Probing the Mechanism of Action of Bis(phenolato) Amine (ONO Donor Set) Titanium(IV) Anticancer Agents

The need for anticancer therapies that overcome metallodrug resistance while minimizing adverse toxicities is targeted, herein, using titanium coordination complexes. Octahedral titanium(IV) trans,mer-[Ti{R1N(CH2–2-MeO-4-R1-C6H2)2}2] [R1 = Et, allyl, n-Pr, CHO, F, CH2(morpholino), the latter from the formyl derivative; R2 = Me, Et; not all combinations] are attained from Mannich reactions of commercial 2-methoxyphenols (27–74% overall yield, 2 steps). These crystalline (four X-ray structures) Ti(IV)-complexes are active against MCF-7, HCT-116, HT-29, PANC-1, and MDA-MB-468 cancer cell lines (GI50 = 0.5–38 μM). Their activity and cancer selectivity (vs nontumor MRC-5 cells) typically exceeds that of cisplatin (up to 16-fold). Proteomic analysis (in MCF-7) supported by other studies (G2/M cell cycle arrest, ROS generation, γH2AX production, caspase activation, annexin positivity, western blot, and kinase screens in MCF-7 and HCT-116) suggest apoptosis elicited by more than one mechanism of action. Comparison of these data to the modes of action proposed for salan Ti(IV) complexes is made.


■ INTRODUCTION
−7 Titanium(IV) anticancer agents are of contemporary interest as, to the best of our knowledge, there is no reported example of any such species leading to development of acquired cancer cell line resistance, 1−7 as is commonly observed for cisplatin. 8Indeed, salan Ti complexes have demonstrated activity against cisplatin-resistant A2780 ovarian cancer cells. 9This may indicate that titanium(IV) species can target conserved cellular processes that cannot be out-evolved.−12 Unfortunately, historical cellular mechanistic investigations of titanium-based agents have been dogged by paradoxes rather than insight for >40 years, even for those species previously trialed in the clinic. 2The exact processes by which titanium cellular uptake occurs are still not fully defined.For the few anticancer titanium agents where cellular titanium burdens have been determined, 3,6,7 only femtomol (10 −15 mol) amounts of the metal per treated cancer cell have been found (see also the Supporting Information, Figure S62).Currently, only fluorescence imaging techniques (requiring tagged model compounds, sometimes of unknown relevance to the actual titanium drugs) have proved sensitive enough to allow the detection of such levels of titanium in treated cancer cells.−16 This is contrary to early reports, 17 that proposed that the clinical candidate titanocene dichloride (η-C 5 H 5 ) 2 TiCl 2 accumulated mostly in the nucleus and/or chromatin.Contemporary data on substituted titanocenes are not in accord with that proposal. 18The site(s) of cancer cell Ti binding have been the subject of wide debate, 2 and 2013 work by Tshuva et al. added the mitochondrion as a further potential target organelle for titanium agents. 19In 2020, accumulation of phenolate-ligated titanium agents within the endoplasmic reticulum (ER) of MCF-7 cells was proposed. 20The resultant ER stress, causing protein misfolding, was proposed as the preferred "mechanism of action" for agents of motif type B (Figure 1). 5,20The induced misfolded protein was suggested as the trigger for observed MCF-7 apoptosis brought about by such agents.As we recently (2019) identified the additional new motif D as an active titanium agent (Figure 1), 7 we were intrigued to see if our species was related to class B in its mode(s) of action.

