Ferrocene modified analogues of imatinib and nilotinib as potent anti-cancer agents

Abstract

The unique features of ferrocene and the need for development of targeted anticancer drugs inspired the design, synthesis and biological evaluation of ferrocenyl modified tyrosine kinase inhibitors by replacing the pyridyl moiety in imatinib and nilotinib generalized structures with a ferrocenyl group. A series of seven new ferrocene analogues were synthesized and evaluated for their anticancer activity in a panel of bcr–abl positive human malignant cell lines using imatinib as a reference drug. The metallocenes exhibited a dose-dependent inhibition on malignant cell growth with varying antileukemic activity. The most potent analogues were compounds 9 and 15a showing comparable or even superior efficacy to the reference. Their cancer selectivity indices suggest a favorable selectivity profile, indicating a 250 times higher preferential activity of 15a towards malignantly transformed K-562 cells and an even twice greater one (500) of 9 in the LAMA-84 leukemic model as compared to the normal murine fibroblast cell line.

Analogues of imatinib and nilotinib, where the pyridine ring is replaced by a ferrocenyl moiety, were synthesized and evaluated for their anti-cancer activity. Two of the metallocenes exhibited highly selective tumor cell growth inhibition.

Introduction

Ferrocene (Fc) is an 18-electron organometallic complex composed of a central iron(ii) ion sandwiched between two parallel cyclopentadienyl (Cp) ligands. Due to its unique geometry and chemical features, the ferrocene moiety has been an enticing research subject in organometallic, material and medicinal sciences.^1,2 Its unusual structure has been exploited for a number of reasons: the high stability in aqueous and aerobic media;³ the planar chirality of the 1,2-disubstituted Fc scaffolds, which found broad application in the development of catalysts with high asymmetric induction abilities;⁴ the fast and reversible redox change of the extremely stable Fc/Fc⁺ scaffold that makes it a suitable building block for the design of switchable functional systems in supramolecular chemistry and might account for potential selectivity in anticancer activity;^2,5,6 the aromaticity and high lipophilicity of the Fc core make the structure readily membrane permeable and suitable for molecular tailoring;⁷ and most importantly, the non-toxicity and the manifold effects of ferrocene derivatives on cellular function lend them a special place in medicinal and pharmaceutical chemistry.^5,8

The initial prominent anticancer activity of Fc analogues has been displayed by some simple ferrocenium salts (Fig. 1), which continue to be an attractive research topic.^9,10 The exact mechanism of their antiproliferative action is still not fully understood, however, ROS (reactive oxygen species) generation has been proposed.^9,11 One of the distinguishing features of malignant cells is the enhanced production of reactive oxygen species (ROS) that makes them more vulnerable to oxidative stress. Since the ferrocene structure exists in a state of redox equilibrium, it can elicit ROS formation under physiological conditions and induce oxidative damage to DNA and protein biostructures.^3,9,12 Furthermore, the occurrence of iron-dependent lipid peroxidation leads to membrane destabilization and initiates a specific kind of non-programmed cell death pathway, called ferroptosis.^13–15 Functionalization of the metal complex by attaching various pharmacophore groups to the ferrocene system has been another trend in the design and synthesis of innovative ferrocene derivatives with extended biological activity.^16–19

Fig. 1. Structures of ferrocene (Fc), ferrocenium salts (A⁻ = FeCl₄⁻, I₃⁻), ferrocifen (R = H), hydroxyferrocifen (R = OH), and the TKIs imatinib and nilotinib.

The concept of using a ferrocenyl group as a bioisostere of phenyl and heteroaromatic rings is developed based on the lipophilicity of the Fc's bulky sandwich structure, which offers potentially better fitting into hydrophobic pockets.^5,9,20 The true potential of this idea is displayed after the discovery of ferrocifen and hydroxyferrocifen (Fig. 1) – structural analogues of tamoxifen and hydroxytamoxifen, respectively, where a phenyl ring is replaced by a ferrocenyl moiety.²¹ Interestingly, the ferrocifens exert high antiproliferative activity both on hormone-dependent and hormone-independent breast cancer cells, whereas the activity of the prototype SERM (selective estrogen receptor modulator) drug tamoxifen is restricted only to the hormone-sensitive types.²¹ Ferrocifens are currently in pre-clinical trials. However, their high lipophilicity restricts solubility in a physiological environment and requires suitable formulations for in vivo administration like the use of lipid nanocapsules.²¹ Enhancement of the hydrophilicity of ferrocenes, meeting the challenges of in vivo studies, was further developed towards ferrocene-based imidazoles, triiron vinyliminium complexes, and oxindole derivatives.²² Ferrocene has also been used as a bioisostere for the phenyl ring or heteroaromatic groups in GPCRs modulators, kinase inhibitors and histone deacetylase (HDAC) inhibitors.²³ The displayed activity against drug resistant cancer cells, combined with selectivity, justifies the further attempts to apply the ferrocenyl moiety as a template for the design and synthesis of novel anticancer agents.²⁴

The discovery of imatinib (Glivec®/Gleevec®, Fig. 1) and nilotinib (Tasigna®, Fig. 1) represents a milestone in the field of oncopharmacology, giving birth to genome-targeted anticancer therapies.²⁵ The presence of site-specific mutations in various proliferative diseases makes the molecular profiling of tumor pathology possible, thus giving specific protein drug targets to medicinal chemists.²⁶ Chronic myeloid leukemia (CML) is a hematopoietic disorder characterized by a pathognomonic reciprocal translocation between chromosomes 9 and 22: t(9;22)(q34;q11), resulting in the formation of a fusion Bcr–Abl oncoprotein.^27,28 The prototype drug imatinib and its advancer nilotinib are very potent, competitive Bcr–Abl tyrosine kinase inhibitors (TKIs), which act on the ATP binding site in the catalytic domain of the Abl kinase.^25,29,30 The latter consists of several regions – a hydrophobic pocket, a hydrophobic channel, an adenine, a sugar, and a phosphate-binding region.³¹ Crucial structural elements for the active site binding interactions are: the pyridyl–pyrimidine fragment in the hydrophobic pocket of the hinge region; the aniline-NH hydrogen bonding with the Thr315 gatekeeper residue; and the aromatic amide bond serving as an anchoring group.³² Substitution point mutations in the Thr315 residue (specifically T315I) result in the elimination of a key OH-group required for the hydrogen bonding of the imatinib to the enzyme and mediate resistance to Ist (imatinib) and all of the IInd generation (nilotinib, bosutinib, dasatinib) TKIs.^33,34

Inspired by the unique properties of ferrocene and the need for development of advanced Bcr–Abl targeted drugs, this study is aimed at the design and synthesis of potential ferrocenyl modified TKIs. Herein, the bioisostere concept is further developed by replacing the pyridyl moiety in imatinib and nilotinib generalized structures with a ferrocenyl group. A series of seven new compounds were synthesized and evaluated for their antiproliferative effect on a panel of (Ph+) human malignant cell lines of CML origin. To the best of our knowledge, this is the first study of ferrocenyl modified imatinib and nilotinib structural analogues.

