The observed packing peculiarities of indoloquinolines and their metal complexes suggest, that these compounds in contrast to indolobenzazepines possess more potential to act as DNA intercalators.

Optical spectra and aqueous solution behaviour

Fig. 4 depicts the UV–Vis spectra of the free ligand HL and its Ru- and Os-complex (1a and 1b, respectively) in methanol. The free ligand exhibits strong absorptions with maxima at 229 and 265 nm due to intraligand π−π∗ transitions. Upon complexation to Ru and Os-arene moieties a strong band at 298 and 303 nm, respectively, appeared. In addition another absorption in the visible region with maximum between 450 and 500 nm is seen, which can presumably be attributed to metal-to-ligand charge transfer.

Fig. 4

UV–Vis spectra of HL (37 μM, green line) and its Ru- and Os complex 1a (36 μM, blue line) and 1b (39 μM, red line) in methanol. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

The aqueous solution behaviour of ruthenium and osmium complexes with respect to hydrolysis was studied at 298 K over 24 h by UV–Vis spectroscopy. The osmium complexes 1b−3b are quite stable in aqueous solution with 1% DMSO, while ruthenium congeners (1a−3a) undergo hydrolysis to a certain extent (Figure S8). To further investigate the nature of changes observed in UV–Vis spectra the experiment was repeated by using higher concentrations of ruthenium complex in 10% DMSO-d in D2O as a solvent and the process was monitored by 1H NMR spectroscopy. There were no signs of decomposition or release of ligands after 24 h, possibly owing to the much higher concentration of the complex in the NMR tube (ca 2 mM compared to 35–40 μM in the UV–Vis experiment). Partial dissociation of the ruthenium(II) complexes can be expected (in case of not very high thermodynamic stability) with decreasing analytical concentrations. Therefore, the aqueous behaviour was further investigated by analytical HPLC. The results for complexes 3a and 3b are illustrated in Figure S9. Whereas the initial ruthenium complex underwent at least two minor reactions, the osmium complex 3b remained intact over 24 h, what is typical for other osmium(II)-arene complexes [37,46,47], which are markedly more inert compared to ruthenium congeners.

Cytotoxicity in cancer cells

To determine the cytotoxicity of the ruthenium(II)- and osmium(II)-arene complexes of indoloquinolines, a colorimetric microculture assay (MTT assay) was used in three human cancer cell lines (A549, CH1, SW480), yielding IC50 values in the 10−7–10−6 M range after continuous exposure for 96 h (Table 2). A549, a generally more chemoresistant cell line, is the least sensitive to all the tested compounds, whereas in CH1 and SW480 cells up to one order of magnitude lower IC50 values were found. The uncomplexed ligands could not be tested because of their insufficient solubility.

Table 2

Comparison of cytotoxicity of ruthenium(II)- and osmium(II)-arene complexes with modified indoloquinoline ligands 1a,b−3a,b vs. 4a,b−6a,b (reported previously [38]) in three human cancer cell lines.

Compound	IC50a (μM), 96 h
	A549	SW480	CH1
1a	2.3 ± 0.7	0.13 ± 0.03	0.22 ± 0.03
1b	1.9 ± 0.2	0.67 ± 0.14	0.22 ± 0.01
2a	2.2 ± 0.7	0.38 ± 0.06	0.34 ± 0.09
2b	2.5 ± 0.4	0.83 ± 0.24	0.23 ± 0.04
3a	1.6 ± 0.3	0.33 ± 0.04	0.20 ± 0.05
3b	2.0 ± 0.2	0.44 ± 0.15	0.20 ± 0.03
4a	2.0 ± 0.4	0.28 ± 0.02	0.19 ± 0.02
4b	3.2 ± 0.4	0.57 ± 0.20	0.19 ± 0.08
5a	9.3 ± 3.4	5.0 ± 1.0	3.8 ± 0.6
5b	3.9 ± 0.5	1.2 ± 0.3	0.55 ± 0.14
6a	7.2 ± 1.7	1.5 ± 0.6	1.3 ± 0.2
6b	7.8 ± 2.1	2.3 ± 0.4	1.0 ± 0.4
[η6-p-cymene)Ru(II)(en)Cl](PF6)b	7.1 ± 1.1	3.5 ± 0.5	4.4 ± 0.9

50% Inhibitory concentrations (means ± standard deviations from at least three independent experiments), as obtained by the MTT assay using exposure times of 96 h.

Taken from Ref. [48].

Comparison of the ruthenium(II) with their osmium(II) analogues shows up to five times higher activity of the former in SW480 cells, whereas in CH1 and A549 cells all IC50 values are in a comparable range. The different substituents, methyl (1a, 1b), chloro (2a, 2b) or bromo (3a, 3b) in position 8, have no pronounced (if any) effect on cytotoxicity of the complexes (Table 2, Fig. 5).

Fig. 5

Concentration-effect curves of complexes 1a, 1b, 2a, 2b, 3a and 3b in the human cancer cell lines A549 (A), CH1 (B), SW480 (C), all determined by the MTT assay using continuous exposure for 96 h.

This is in accordance with structure–activity relationships of related copper(II) complexes with indoloquinolines reported previously [39], showing that major effects on cytotoxicity were only observed for the presence/absence of a methyl group at c14 in the ligand side chain, which is invariably methylated in all the ruthenium(II) and osmium(II) complexes studied here. The copper(II) complexes of the ligands methylated at c14 are 10–50 times more cytotoxic in all tested cell lines. Related ruthenium(II) and osmium(II) complexes reported previously [38], in which the indoloquinoline ligands are substituted in position 2, showed differences dependent on the substituent identity, with the methyl derivative exhibiting the strongest activity, followed by bromo and chloro substituted species. Halo substitution in position 8 is more favourable for cytotoxicity than that in position 2 (up to a factor 13 for the chloro couple 2a,b versus 5a,b and 6.5 for the bromo complexes 3a,b versus 6a,b, respectively), whereas the position of the methyl substituent has nearly no impact on cytotoxicity. [η6-p-cymene)Ru(en)Cl](PF6), [48] taken as a well-known reference compound, is by a factor of 2 in A549, and up to a factor of 10 in CH1 and SW480 cells less cytotoxic (Table 2), while another compound, [η6-p-cymene)Ru(azpy)Cl](PF6), where azpy = phenylazopyridine, which is chemically more closely related to 1–6 shows IC50 values higher than 100 μM in A549 cells [49]. Hence, the indoloquinoline moiety makes a major contribution to the activity of these compounds.

Concluding remarks

Six new ruthenium- and osmium-arene complexes bearing modified indoloquinoline ligands were synthesized. The compounds were comprehensively characterised by 1D and 2D NMR techniques, ESI mass spectrometry, optical spectroscopy and X-ray crystallography (three ligands and three complexes). Complexation of indoloquinoline ligands with ruthenium(II) or osmium(II) resulted in improved solubility in biological media enabling biological assays. Substitution in position 8 seems to be more favourable for the cytotoxic activity than that in position 2, at least for the halogenated indoloquinoline complexes. Especially the most effective compounds with IC50 values in the submicromolar concentration range are worth being studied for their suitability as potential anticancer drugs.