■ RESULTS AND DISCUSSION
One common issue for titanium(IV) complexes as potential therapeutics is establishing mild functional group tolerant methodologies that easily install points of derivatization (i.e., via R n , Figure 1) into their ligands.This is desirable for simplifying drug library design in structure−activity studies, and ultimately for optimization of drug delivery and related pharmacokinetic factors.For example, while our own motif D lead ligand 1a (Scheme 1) used simple Mannich chemistry for its preparation, the forcing conditions (125−150 °C, acid solvent) 7 used were incompatible with many useful functional groups for derivative formation.We now find that more concentrated solutions of 2-methoxyphenol derivatives (1 M, MeOH) give acceptable yields (35−88%) of symmetrical ligands 1a−i typically at room temperature (followed, in some cases, by mild warming).These reactions proceed via the intermediate benzoxazines 2, and, in favorable cases (2c, 2e), these can be isolated at ambient temperature (with the mass balance being the derived ligands 1c, 1e, and starting phenol).As both alkenes and formyl groups are tolerated under these new conditions, low cost renewable eugenol (R 1 = allyl) and vanillin (R 1 = CHO) become attractive starting materials.The nonparticipation of the formyl substituent in the reaction of vanillin is ascribed to its conjugation to the phenol OH.Ambient temperature chemoselective Mannich reactions of vanillin are apparently very rare. 21The isolated benzoxazines 2 thus allow for the formation of mixed systems, as in the preparation of 1k.Bis(formyl) 1e readily undergoes reductive amination with amine/pinacol•borane mixtures.Representative morpholine is shown, providing mild diversification to exemplar 1j.One slight complication is the tendency of 1j to form borate complexes with the borane-derived byproducts.However, decomplexation is affected by a simple acid treatment (HCl, 12 M).
The derived bis(phenolato) amine titanium(IV) complexes 3 are easily prepared by the reaction of 1 with Ti(Oi-Pr) 4 (Scheme 2).All these are highly crystalline and easily purified (≥99% by CHN analysis) to levels appropriate for biological studies.Crystallographic studies show the compounds 3 are nearly isostructural in the solid-state (box within Scheme 2 and the Supporting Information, Figures S47−S58).All of structures 3a−c and 3i−j show Ti−O and Ti−N bond lengths in the range expected for Ti(IV) phenolate complexes, Ti−O: 1.858−1.909Å and Ti−N: 2.244−2.269Å (see the Supporting Information, Table S1 and associated CIF files).Complex 3i shows the largest structural difference compared to the parent 3a, which may be pertinent to its poorer biological activity (see later).Region R 2 is poorly tolerant of steric factors; all attempts to complex ligands 1 bearing R 2 = i-Pr to titanium(IV) failed.This distortion is reflected in the N−Ti−N bond angle, which rises from 173.2 to 179.1°in structure 3b vs 3i.The X-ray study of 3k shows the presence of a water solvate hydrogen bonded to one of the morpholine nitrogens, consistent with its lower C Log P value (Table 1).
Concentrations of titanium complexes 3b−k that inhibited cell growth by 50% (GI 50 values) in six cell lines were obtained from MTT studies and are shown in Tables 1 and 2, with comparison to literature 7 3a where possible.
Tables 1 and 2 indicate that greater activity compounds are attained when R 2 = Me.Using the data of Table 1, Figure 2 plots three simple ligand features supporting the following conclusions: (i) ligand 1 polarities of 2.7−4.2(C Log P) are typically associated with the highest activities, (ii) scope exists for a wide range of R 1 volumes to be accommodated (without dramatically lowering overall activity), and (iii) mesomeric electron withdrawing groups at R 1 lower overall anticancer activity.These three main ligand features cointeract providing the observed SAR.Additionally, while opportunities exist for maximizing activity for cancer cell lines (over representative noncancer MRC-5 cells), those factors are too complex to model accurately at present.Selectivity indices of 0.5−18.6 can be derived from the values of Table 1 (see also the Supporting Information, Table S4).
Based on Tables 1 and 2, new agents 3b−c were selected for further scrutiny, focusing mainly on MCF-7 and HCT-116 cell lines.Cell counts and clonogenic assays of 3b−k confirm the cytotoxic nature of antitumor activity detected in the initial MTT studies (see the Supporting Information, Figure S58).