Results and discussion

Synthesis

Our initial investigations were focused on the construction of the key ferrocenyl-pyrimidine fragment. Optimization of the synthetic pathway was envisaged by the preparation of simplified model compound 3 (Scheme 1). Overnight condensation of acetylferrocene with dimethylformamide dimethyl acetal (DMF-DMA) in DMF at 125 °C afforded (E)-3-(dimethylamino)-1-ferrocenyl-2-propen-1-one 1.³⁵o-Toluidine was converted into its hydrochloric salt and subsequently melted with cyanamide to give 1-(o-tolyl)guanidine hydrochloride 2 in quantitative yield.^36,37 Finally, the pyrimidine core was constructed by cyclocondensation of enaminone 1 and guanidinium salt 2. Optimal conditions were achieved with K₂CO₃ as a base in refluxing n-butanol for 48 h. Purification by column chromatography gave pure product 3 in 57% yield.

Scheme 1. Optimized synthesis of the key ferrocenyl-pyrimidine fragment.

Our next goal was the synthesis of ferrocenyl modified analogues of imatinib. Reaction of commercially available 2-methyl-5-nitroaniline with cyanamide furnished the corresponding guanidine 4 following the already described two-step procedure (Scheme 2). It was planned construction of the pyrimidine ring, subsequent reduction of the nitro group, and finally amide bonding with benzoyl chloride. Unfortunately, the cyclocondensation step was not successful due to decomposition of the product during the purification. We assume that the ferrocene scaffold is not stable in the presence of a nitro group.

Scheme 2. Synthesis of compound 9.

The original route to imatinib analogues was replaced with the alternative where the pyrimidine ring formation is the final synthetic step (Scheme 2). Initially, an amide bonding of 4-methyl-3-nitroaniline with benzoyl chloride to nitrobenzamide 6 was accomplished. Reduction of the nitro group to amino in analogous structures required either expensive catalysts for hydrogenation,³⁷ or the use of excess SnCl₂, resulting in moderate yields after purification.³⁸ We were able to overcome these problems by using Fe dust and aqueous NH₄Cl in boiling ethanol media.³⁹ This cheap, environmentally friendly and easy to perform reaction afforded N-(3-amino-4-methylphenyl)benzamide 7 quantitatively without purification. The latter was transformed into its guanidinium salt 8 in 96% yield following the already described procedure. Subsequent cyclocondensation with enaminone 1 afforded the target compound 9 in 39% yield after flash column chromatography.

We extended our investigations to ferrocenyl modified nilotinib structural analogues benefiting from our previous synthetic experience. Nitration of methyl 4-methylbenzoate gave the starting nitrobenzoate 10, which was successfully reduced under the above described Fe/NH₄Cl reaction conditions to the corresponding amine 11 (Scheme 3). The reaction again proceeded with excellent yield without the need for further product purification. The subsequent transformation of the aniline into guanidinium salt was accomplished quantitatively.³⁷ Cyclocondensation with enaminone 1 proceeded with moderate yield and interestingly due to the prolonged reaction time a transesterification took place. Basic hydrolysis furnished the desired acid, which was submitted to an amide coupling reaction with aniline and a number of amino acid esters by using N-(3-dimethylaminopropyl)-N′-ethylcarbodiimide hydrochloride (EDC) and 1-hydroxybenzotriazole hydrate (HOBT) as activating agents, and CH₃CN as a solvent. The desired target compounds 15a–e were isolated after flash column chromatography on silica gel with good to excellent yields. The structural identity of each compound was confirmed by ¹H and ¹³C NMR spectroscopy including mass spectrometry.

Scheme 3. Synthesis of compounds 15a–e.

The stability of representative compound 15a was monitored by ¹H NMR spectrometry over 4 days in DMSO-d₆ : PBS (pH = 7.4) 4 : 1 (v : v). No sign of decomposition and/or iron oxidation was observed.

Antiproliferative activity

The newly synthesized compounds were subjected to a colorimetric MTT assay to evaluate their antiproliferative potential in a panel of bcr–abl positive human leukemia cell lines. The prototype TKI imatinib was used as a reference drug (Table 1). Column graphs were plotted (mean ± SD) against untreated control to compare the cytotoxicity of the tested compounds at each treatment concentration (Fig. 2). Finally, cancer selectivity indices (CSI) were estimated for the two most promising compounds (9 and 15a) against the normal fibroblast murine cell line CCL-1 (Table 2).

In vitro cytotoxicity (mean IC₅₀ value, [μM ± SD]) of the tested compounds against bcr–abl positive human leukemic cell lines³⁸ and normal murine fibroblast cells.

Cell line/fusion genes Cmpd	AR-230a e19a2 (p230)	BV-173b e13a2/B2-a2 (p210)	LAMA-84c e14a2/B3-a2 (p210)	K-562d e14a2/B3-a2 (p210)	CCL-1e
3	>400	70.1 ± 6.4	118.6 ± 7.4	66.5 ± 4.3	—
9	5.7 ± 0.9	48.0 ± 7.2	0.9 ± 0.2	28.9 ± 4.1	>200
15a	30.3 ± 5.5	10.1 ± 2.7	0.4 ± 0.1	0.8 ± 0.2	>200
15b	123.7 ± 12.2	64.6 ± 4.1	72.9 ± 3.3	11.7 ± 1.2	—
15c	>200	67.2 ± 5.1	199 ± 10.1	>200	—
15d	96.8 ± 8.3	>400	142.5 ± 13.4	56.3 ± 4.7	—
15e	71.4 ± 7.8	36.7 ± 4.3	88.7 ± 5.4	74.3 ± 6.2	—
Imatinib	4.7 ± 1.0	22.8 ± 4.8	0.5 ± 0.05	45.5 ± 5.0	—

^aCML in blast crisis (bcr–abl+).

^bB-cell precursor leukemia (bcr–abl+).

^cCML in blast crisis (bcr–abl+).

^dBlast phase CML (bcr–abl+).

^eNormal murine fibroblast cells.

Fig. 2. Inhibition of cell viability by the lead compounds in the series of the newly synthesized ferrocene-based TKIs and the reference drug imatinib in a panel of bcr–abl+ leukemia cell lines.

Estimated CSIs (cancer selectivity indices) for the two lead compounds 9 and 15a in the screened leukemic cell lines.

Cmpd	CSI AR-230a (IC50 CCLe/IC50 AR-230)	CSI BV-173b (IC50 CCL/IC50 BV-173)	CSI LAMA-84c (IC50 CCL/IC50 LAMA-84)	CSI K-5623d (IC50 CCL/IC50 K-5623)
9	≈35	≈4.1	≈222	≈6.9
15a	≈6.6	≈20	≈500	≈250

^aCML in blast crisis (bcr–abl+).

^bB-cell precursor leukemia (bcr–abl+).

^cCML in blast crisis (bcr–abl+).

^dBlast phase CML (bcr–abl+).

^eCCL – Normal murine fibroblast cells.

At first glance, all of the screened experimental TKIs exhibited a dose-dependent inhibition on malignant cell growth, however, their antileukemic activity varied within a wide range in a cell-line specific manner (Table 1, Fig. 2). The least effective within the series were compounds 3 and 15d, whereas compounds 9 and 15a can be outlined as the most potent analogues with comparable or even superior cytotoxicity to the reference drug in all leukemic models. Remarkably, the lead compound 15a produced a 50-fold higher antitumor efficacy against K-562 leukemia blast cells as compared to imatinib (IC₅₀ 0.8 μM vs. 56.3 μM, respectively).

The prominent effects of both 9 and 15a may be related to their closest structural resemblance to the TKIs imatinib and nilotinib, where the pyridine ring is replaced with a ferrocenyl moiety (Fig. 3). The functional groups on the aromatic ring are eliminated, however, the crucial structural elements for active site interactions are preserved: the aniline-NH hydrogen bonding with the Thr315 gatekeeper residue and the aromatic amide bond serving as the anchoring group are the same as in the parent molecules. The only structural modification in the newly designed 15a and 9 derivatives is the replacement of the pyridil-pyrimidine fragment with a ferrocenyl-pyrimidine one.

Fig. 3. Comparison of the structures of imatinib and nilotinib with compounds 9 and 15a.

The poor activity of compound 3 can be explained by the missing aromatic amide bond serving as an anchoring group in the parent compound. We were challenged to further explore the functional significance of this structural element by replacing the aromatic amide bond with an amino acid amide bond in the corresponding 15b–e analogues. Since peptides and peptide-related structures have a wide variety of physiological and pharmacological actions, the concept of peptidomimetics was aimed by the introduction of aromatic amino acids or histidine in place of the benzamide motif. Additionally, we aimed at the improvement of the compounds' pharmacokinetic properties by possible transformation of the amino-acid moieties into the corresponding salts. Unfortunately, compounds 15b–e displayed poor to moderate antimyeloid activity. Notable cytotoxicity was only observed in the case of the phenylalanine derivative 15b, whose half-inhibitory concentration is nearly 5 times lower than that of imatinib in the K-562 cell line (11.7 μM vs. 45.5 μM).

As shown in Table 2, CSIs were calculated for the two lead compounds 9 and 15a, defined by the ratio between their IC50 against the normal CCL-1 cell line and their respective IC50 values in the four malignant leukemia types. Once again, a slightly more favorable selectivity profile was displayed by 15a, indicating a 250 times higher preferential activity towards the K-562 cell line and an even twice greater one (500) against the LAMA-84 leukemic model.

The different susceptibility of the four bcr–abl+ leukemia cell lines to the newly synthesized TKIs and imatinib may be associated with the formation of alternative bcr–abl fusion oncogenes, whose enzymatic transcripts have been shown to vary in the level of expression, tyrosine kinase activity and response to targeted therapy.^40–44 For example, the AR-230 cells producing the p230 (e19a2) variant of the protein showed the lowest chemosensitivity towards most experimental compounds, but preserved a descent one towards compound 9 and imatinib. On the contrary, the proliferation of the BV-173, LAMA-84 and K-562 cell lines with a p210 phenotype was inhibited in a more consistent manner by most analogues in the series but varied in the imatinib reference treatment. However, drawing a correlation between the tyrosine kinase type and the observed degree of effective enzyme inhibition goes beyond the scope of this study and may be addressed in future research. Additionally, the high lipophilicity of the compounds restricts their bioavailability and requires further synthetic modifications aiming at improvement of their pharmacokinetic properties.

Conclusions

In the present study, we have reported on the design, synthesis and in vitro antileukemic activity of novel ferrocene-based anti-cancer agents. Minor changes in imatinib (9) and nilotinib (15a) structures where the pyridine ring is replaced with a ferrocenyl moiety resulted in a highly selective tumor cell growth inhibition, comparable or surpassing that of the parent molecule. Our promising initial results open the field for ferrocene derivatives in tailoring targeted drugs as an effective approach to boost their efficacy and circumvent arising resistance and safety issues.

Experimental

Synthetic procedures

General

Reagents were of commercial grade and used without further purification. Thin layer chromatography (TLC): aluminum sheets pre-coated with silica gel 60 F254 (Merck). Flash column chromatography was carried out using Silica Gel 60 230–400 mesh (Fluka). Commercially available solvents were used for reactions, TLC and column chromatography. Melting points were determined in a capillary tube on a BÜCHI melting point B-540 apparatus (uncorrected). The NMR spectra were recorded on a Bruker Avance II+ 600 (600.13 for ¹H NMR and 150.92 MHz for ¹³C NMR) or a Bruker Avance NEO 400 MHz (400.13 for ¹H NMR and 100.6 MHz for ¹³C NMR) spectrometer with TMS as the internal standard for chemical shifts. ¹H and ¹³C NMR data are reported as follows: chemical shift, multiplicity (s = singlet, d = doublet, t = triplet, q = quartet, br = broad, m = multiplet), coupling constants (Hz), integration, identification. The assignment of the ¹H and ¹³C NMR spectra was made on the basis of COSY and HSQC experiments. The following abbreviations have been used: Pyr (pyrimidine), Ar (aryl), Trp (tryptophan), His (histidine), Tyr (tyrosine). The high-resolution mass spectra (HRMS) of the compounds were recorded on a Thermo Scientific Q Exactive Plus Hybrid Quadrupole-Orbitrap mass spectrometer. MS acquisition was carried out with heated electrospray ionization (HESI) in positive mode.