Microscopy imaging and flow cytometry (cell cycle, annexin-V/PI) show that agents 3b−c induce apoptosis in both MCF-7 and HCT-116 cells (Figure 3).Both chromatin condensation and membrane blebbing are imaged in 3b-and 3c-treated cells (Figure 3A and Supporting Information, Figure S72), which are characteristic morphological features of apoptosis. 24The fraction of cells in G2/M phases is significantly increased for both MCF-7 (1.4× control) and HCT-116 (1.8× control) cells for both 3b and 3c (Figure 3B, see also Supporting Information, Figures S68 and S69).Both complexes result in associated increases in the percentage of cells undergoing apoptosis (Figure 3C and Supporting Information, Figure S70), similar to that observed for cisplatin.Additional evidence that 3b−c trigger apoptosis in response to catastrophic DNA damage is provided by γ-H2AX detection and caspase 3/7 activation (Figure 4 and Supporting Information).Both the presence of extensive DNA double strand breaks and enhanced caspase 3/7 activity are fully consistent with apoptosis triggered by DNA damage, causing arrest at the G2/M cell cycle check point, as has been observed in other phenolatebased titanium agents. 7,20he biological features observed in Figures 3 and 4 may correlate with signaling initiated by an extracellular ligand trigger.Evidence to support this thesis can be seen in UV−vis spectra (10 μM, ligand-to-titanium charge transfer band at 330 nm in cell growth medium) of 3c.These show no change over 24 h, indicating that 3c is stable in aqueous media for at least that period.After that time, slow hydrolysis of 3c is detected by UV−vis spectroscopy, amounting to ca. 2% of 3c per day.In the same culture medium containing MCF-7 cells, freshly prepared 3c (10 μM initial concentration) is consumed at a much higher rate: [0.024(2) h −1 ], identical within experimental error to that of the growth of the treated MCF-7 cells [0.028(5) h −1 ] for the first 8 h (Supporting Information, see Figure S59).In the absence of 3c, MCF-7 grows at a faster rate [0.043(4) h −1 ] with doubling times (16 h) that are identical, within experimental error, to literature values. 25MCF-7 cell driven consumption of 3c starts immediately after the addition of 3c and amounts to ca. 10 femtomol per cell at 8 h.After 8 h, the treated MCF-7 cell growth rate recovers partially [0.06(1) h −1 ], and the rate of 3c depletion also increases [to 0.036(3) h −1 ].From this time, populations of dead MCF-7 cells begin to emerge at [0.41(4) h −1 ].The overall behavior is consistent with 3c being the ultimate source of the growth inhibition that is recorded as MCF-7 G2/M arrest and which subsequently promotes apoptosis.The causative agent is clearly 3, or a species derived from it.In that regard, we could detect the formation of free ligand 2c, by LCMS, under aqueous conditions mirroring preparation of biological stock solutions of 3c (200 μM, see Supporting Information, Figure S60).After 2 days, ≤4% ligand was detected, rising to 16% 2c after 5 days.Similarly 3c (200 μM in 4:1 DMSO−D 6 /D 2 O corresponding to the presence of 9.4 M water/4.7 × 10 4 equivalents of D 2 O) remained completely intact for at least 2 days (see Supporting Information, Figure S61).At longer times, smooth formation of a new titanium species over 4 weeks is seen.The +ESI mass ion observed for the hydrolysis species is consistent with the formation of [LTi(OH)(OH 2 )] + where L is the bis(phenolato) dianion of 2c by hydrolytic cleavage of one of 2c ligands.While the ability of complex 3, or derived species, to bind to cells is demonstrated, we have not yet identified their localization in specific organelles.
To try to understand the G2/M block/apoptosis response elicited and elucidate possible molecular (cellular) targets for 3, we undertook a proteomic analysis of the ca.3100 proteins quantifiable by DIA(SWATH) LC-MSMS after MCF-7 cells are treated with 3c (10 μM, 24 h) and lysed (see Supporting Information, Excel File "Proteomics").In comparison to untreated MCF-7, those exposed to agent 3c show significant (−3.49962 log 2 -fold change, ρ 0.0022) downregulation of protein CDK1 (the serine/threonine protein kinase critically involved in cell cycle regulation 26 ).CDK1 is specifically responsible for enabling the G2/M phase transition; thus, onward cell division is impeded by its low availability.However, the concentration of CDK1's archetypal p21 interaction protein, also associated with the normal inhibition of CDK1, 27 was not affected (log 2 -fold changes +0.04 vs the control, Supporting Information, Excel File "Proteomics").Similarly, all the other CDK proteins we could analyze in our study (2/5/6/7/9/11) were only modestly upregulated (log 2fold change +0.20 to +1.40, Supporting Information, Excel File "Proteomics").CDK2 (which facilitates nuclear export) was the next most affected protein, consistent with DNA damage being present at the G2/M checkpoint. 28Selective CDK1 inhibition is rare 29 being previously seen only for competitive ATP binding to that kinase.However, as the Wee1 and Cdc25 proteins (dictating the "hold" or "go" signals to CDK1) could not be analyzed in our protein set, definitive confirmation of CDK1 inhibition would need future work.Proteomic pathway analysis suggests a modified immune response as an alternative likely inhibition process.Immune response evasion is a   20 and Hammett parameters 23 from literature sources.hallmark of cancer, and the suppression of this innate ability by 3c is also a viable possibility.Typically, growth factor signaling is required to initiate MCF-7's entry into its division cycle.Interaction, at the cell surface, with receptor tyrosine kinase (RTK) sites initiates all subsequent events.The RTK-linked membrane bound ras-GDP signals to the mitogen-activated protein kinase (MAPK)/ extracellular signal-regulated kinase (ERK) cascade causing ERK translocation to the nucleus where multiple target interactions are possible.In line with this proposal, in 3c treated MCF-7 cells, the Ras GTPase-activating protein is significantly downregulated (−2.00630 log 2 -fold change, Supporting Information, Excel File "Proteomics"), with a smaller, but notable effect on MAPK-1 (−0.29654 log 2fold change).We note that perturbation of MAPK signaling has been noted in previously studied titanium complexes. 18hree of the key downregulated proteins we identify (CDK1, PABPC, and NQ01) have equivalent genes that are implicated in anticancer activity in related RNA studies of Tshuva and co-workers. 20Our observation of CKD1 downregulation is in agreement with that earlier study, but their observation of PABPC upregulation (at 24 h) is different from our own findings, even though the cell line studied is identical (MCF-7) in both cases.In the study by Tshuva and co-workers, 20 NQ01 regulation depends on the time course of exposure (upregulated between 15 and 24 h, then downregulated at 48 h).Our proteomic analyses imply very significant downregulation for this protein/gene at 24 h.As NQ01 is a superoxide scavenger, its downregulation is consistent with the reactive oxygen species (ROS) generation uptick we also observe (see later).To support our proteomic studies, we also investigated the qualitative behavior of the 23 proteins that we could map to the earlier RNA study 20 (Table 3).The expression trend analysis within our proteomic study supports Tshuva's implication of the involvement of the mitochondrial translation pathway.However, it did not completely support the suggested protein processing involvement in the ER pathway, even though MCF-7 was used in both studies.Overall, our data analysis indicates the possibility that perturbations of a rich and diverse pharmacology are viable, even for closely related titanium agents.
As we identified apoptosis as the mechanism of cell death, and Bcl-2 and Mcl-1 are key pro-survival proteins, we interrogated by western blot, changes in expression of these key cancer survival (antiapoptotic) proteins (Bcl-2 and Mcl-1) in MCF-7 and HCT 116 cells following exposure to 3b/3c (Figure 6 and Supporting Information, Figures S74 and  .Analysis of the growth inhibitory activity (mean ± SD GI 50 value in μM, MTT assay, 72 h) of complexes 3a−k vs selected ligand features of 1: (A) C Log P; (B) V(R 1 ); and (C) σ p (R 1 ).Key: (Royal blue) ■ MCF-7, (mustard) ×HCT-116, (gray) ΔHT-29, (yellow) •PANC-1, (pale blue) ◆ MDA-MB-468, (green) �MRC-5, and (black) ---average GI 50 value for all six cell lines studied.S75). 30Proteins Bcl-2 31,32 and Mcl-1 33,34 work together to counter apoptosis by limiting mitochondrial membrane release of cytochrome c; 35,36 they are typically overproduced in cancer cells, and their expression is closely associated with poor prognosis and drug-resistance.Profoundly reduced expression of Bcl-2 and Mcl-1 in cells treated with 3b/3c is in line with apoptosis induction and reduced immune response evasion indicated in the proteomic study.This pathway is also consistent with receptor-mediated (MAPK/ERK) apoptosis 37 (see later).
In addition, because we detect DNA double strand breaks and metal complexes are known to damage DNA through ROS generation, we investigated whether 3b/3c-induced MCF-7 and HCT 116 intracellular ROS formation. 38ROS are formed in mitochondria by electron chain reduction of O 2 to form superoxide or by peroxisomes (through fatty acid oxidation) or by ER oxidation of proteins. 39Treated MCF-7 (3c, 10 μM, 24 h, Figure 6) shows nearly double the ROS of untreated controls.This is consistent with the NQ01 downregulation observed in our proteomic study.Excess ROS populations are known to promote cytochrome c release, through oxidation of mitochondrial pores, causing caspase activation and ultimately apoptosis. 40Consistently, ROS generation is noted in both HCT 116 and MCF-7 cells treated with 3b as well as 3c (Figure 5).