(E)-3-(Dimethylamino)-1-ferrocenyl-2-propen-1-one (1)

To a stirred solution of acetylferrocene (0.912 g, 4.0 mmol) in dry DMF (2 mL) under argon atmosphere was added dropwise N,N-dimethylformamide diethyl acetal (2.66 mL, 20.0 mol) and the mixture was heated at 125 °C for 24 h. The reaction was cooled to r.t. and quenched with water. The mixture was extracted with ethyl acetate. The combined organic phases were washed three times with water, dried over MgSO₄, filtered and the solvent was removed under reduced pressure. The residue was purified by flash column chromatography (silica gel, CH₂Cl₂/CH₃OH = 50 : 1) to give 1.030 g (91%) of 1 as an orange solid. ¹H NMR (CDCl₃, 600 MHz): δ 7.70 (d, J = 12.4 Hz, 1H, CHN), 5.35 (d, J = 12.4 Hz, 1H, COCH), 4.77 (s, 2H, Cp–H), 4.38 (s, 2H, Cp–H), 4.15 (s, 5H, Cp5), 2.99 (brs, 6H, (CH₃)₂N) ppm.

1-(o-Tolyl)guanidine hydrochloride (2)

To an ice-cold solution of o-toluidine (1.600 g, 15.0 mmol) in Et₂O (20 mL) was added dropwise HCl 2 N solution in Et₂O (8 mL, 15.8 mmol). The mixture was stirred for 0.5 h at r.t. and then concentrated to dryness under reduced pressure. The obtained hydrochloride was slowly added under stirring to molten cyanamide (1.890 g, 45.0 mmol) at 50 °C. The temperature was then raised to 65–70 °C and the melt was gently stirred for additional 1.5 h. The reaction was then cooled to r.t., Et₂O was added to the oily residue and the mixture was vigorously stirred for 30 min. The crude precipitate was filtered off, washed with Et₂O and dried under vacuum. Yield 96%. ¹H NMR (DMSO-d₆, 400 MHz): δ 9.74 (s, 1H, NH), 7.40 (s, 3H, NH, NH₂), 7.36–7.32 (m, 1H, arom.), 7.30–7.26 (m, 2H, arom.), 7.22–7.18 (m, 1H, arom.), 2.23 (s, 3H, CH₃) ppm.

4-Ferrocenyl-N-(o-tolyl)pyrimidin-2-amine (3)

To a suspension of 1-(o-tolyl)guanidine hydrochloride (0.100 g, 0.546 mmol) and K₂CO₃ (0.151 g, 1.092 mmol) in n-butanol (5 mL) was added enaminone 1 (0.103 g, 0.364 mmol). The mixture was refluxed for 48 h, cooled to r.t. and filtered through a pad of silica gel (EtOAc). The filtrate was concentrated under reduced pressure and subjected to purification by flash column chromatography (silica gel, CH₂Cl₂/EtOAc = 50 : 1) to give 0.076 g (57%) of 3 as orange solid; mp 156–158 °C. ¹H NMR (CDCl₃, 600 MHz): δ 8.25–8.24 (m, 2H, PyrH-6, ArH-6), 7.29 (d, J = 7.3 Hz, 1H, ArH-5), 7.22 (d, J = 7.4 Hz, 1H, ArH-3), 7.02 (dt, J = 7.4 Hz, 1H, ArH-4), 6.85 (brs, 1H, NH), 6.77 (d, J = 5.2 Hz, 1H, PyrH-5), 4.94 (t, J = 1.9 Hz, 2H, Cp–H), 4.38 (t, J = 1.9 Hz, 2H, Cp–H), 4.09 (s, 5H, Cp5), 2.37 (s, 3H, CH₃) ppm. ¹³C NMR (CDCl₃, 150.9 MHz): δ = 168.68 (PyrC-2), 160.27 (ArC-1), 157.07 (PyrCH-6), 137.99 (PyrC-4), 130.43 (ArCH-3), 127.72 (ArC-2), 126.56 (ArCH-5), 122.93 (ArCH-4), 120.90 (ArCH-6), 108.38 (PyrCH-5), 80.80 (C, Cp), 70.92 (2CH, Cp), 70.01 (5CH, Cp5), 68.10 (2CH, Cp), 18.25 (CH₃) ppm. HRMS (HESI): found for C₂₁H₂₀FeN₃ [M + H]⁺m/z 370.1003, calcd. m/z 370.1001.

N-(4-Methyl-3-nitrophenyl)benzamide (6)

To a solution of 4-methyl-3-nitroaniline (1.000 g, 6.57 mmol) and Et₃N (1.660 g, 16.43 mmol) in dichloromethane (10 mL) was added dropwise benzoyl chloride at 0 °C. The reaction was stirred for 2 h at r.t. and quenched with water. The mixture was extracted with dichloromethane. The combined organic phases were washed with water, dried over MgSO₄, filtered and concentrated under reduced pressure. The residue was purified by crystallization from petroleum ether/CH₂Cl₂ (10 : 1) to give 1.204 g (72%) of 6. ¹H NMR (CDCl₃/CD₃OD, 400 MHz): δ 8.28 (d, J = 2.2 Hz, 1H, H-2), 7.97 (dd, J = 8.4, 2.2 Hz, 1H, H-6), 7.92–7.91 (m, 1H, Ph), 7.90–7.89 (m, 1H, Ph), 7.59–7.55 (m, 1H, Ph), 7.51–7.47 (m, 2H, Ph), 7.33 (d, J = 8.4 Hz, 1H, H-5), 2.58 (s, 3H, CH₃) ppm. ¹³C NMR (CDCl₃/CD₃OD, 100.6 MHz): δ = 166.60 (CO), 137.18 (C-1), 137.24 (C-3), 133.14 (CH, arom.), 132.13 (CH, arom.), 129.05 (2C, arom.), 128.71 (CH, arom.), 127.26 (CH, arom.), 124.86 (CH, arom.), 116.14 (CH, arom.), 19.93 (CH₃) ppm.

N-(3-Amino-4-methylphenyl)benzamide (7)

To a solution of nitro compound 6 (0.384 g, 1.5 mmol) in ethanol (23 mL) was added Fe dust (0.335 g, 6.0 mmol). The suspension was heated to reflux and aq. NH₄Cl (0.802 g, 15.0 mmol in 4.5 mL H₂O) was added dropwise at this temperature. After complete addition, the mixture was refluxed for additional 2 h, then cooled to room temperature, filtered through folded filter paper and washed with ethanol. The filtrate was concentrated under reduced pressure. The residue was dissolved in CH₂Cl₂ and washed three times with water, followed by brine. The organic layer was dried over MgSO₄, filtered and concentrated under reduced pressure to give 0.328 g, 97% of the desired product 7. No further purification was required. ¹H NMR (CDCl₃, 400 MHz): δ 7.85–7.83 (m, 2H, arom.), 7.74 (brs, 1H, NH), 7.55–7.51 (m, 1H, arom.), 7.48–7.44 (m, 2H, arom.), 7.30 (d, J = 1.9 Hz, 1H, H-2), 7.00 (d, J = 8.0 Hz, 1H, H-6), 6.74 (dd, J = 8.0, 2.1 Hz, 1H, H-5), 3.31 (brs, 2H, NH₂), 2.14 (s, 3H, CH₃) ppm. ¹³C NMR (CDCl₃, 100.6 MHz): δ = 165.56 (CO), 145.03 (C-3), 136.89 (C-1), 135.24 (C, arom.), 131.67 (CH, arom.), 130.67 (CH, arom.), 128.75 (2CH, arom.), 126.97 (2CH, arom.), 118.76 (C, arom.), 110.19 (CH, arom.), 106.91 (CH, arom.), 16.92 (CH₃) ppm.

N-(3-Guanidino-4-methylphenyl)benzamide hydrochloride (8)

To an ice-cooled solution of 7 (0.320 g, 1.410 mmol) in Et₂O/CH₂Cl₂ (5/5 mL) was added dropwise HCl 2 N solution in Et₂O (0.8 mL, 1.490 mmol). The mixture was stirred for 0.5 h at r.t. and then concentrated under reduced pressure. The obtained hydrochloride was slowly added under stirring to molten cyanamide (0.178 g, 4.230 mmol) at 50 °C. The temperature was then raised at 65–70 °C and the melt was gently stirred for additional 1.5 h. The reaction was then cooled to r.t., Et₂O and a few drops of MeOH were added to the oily residue and the mixture was vigorously stirred for 30 min. The crude precipitate was filtered off, washed with Et₂O and dried under vacuum. Yield: 96%. ¹H NMR (DMSO-d₆, 400 MHz): δ 10.40 (s, 1H, NH), 9.72 (s, 1H, NH), 7.98–7.96 (m, 2H, arom.), 7.78 (d, J = 1.9 Hz, 1H, H-6), 7.73 (dd, J = 8.3, 2.1 Hz, 1H, H-4), 7.62–7.59 (m, 1H, arom.), 7.56–7.52 (m, 2H, arom.), 7.40 (brs, 3H, NH, NH₂), 7.31 (d, J = 8.3 Hz, 1H, H-3), 2.20 (s, 3H, CH₃) ppm. ¹³C NMR (DMSO-d₆, 100.6 MHz): δ = 166.00 (CO), 156.87 (CNH), 138.65 (C-1), 135.18 (C-5), 133.58 (C, arom.), 132.15 (CH, arom.), 131.55 (CH, arom.), 130.24 (C, arom.), 128.90 (2CH, arom.), 128.11 (2CH, arom.), 119.66 (CH, arom.), 119.14 (CH, arom.), 17.34 (CH₃) ppm.

N-(4-Methyl-3-((4-ferrocenylpyrimidin-2-yl)amino)phenyl)-benzamide (9)

To a suspension of 8 (0.214 g, 0.636 mmol) and K₂CO₃ (0.220 g, 1.590 mmol) in n-butanol (10 mL) was added enaminone 1 (0.150 g, 0.530 mmol). The mixture was refluxed for 48 h, cooled to r.t. and filtered through a pad of silica gel (EtOAc). The filtrate was concentrated under reduced pressure and subjected to purification by flash column chromatography (silica gel, CH₂Cl₂/EtOAc = 10 : 1) to give 0.101 g (39%) of 9 as an orange solid; mp 216–218 °C. ¹H NMR (CDCl₃/CD₃OD, 400 MHz): δ 8.23 (d, J = 2.2 Hz, 1H, ArH-2), 8.12 (d, J = 5.3 Hz, 1H, PyrH-6), 7.86–7.84 (m, 2H, Ph), 7.48–7.39 (m, 3H, Ph), 7.33–7.31 (m, 1H, ArH-6), 7.14 (d, J = 8.2 Hz, 1H, ArH-5), 6.74 (d, J = 5.3 Hz, 1H, PyrH-5), 4.89 (t, J = 1.9 Hz, 2H, Cp–H), 4.41 (t, J = 1.9 Hz, 2H, Cp–H), 4.01 (s, 5H, Cp5), 2.27 (s, 3H, CH₃) ppm. ¹³C NMR (CDCl₃/CD₃OD, 100.6 MHz): δ = 169.46 (PyrC-2), 166.83 (CO), 160.00 (ArC-3), 156.44 (PyrCH-6), 137.77 (PyrC-4), 136.33 (ArC-1), 134.95 (ArC-4), 131.65 (CH, Ph), 130.59 (ArCH-5), 128.52 (2CH, Ph), 127.23 (2CH, Ph), 125.46 (C, Ph), 116.48 (ArCH-6), 114.77 (ArCH-2), 108.47 (PyrCH-5), 80.09 (C, Cp), 71.15 (2CH, Cp), 69.99 (5CH, Cp5), 68.07 (2CH, Cp), 17.44 (CH₃) ppm. HRMS (HESI): found for C₂₈H₂₅FeN₄O [M + H]⁺m/z 489.1375, calcd. m/z 489.1372.

Methyl 4-methyl-3-nitrobenzoate (10)

A solution of methyl 4-methylbenzoate (4.5 g, 30 mmol) in conc. H₂SO₄ (3 mL) was cooled to 0 °C and a mixture of 65% HNO₃ (3 mL) and conc. H₂SO₄ (3 mL) was added dropwise over a period of 1 h. The reaction was stirred at r.t. for 2 h, then poured on crushed ice, and extracted with EtOAc. The organic phases were separated, washed with water, aq. NaHCO₃, and brine, dried over MgSO₄, filtered and concentrated under reduced pressure. Recrystallization from ethanol provided 4.286 g (73%) of the desired product 10. ¹H NMR (CDCl₃, 400 MHz): δ 8.62 (d, J = 1.7 Hz, 1H, H-2), 8.15 (dd, J = 8.0, 1.7 Hz, 1H, H-6), 7.45 (d, J = 8.0 Hz, 1H, H-5), 3.96 (s, 3H, OCH₃), 2.67 (s, 3H, CH₃) ppm.

Methyl 3-amino-4-methylbenzoate (11)

To a solution of nitro compound 10 (2.00 g, 10.25 mmol) in ethanol (20 mL) was added Fe dust (2.29 g, 41.00 mmol). The suspension was heated to reflux and aq. NH₄Cl (5.48 g, 102.5 mmol in 7 mL H₂O) was added dropwise at this temperature. After complete addition the mixture was refluxed for additional 3 h, then cooled to room temperature, filtered through folded filter paper and washed with ethanol. The filtrate was concentrated under reduced pressure. The residue was dissolved in EtOAc and washed three times with water, followed by brine. The organic layer was dried over MgSO₄, filtered and concentrated under reduced pressure to give 1.470 g, 87% of the desired product 11. No further purification was required. ¹H NMR (CDCl₃, 400 MHz): δ 7.38 (dd, J = 7.7, 1.6 Hz, 1H, H-6), 7.36 (d, J = 1.6 Hz, 1H, H-2), 7.10 (d, J = 7.7 Hz, 1H, H-5), 3.88 (s, 3H, OCH₃), 2.21 (s, 3H, CH₃) ppm.