Complexes 3b−c induce DNA damage.Poly(ADP-ribose) polymerase 1 (PARP1), an ADP-ribosylating enzyme becomes activated upon binding to DNA single strand and double strand breaks and is essential for initiating various forms of DNA repair. 42PARP is also a substrate for caspases�activated during apoptosis.Caspases cleave PARP during apoptosis and thus cleaved PARP emergence accompanied by downregulation of whole PARP has become a marker for apoptosis. 43MCF-7 cells treated with 3b or 3c (10 μM, 24 h) show, through western blots, highly reduced PARP populations (13−15% of control).Similar behavior is seen for the same two complexes in the HCT-116 cell line (12− 13% of control).The reduced PARP (and associated increased cleaved PARP) are clear markers for DNA damage-induced apoptosis (see the Supporting Information, Figure S73).
The heterogeneous nuclear ribonucleoprotein M is strongly downregulated (−4.15369 log 2 -fold change, ρ 0.0005) in our proteomic study.The latter nuclear proteins sequester and transport RNA out of the nucleus and are closely associated with cell cycle regulation at the G2/M DNA checkpoint.Heterogeneous nuclear ribonucleoproteins regulate the surface receptor glycoprotein CD44 44 that acts via a range of signaling kinases, but especially Ras-MAPK pathways, terminating in gene transcription at the nucleus.We thus used western blot techniques to quantify the marker ERK1/2 protein of the Ras-MAPK cascade.ERK1/2 cascade activation is typically initiated by membrane receptors such as RTKs, G proteincoupled receptors (GPCRs), and ion channel receptors, for example. 45These receptors transmit signals by recruiting adaptor proteins (e.g., Grb2) and exchange factors (e.g., son of sevenless, SOS), which in turn activate Ras.The active, GTPbound Ras then delivers the signal by activating the Raf-1, B-Raf, and A-Raf (Rafs) protein kinases within the MAPK cascade. 45,46Ras/Raf/MEK/ERK1/2 signaling is triggered via a small GTPase-mediated activation of activated tyrosine receptors and cytoplasmic kinase signaling cascades. 47The key point of activation is the transmission of a signal from tyrosine kinase receptors, including the epidermal growth factor receptor, which then recruit SOS via intracellular Shc and Grb2 domains, catalyzing the conversion of inactive Ras/ guanosine diphosphate to an active Ras/guanosine triphosphate complex. 47,48This ERK1/2 cascade is also a major signaling system that regulates not only many activated cellular activities, most notably proliferation, differentiation, and survival, but also apoptosis and stress response. 45The effect of compounds 3b−c on ERK1/2 within the MAPK cascade is shown schematically in Figure 6 (see also the Supporting Information, Figures S74 and S75).
The reduction of ERK1/2 induced by 3b−c is opposite to the increase seen in our earlier studies of the action of chiral titanocenes. 7However, in that case paraptotic cell death is induced by acceleration of cellular processes (paraptosis) vs the inhibition (apoptosis) seen here.This behavior is exemplary of the increasing repertoire of compounds that mediate ERK activation leading to apoptosis.While activation of ERK1/2 typically promotes cell proliferation, some compounds induce ERK activation while exerting antiproliferative effects.It is acknowledged that the mechanisms underlying ERK1/2-mediated cell death are still to be fully defined. 30,38

CONCLUSIONS
To conclude, we describe the synthesis of a series of octahedral titanium(IV) complexes whose anticancer activity and putative  51 demonstrated selective inhibition of multiple kinases by metal (Zn and Cu) complexes in human carcinoma cell lines.Overall, a picture emerges that titanium(IV) anticancer agents evoke responses in multiple, highly conserved, cellular processes that significantly limits the evolution of resistant cancer cell types.Thus, research into Ti(IV) complexes for treatment of intractable malignancies is worthy of continued development, to progress full elucidation of molecular/cellular mechanisms, preclinical biodistribution, pharmacokinetics, tolerability, and efficacy studies, and eventual clinical evaluation.