Methyl 3-guanidino-4-methylbenzoate hydrochloride (12)

To an ice-cold solution of amine 11 (1.34 g, 8.11 mmol) in Et₂O (40 mL) was added dropwise HCl 2 N solution in Et₂O (4.5 mL, 9.00 mmol). The mixture was stirred for 0.5 h at r.t. and then concentrated under reduced pressure. The obtained hydrochloride was slowly added under stirring to molten cyanamide (1.62 g, 8.03 mmol) at 50 °C. The temperature was then raised at 65–70 °C and the melt was gently stirred for additional 2 h. The reaction was then cooled to r.t., Et₂O was added to the oily residue and the mixture was vigorously stirred for 30 min. The crude precipitate was filtered off, washed with Et₂O and dried under vacuum. Yield: 99%. ¹H NMR (CD₃OD, 600 MHz): δ 7.95 (dd, J = 8.0, 1.6 Hz, 1H, H-4), 7.87 (d, J = 1.6 Hz, 1H, H-6), 7.49 (d, J = 8.0 Hz, 1H, H-3), 4.88 (s, 4H, 2NH, NH₂), 3.90 (s, 3H, OCH₃), 2.37 (s, 3H, CH₃) ppm. ¹³C NMR (CD₃OD, 150.9 MHz): δ = 166.09 (CN), 157.00 (CO), 141.76 (C-1), 133.19 (C-5), 131.53 (CH, arom.), 129.50 (C-2), 129.34 (CH, arom.), 128.70 (CH, arom.), 51.41 (OCH₃), 16.52 (CH₃) ppm.

Butyl 4-methyl-3-(4-ferrocenylpyrimidin-2-ylamino)benzoate (13)

To a suspension of 12 (0.130 g, 0.530 mmol) and K₂CO₃ (0.150 g, 1.060 mmol) in n-butanol (5 mL) was added enaminone 1 (0.100 g, 0.353 mmol). The mixture was refluxed for 48 h, cooled to r.t. and filtered through a pad of silica gel (EtOAc). The filtrate was concentrated under reduced pressure and subjected to purification by flash column chromatography (silica gel, petroleum ether/EtOAc = 2 : 1) to give 0.078 g (47%) of 13 as an orange solid. ¹H NMR (CDCl₃, 400 MHz): δ 9.09 (d, J = 1.7 Hz, 1H, ArH-2), 8.26 (d, J = 5.3 Hz, 1H, PyrH-6), 7.70 (dd, J = 7.8, 1.7 Hz, 1H, ArH-6), 7.27 (d, J = 8.0 Hz, 1H, ArH-5), 7.00 (brs, 1H, NH), 6.81 (d, J = 5.3 Hz, 1H, PyrH-5), 5.03 (t, J = 1.9 Hz, 2H, Cp–H), 4.49 (t, J = 1.9 Hz, 2H, Cp–H), 4.38 (t, J = 6.8 Hz, 2H, OCH₂), 4.10 (s, 5H, Cp5), 2.42 (s, 3H, CH₃), 1.84–1.77 (m, 2H, OCH₂CH₂), 1.53–1.47 (m, 2H, OCH₂CH₂CH₂), 0.99 (s, 3H, CH₂CH₃) ppm. ¹³C NMR (CDCl₃, 100.6 MHz): δ 168.98 (PyrC-2), 166.94 (CO), 159.92 (ArC-3), 157.05 (PyrCH-6), 138.12 (PyrC-4), 132.34 (ArC-1), 130.31 (ArCH-5), 129.08 (ArC-4), 123.87 (ArCH-6), 121.47 (ArCH-2), 108.58 (PyrCH-5), 80.44 (C, Cp), 71.10 (2CH, Cp), 70.73 (5CH, Cp5), 68.20 (2CH, Cp), 64.76 (OCH₂), 30.95 (OCH₂CH₂), 19.31 (OCH₂CH₂CH₂), 18.42 (CH₃), 13.87 (CH₂CH₃) ppm.

4-Methyl-3-(4-ferrocenylpyrimidin-2-ylamino)benzoic acid (14)

To a solution of 13 (0.158 g, 0.337 mmol) in ethanol (10 mL) was added 5% NaOH (2.5 mL, 3.03 mmol) and the mixture was stirred for 2.5 h at 50 °C. The reaction mixture was concentrated under reduced pressure, cooled to 0 °C and neutralized with 1 N HCl. The mixture was extracted three times with EtOAc. The combined organic layers were washed with water, followed by brine, dried over MgSO₄, filtered and concentrated under reduced pressure to give 0.122 g (88%) of 14. ¹H NMR (DMSO-d₆, 400 MHz): δ 12.27 (brs, 1H, OH), 8.63 (s, 1H, ArH-2), 8.48 (s, 1H, PyrH-6), 7.61 (d, J = 7.7 Hz, 1H, ArH-6), 7.33 (d, J = 7.9 Hz, 1H, ArH-5), 7.00 (d, J = 5.1 Hz, 1H, PyrH-5), 5.01 (s, 2H, Cp–H), 4.49 (t, 2H, Cp–H), 4.11 (s, 5H, Cp5), 2.35 (s, 3H, CH₃) ppm. ¹³C NMR (DMSO-d₆, 100.6 MHz): δ 168.06 (PyrC-2), 160.75 (CO), 157.88 (PyrCH-6, ArC-3), 138.90 (PyrC-4), 136.08 (ArC-1), 130.71 (ArCH-5, (ArC-4), 124.82 (ArCH-6), 124.58 (ArCH-2), 108.55 (PyrCH-5), 81.07 (C, Cp), 71.17 (2CH, Cp), 70.22 (5CH, Cp5), 68.28 (2CH, Cp), 18.79 (CH₃) ppm.

General procedure for the synthesis of the target compounds 15a–15e

To a solution of acid 14 (1 equiv.) in CH₃CN was added N-[3-(dimethylamino) propyl]-N-ethylcarbodiimide (1 equiv.), 1-hydroxybenzotriazole (1 equiv.) and the appropriate amine (1.2 equiv.). The mixture was stirred at r.t. and the product formation was monitored by TLC (7d for 15a; 24 h for 15b–e). The reaction mixture was concentrated under reduced pressure and subjected to purification by flash column chromatography (silica gel, CH₂Cl₂/EtOAc for 15a; CH₂Cl₂/MeOH for 15b–e). Specific details are given for each compound below.

4-Methyl-N-phenyl-3-((4-ferrocenylpyrimidin-2-yl)amino)benzamide (15a)

Yield: 94%; orange solid; mp 117–119 °C. ¹H NMR (CDCl₃, 400 MHz): δ 9.02 (d, J = 1.8 Hz, 1H, ArH-2), 8.28 (d, J = 5.3 Hz, 1H, PyrH-6), 8.05 (brs, 1H, NH), 7.71–7.69 (m, 2H, Ph), 7.54 (dd, J = 7.8, 1.9 Hz, 1H, ArH-6), 7.40–7.36 (m, 2H, Ph), 7.32 (d, J = 7.8 Hz, 1H, ArH-5), 7.18–7.13 (m, 1H, Ph), 6.98 (brs, 1H, CONH), 6.83 (d, J = 5.3 Hz, 1H, PyrH-5), 4.97 (t, J = 1.9 Hz, 2H, Cp–H), 4.45 (t, J = 1.9 Hz, 2H, Cp–H), 4.07 (s, 5H, Cp5), 2.42 (s, 3H, CH₃) ppm. ¹³C NMR (CDCl₃, 100.6 MHz): δ 168.97 (PyrC-2), 165.98 (CO), 159.86 (ArC-3), 157.07 (PyrCH-6), 138.44 (PyrC-4), 138.24 (C, Ph), 133.60 (ArC-1), 130.83 (ArCH-5), 130.45 (ArC-4), 129.05 (2CH, Ph), 124.34 (CH, Ph), 121.32 (ArCH-6), 120.17 (2CH, Ph), 117.91 (ArCH-2), 108.92 (PyrCH-5), 80.45 (C, Cp), 71.22 (2CH, Cp), 70.10 (5CH, Cp5), 68.07 (2CH, Cp), 18.23 (CH₃) ppm. HRMS (HESI): found for C₂₈H₂₅FeN₄O [M + H]⁺m/z 489.1374, calcd. m/z 489.1372.

Methyl (4-methyl-3-((4-ferrocenylpyrimidin-2-yl)amino)benzoyl)-l-phenylalaninate (15b)

Yield: 97%; orange solid; mp 66–68 °C. ¹H NMR (CDCl₃, 400 MHz): δ 8.77 (d, J = 1.6 Hz, 1H, ArH-2), 8.25 (d, J = 5.2 Hz, 1H, PyrH-6), 7.35 (dd, J = 7.7, 1.8 Hz, 1H, ArH-6), 7.28–7.22 (m, 4H, Ph), 7.17–7.15 (m, 2H, ArH-5, Ph), 6.90 (brs, 1H, NH), 6.82 (d, J = 5.2 Hz, 1H, PyrH-5), 6.65 (brs, 1H, CONH), 5.12 (dq, J = 7.5, 5.8 Hz, 1H, NHCHCO₂CH₃), 4.97 (t, J = 1.9 Hz, 2H, Cp–H), 4.47 (t, J = 1.9 Hz, 2H, Cp–H), 4.10 (s, 5H, Cp5), 3.74 (s, 3H, OCH₃), 3.31–3.21 (m, 2H, CH₂Ph), 2.40 (s, 3H, CH₃) ppm. ¹³C NMR (CDCl₃, 100.6 MHz): δ 172.11 (COOCH₃), 169.00 (PyrC-2), 167.00 (CONH), 159.97 (ArC-3), 156.98 (PyrCH-6), 138.36 (PyrC-4), 136.01 (C, Ph), 133.52 (ArC-1), 131.24 (ArC-4), 130.52 (ArCH-5), 129.37 (2CH, Ph), 128.60 (2CH, Ph), 127.12 (CH, Ph), 121.06 (ArCH-6), 119.26 (ArCH-2), 108.84 (PyrCH-5), 80.46 (C, Cp), 71.12 (2CH, Cp), 70.08 (5CH, Cp5), 68.15 (2CH, Cp), 53.63 (NHCHCO₂CH₃), 52.34 (OCH₃), 38.09 (CH₂Ph), 18.27 (CH₃) ppm. HRMS (HESI): found for C₃₂H₃₁FeN₄O₃ [M + H]⁺m/z 575.1747, calcd. m/z 575.1740.

Methyl (4-methyl-3-((4-ferrocenylpyrimidin-2-yl)amino)benzoyl)-l-tryptophanate (15c)

Yield: 74%; orange solid; mp 100–102 °C. ¹H NMR (CDCl₃, 400 MHz): δ 8.72 (s, 1H, ArH-2), 8.22–8.20 (m, 1H, PyrH-6), 8.10 (brs, 1H, TrpNH), 7.56 (d, J = 7.9 Hz, 1H, TrpH), 7.33–7.28 (m, 2H, ArH-6, TrpH), 7.20 (d, J = 7.8 Hz, 1H, ArH-5), 7.15–7.12 (m, 1H, TrpH), 7.06–7.03 (m, 1H, TrpH), 6.99 (s, 1H, TrpH), 6.88 (brs, 1H, NH), 6.80 (dd, J = 5.2, 0.8 Hz, 1H, PyrH-5), 6.75 (brs, 1H, CONH), 5.19–5.14 (m, 1H, NHCHCO₂CH₃), 4.96–4.94 (m, 2H, Cp–H), 4.43–4.42 (m, 2H, Cp–H), 4.08 (s, 5H, Cp5), 3.69 (s, 3H, OCH₃), 3.45–3.43 (m, 2H, CH₂), 2.37 (s, 3H, CH₃) ppm. ¹³C NMR (CDCl₃, 100.6 MHz): δ 172.48 (COOCH₃), 168.96 (PyrC-2), 167.11 (CONH), 159.96 (ArC-3), 156.97 (PyrCH-6), 138.26 (PyrC-4), 136.10 (TrpC), 133.00 (TrpC), 132.51 (ArC-1), 131.37 (ArC-4), 130.47 (ArCH-5), 127.63 (TrpC), 122.90 (TrpCH), 122.18 (TrpCH), 121.25 (TrpCH), 119.65 (TrpCH), 119.43 (ArCH-2), 118.70 (ArCH-6), 111.27 (TrpCH), 108.80 (PyrCH-5), 80.47 (C, Cp), 71.14 (2CH, Cp), 70.09 (5CH, Cp5), 68.14 (2CH, Cp), 53.42 (NHCHCO₂CH₃), 52.38(OCH₃), 27.80 (CH₂), 18.25 (CH₃) ppm. HRMS (HESI): found for C₃₄H₃₂FeN₅O₃ [M + H]⁺m/z 614.1855, calcd. m/z 614.1849.