■ EXPERIMENTAL SECTION Chemical Synthesis.Reactions were carried out under appropriate conditions using commercial reagents of ≥98% purity.Solvents were dried (4A molecular sieves) when appropriate.TLC analyses were performed on foil-backed plates coated with Merck silica gel 60 F 254 .Ultraviolet light and basic aqueous potassium permanganate were used to visualize the plates.Liquid chromatography was performed using forced flow (flash column) techniques with the solvent systems indicated.The stationary phase used was silica gel 60 (220−240 mesh) supplied by Fluorochem.Fouriertransform infrared spectra were recorded on a Bruker Alpha Platinum spectrometer.Nuclear magnetic resonance spectra were recorded on Bruker AV(III)400 (400.1 MHz), Bruker AV400 (400.1 MHz), Bruker Ascend 400 (400.1 MHz), or Bruker Ascend 500 (500.1 MHz) spectrometers at ambient temperature (unless otherwise stated).Chemical shifts are quoted in parts per million (ppm).Coupling constants (J) are quoted in hertz.Couplings are written using the following abbreviations: br (broad), s (singlet), d (doublet), t (triplet), q (quartet), m (multiplet), and app (apparent).Carbon NMR multiplicities and connectivities were assigned using DEPT and the relevant 2D NMR experiments.Mass spectrometry was performed using a VG Micromass AutoSpec spectrometer (EI) or Bruker MicroTOF (ESI) instrument as noted.Theoretical HRMS molecular weights were taken from the spectrometer output file, and for HRMS analyses, deviations from expected values (σ) are given in ppm.Melting points were measured on a Gallenkamp melting point apparatus and are uncorrected.Liquid chromatography−mass spectrometry (LCMS) analysis was performed by using an Agilent 1260 Infinity HPLC with a 6120 Quadrupole mass spectrometer with a multimode source.Chromatography conditions: XBridge C18, 3.5 μm, 2.1 mm × 30 mm column.Mobile phase A: 0.1% ammonia in water; mobile phase B: acetonitrile.Flow rate: 0.8 mL/min in a gradient of 5−95% mobile phase B over 3.5 min, with UV detection at 210−400 nm, reported at 254 nm.Column temperature was 40 °C.Data on X-ray diffraction were gathered via the University of Nottingham, X-ray Crystallography Service.Appropriate single crystals were selected and submerged in an inert oil.After that, the crystal was fixed to a glass capillary and fastened to a goniometer head.Data were collected on a Bruker X8 Apex or an Agilent Supernova diffractometer using graphite-monochromated Mo−K α radiation (λ = 0.71073 Å) using 1.0°ϕ-rotation frames.The crystal was cooled to 100 K by an Oxford Cryostream low temperature device.
Additional compounds and procedures for this publication are described in the Supporting Information.
General Procedure C: Synthesis of Bis(amine) Tetrakis-(phenolate) Titanium(IV) Complexes (3).The bis(phenol) ligand (1) (1 equiv) was dissolved with stirring (3 min) in toluene (ca. 4 mL per mmol of bisphenol).To the resulting solution, titanium(IV) isopropoxide (0.6 equiv) was added dropwise, and the mixture left to stir (4 h) at RT under a nitrogen atmosphere.The solvent was then removed by trap−trap distillation (ca.1−2 mbar, RT) to afford compound (3) as an orange solid.The product was crystallized by liquid−liquid diffusion using suitable solvents upon cooling to 4 °C.Compound 3a was available from a literature procedure. 7All biologically tested 3 were >99% pure by CHN analysis.
Cancer Biology.Full details of all biological studies are given in the Supporting Information.Exemplary annexin-V and determination of γ-H2AX assays are given below.