Methyl (4-methyl-3-((4-ferrocenylpyrimidin-2-yl)amino)benzoyl)-l-histidinate (15d)

Yield: 70%; orange solid; mp 112–114 °C. Methyl (4-methyl-3-((4-ferrocenylpyrimidin-2-yl)amino)benzoyl)-l-histidinate × 2HCl. ¹H NMR (DMSO-d₆, 353 K, 400 MHz): δ 9.96 (brs, 1H, NH), 8.92 (s, 1H, HisH), 8.82 (d, J = 7.7 Hz, 1H, HisNH), 8.34 (d, J = 5.8 Hz, 1H, PyrH-6), 8.24 (d, J = 1.4 Hz, 1H, HisH), 7.69 (dd, J = 7.9, 1.4 Hz, 1H, ArH-6), 7.39–7.37 (m, 2H, ArH-2, ArH-5), 7.12 (d, J = 5.8 Hz, 1H, PyrH-5), 5.08 (s, 2H, Cp–H), 4.89–4.84 (m, 1H, NHCHCO₂CH₃), 4.68 (s, 2H, Cp–H), 4.18 (s, 5H, Cp5), 3.69 (s, 3H, OCH₃), 3.33–3.30 (m, 2H, CH₂Ph), 2.38 (s, 3H, CH₃) ppm. ¹³C NMR (DMSO-d₆, 353 K, 100.6 MHz): δ 171.47 (COOCH₃), 168.82 (CONH), 166.74 (PyrC-2), 159.26 (ArC-3), 151.19 (PyrCH-6), 136.82 (PyrC-4), 136.37 (HisC), 134.07 (HisCH), 132.42 (ArC-1), 130.79 (ArCH-5), 130.11 (ArC-4), 124.60 (ArCH-6), 124.51 (HisCH), 117.41 (ArCH-2), 108.41 (PyrCH-5), 78.85 (C, Cp), 73.07 (2CH, Cp), 70.88 (5CH, Cp5), 69.44 (2CH, Cp), 55.27 (NHCHCO₂CH₃), 52.61 (OCH₃), 26.41 (CH₂), 18.47 (CH₃) ppm. HRMS (HESI): found for C₂₉H₂₉FeN₆O₃ [M + H]⁺m/z 565.1649, calcd. m/z 565.1645.

Methyl (4-methyl-3-((4-ferrocenylpyrimidin-2-yl)amino)benzoyl)-l-tyrosinate (15e)

Yield: 70%; orange solid; mp 94–96 °C. ¹H NMR (CDCl₃, 400 MHz): δ 8.71 (d, J = 1.7 Hz, 1H, ArH-2), 8.22 (d, J = 5.3 Hz, 1H, PyrH-6), 7.31 (dd, J = 7.8, 1.7 Hz, 1H, ArH-6), 7.18 (d, J = 7.8 Hz, 1H, ArH-5), 7.00–6.96 (m, 3H, NH, 2H TyrH), 6.80 (d, J = 5.3 Hz, 1H, PyrH-5), 6.72–6.70 (m, 3H, CONH, 2H TyrH), 5.05 (dq, J = 7.7, 5.8 Hz, 1H, NHCHCO₂CH₃), 4.96–4.95 (m, 2H, Cp–H), 4.46–4.45 (m, 2H, Cp–H), 4.08 (s, 5H, Cp5), 3.72 (s, 3H, OCH₃), 3.20 (dd, J = 14.0, 5.8 Hz, 1H, CH₂Ph), 3.12 (dd, J = 14.0, 5.8 Hz, 1H, CH₂Ph), 2.32 (s, 3H, CH₃) ppm. ¹³C NMR (CDCl₃, 100.6 MHz): δ 172.28 (COOCH₃), 169.34 (PyrC-2), 167.33 (CONH), 159.78 (ArC-3), 156.69 (PyrCH-6), 155.43 (C, Tyr), 138.18 (PyrC-4), 132.36 (ArC-1), 131.63 (ArC-4), 130.58 (ArCH-5), 130.44 (2CH, Tyr), 127.30 (C, Tyr), 121.24 (ArCH-6), 119.45 (ArCH-2), 115.69 (2CH, Tyr), 108.84 (PyrCH-5), 80.90 (C, Cp), 71.27 (2CH, Cp), 70.12 (5CH, Cp5), 68.21 (2CH, Cp), 53.82 (NHCHCO₂CH₃), 52.39 (OCH₃), 37.25 (CH₂), 18.17 (CH₃) ppm. HRMS (HESI): found for C₃₂H₃₁FeN₄O₄ [M + H]⁺m/z 591.1695, calcd. m/z 591.1689.

Cell lines and culture conditions

In vitro cytotoxicity was assessed in a panel of human bcr–abl+ malignant cell lines of leukemic origin (AR-230, BV-173, LAMA-84, K-562) and against normal fibroblast cells (CCL-1). All cell lines were purchased from the German Collection of Microorganisms and Cell Cultures (DSMZ GmbH, Braunschweig, Germany). All cell cultures were cultivated in growth medium RPMI 1640 supplemented with 10% fetal bovine serum (FBS), 5% l-glutamine and incubated under standard conditions of 310 K and 5% humidified CO₂ atmosphere.

Cell viability assay

Experimental design involved a number of cytotoxicity assays that measured cell growth inhibition by the newly synthesized ferrocene derivatives and imatinib as a reference drug. Cell viability was evaluated using a standard MTT-based colorimetric assay. Exponential-phase cells were harvested and seeded (100 μl per well) in 96-well plates at the appropriate density (3 × 10⁵) for the suspension cultures and 1.5 × 10⁵ for the adherent one (CCL-1). Cells were treated and incubated with various concentrations of the experimental compounds in the concentration range of 200–0.5 μM. After exposure time of 72 h, filter sterilized MTT substrate solution (5 mg ml⁻¹ in PBS) was added to each well of the culture plate. A further 1–4 h incubation allowed for the formation of purple insoluble formazan precipitates. The latter were dissolved in isopropyl alcohol solution containing 5% formic acid prior to absorbance measurement at 550 nm. Collected absorbance values were blanked against MTT and isopropanol solution and normalized to the mean value of untreated control (100% cell viability). Semi-logarithmic “dose–response” curves were constructed and the half-inhibitory concentrations of the screened compounds against each tested cell line were calculated. Values of p ≤ 0.05 were considered statistically significant.

Author contributions

The study concept was designed by GS and GM. GS and IP performed the synthesis and analysis; GM, RM, and RA studied the antiproliferative activity. The manuscript was reviewed by all the authors.

Conflicts of interest

There are no conflicts to declare.