Annexin-V Assay.Cells were seeded in 10 cm diameter Petri dishes with 10 mL of complete medium at a density of 4 × 10 5 cells The cells were incubated for 24 h to allow cell attachment.Following treatment (72 h; 10 μM) with a test compound, the cells were trypsinized with 300 μL of 1× trypsin−EDTA and pooled in a total of 1 mL of complete growth medium.Afterward, the cells were resuspended in 2 mL of cold medium and decanted into labeled FACS tubes and kept on ice to allow recovery from any damage caused by trypsin.Cells were centrifuged at 1200 rpm (Beckman Coulter Allegro centrifuge) for 5 min at 4 °C.The supernatant was discarded, and the pellet broken up by gently flicking the tube.Cold PBS (2 mL) was added, and the cells were centrifuged as before.The supernatant was discarded, and the pellet broken up by gently flicking the tube.Thereafter, 100 μL of 1× annexin binding buffer and 5 μL of annexin V FITC was added to each tube.The tubes were briefly vortexed and kept at room temperature for 15 min in the dark.Annexin binding buffer (400 μL; 1×) and 10 μL of 50 μg/mL PI solution were added to each tube, which was vortexed and kept for 10 min at room temperature in the dark prior to analysis on the flow cytometer.Samples were analyzed within 1 h of preparation to avoid sample deterioration using a FC500 Beckman Coulter flow cytometer, and 20,000 events were evaluated for each sample.The results obtained were analyzed using WEASEL software. 7,52etermination of γ-H2AX Foci Perturbation.For γ-H2AX detection of DNA double strand breaks with concurrent cell cycle analysis, cells were seeded in cell culture dishes at densities of 3−5 × 10 5 cells/dish in 10 mL of medium.Following 72 h treatment, the cells were harvested and pelleted by centrifugation, resuspended and washed (2×) in PBS, pelleted again by centrifugation and fixed in 500 μL of 1% methanol-free formaldehyde in PBS (5 min; room temperature).The cells were permeabilized by adding 500 μL of 0.4% Triton-X-100 in PBS.FBS (1% in PBS; 1 mL) was then added to cells with gentle mixing before incubation at room temperature for 30 min.The cell suspensions were centrifuged and supernatants aspirated.Primary antibody (1°Ab, p-Histone γ-H2AX) solution (200 μL, 1:3333 in 1% FBS) was added to each tube and the samples incubated (1.5 h).PBS (1 mL) was added, and the samples were centrifuged and supernatants discarded.Secondary antibody (2°Ab, Alexa Flour 488 goat secondary antimouse) was introduced (200 μL, 1:750 in 1% FBS) and the samples incubated for 1 h at room temperature before addition of 1 mL of PBS.Samples were centrifuged and supernatant again discarded.The cells were resuspended in PBS containing 300 μL of 50 μg mL −1 PI and 0.1 mg mL −1 RNaseA.Analyses of cells (20,000 events per experimental sample) by flow cytometry followed a 10 min incubation. 18

Figure 1 .
Figure 1.Contemporary titanium(IV)-based anticancer motifs A−D, with representative in vitro activity values against HT-29 and MCF-7 cell lines. 3−7 Cisplatin shows activities of ∼20 and ∼8 μM, respectively, in the same two cell lines.Variation of the substituents (R n ) is possible for each motif (for specific examples see refs 3−7), but typically R n are simply methyl, halogen, or alkoxy units.Scheme 1. Preparation of Bis(phenolato) Amines (1a−k) Used in This Study a

Figure 5 .
Figure 5. Change in species known to either promote ROS or apoptosis (downregulate proteins Bcl-2 and Mcl-1) through cytochrome c release from mitochondria vs controls; see the Supporting Information (Figures S74 and S75) for primary data.The structure of cytochrome c is from ref 41 and used with permission from Elsevier for reproduction, license number, 5531400991823; issue year, 2023.

Table 1 .
Growth Inhibitory Activity of 3a−k (Mean ± SD GI 50 Values in μM, MTT Assay, 72 h) as a Function of Changing R 1 (R 2 = Me in all Cases) and Treated Cell Lines a Data generated from ≥3 independent trials; n = 8 per experimental condition per trial.b Ligand C Log P values from ChemDraw (ver.20); R 1 substituent volumes 22 and Hammett parameters 23 from literature sources.c No Hammett parameter is available for CH 2 (morpholino), the value for CH 2 NMe 2 is given. 23d Average of allyl and Me values. a

Table 2 .
Growth Inhibitory Activity of 3h−j (Mean ± SD GI 50 Value in μM, MTT Assay, 72 h) as a Function of Changing R 1 (R 2 = Et in all Cases) and Treated Cell Lines a a Data generated from ≥3 independent trials; n = 8 per experimental condition per trial.b Ligand C Log P values from ChemDraw (ver.20); R 1 substituent volumes