Synthesis, biological activity and modeling study of some thiopyrimidine derivatives and their platinum(II) and ruthenium(III) metal complexes

1 Introduction

The pyrimidine molecule is a promising structural moiety for drug design because it forms a component in a number of useful drugs and is associated with many biological and therapeutic activities [1]. Besides the biological significance of pyrimidine, several of its derivatives have been developed as chemotherapeutic agents and have found wide clinical applications such as antiviral [2, 3], antiprotozoal [4], antimicrobial, anti-insecticidal [5–8], anticancer [9–14], antihypertensive [15], antihypoglycemic and hypolipidemic [16], and anti-HIV agents [17–19], in addition to their cardiovascular [20] and diuretic [21, 22] properties. Furthermore, some pyrimidines are used as hypnotic and sedative drugs [23, 24]. Two diarylpyrimidines (DAPY), rilpivirine (1) [25] and etravirine [26, 27], have been classified as non-nucleoside reverse transcriptase inhibitors, whereas bacimethrin (4-amino-5-(hydroxymethyl)-2-methoxypyrimidine) (2) (Fig. 1) is a pyrimidine antibiotic that is active against several staphylococcal bacteria [28].

Fig. 1:

Chemical structures of rilpivirine (1), bacimethrin (2) and [M(pyrimidinyl-pyrazole)Cl₂] [M=Pt, Ru(Cl)] complexes (3).

Some platinum and ruthenium complexes play a crucial role in preclinical as well as in clinical studies. Saha and Muherjee [29] have reported the metal complexes of palladium(II) and platinum(II) 3 of (3,5-dimethyl-2′-pyrimidyl)pyrazole as potent anticancer agents (Fig. 1). Furthermore, we have reported the synthesis and anti-HIV activity of a new series of Pt(II), as well as those of other metal complexes with 3-methyl-6,7-diphenyllumazine [30]. In addition, Joshi et al. [31] have discussed the selectivity of anticancer drugs toward cancer cells as one of the main goals of drug optimization, by using the complexes Ru(dppz)₂(CppH)](PF₆)₂ [where CppH = 2-(2-pyridyl)pyrimidine-4-carboxylic acid and dppz = dipyrido[3,2-a:2′,3′-c]phenazine)] as a “photocaging” agent to give a light illumination at 350 nm for detection of cancer cells. This is the first substitutionally inert cytotoxic metal complex to be used as a light-triggered prodrug candidate.

In continuation of our ongoing work on the synthesis of a new anti-HIV agent and our recent antiviral data on new pyrimidine derivatives [32–34], we report here the synthesis of a new series of pyrimidines having arylazo residues, including some of their Pt(II) and Ru(III) metal complexes, and the evaluation of their anti-HIV and antibacterial activities, together with a molecular modeling study.

2 Results and discussion

6-Amino-5-(aryldiazenyl)-N¹,N³-dimethyl-2-thioxo-pyrimidin-4-one derivatives 17–27 were prepared in 40–92 % yield via diazotization reaction by treatment of 6-amino-N¹,N³-dimethyl-2-thioxo-pyrimidin-4-one (4) with various substituted amines – 4-chloro- (6), 4-bromo- (7), 4-nitro- (8), 4-methyl- (9), 4-sulfonamido- (10) and 4-sulfonylactamido- (11) anilines – as well as with p-aminobenzoic acid (12), p-aminosulfonic acid (13), sulfadiazine (14), sulfamerazine (15) and sulfamethazine (16), in the presence of NaNO₂ and HCl at 0–5 °C, then at ambient temperature for 15 h. Under similar conditions, azothiopyrimidine 28 was prepared in 68 % yield from diazotization of 6-amino-N¹-methyl-2-thioxo-pyrimidin-4-one (5) (Scheme 1).

Scheme 1:

Synthesis of the 6-amino-5-aryl-diazenyl-N¹,N³-dimethyl-2-thioxo-pyrimidin-4-one derivatives (17–27) and of the N¹-methyl analog (28).

The structures of 17–28 were assigned on the basis of their ¹H, ¹³C NMR and mass spectra. The ¹H NMR spectra showed similar patterns for the aromatic and pyrimidine protons and carbon atoms, where the NH₂ group of 17, 18, 20 and 23–25 resonated as two separated singlets at the regions δ = 9.14–11.75 ppm and δ = 8.64–9.02 ppm, and this pattern was due to the hydrogen bonding between the NH₂ group and the nitrogen atom of the azo residue. Compound 20 was selected for ¹H NMR measurement at high temperature (364 K), which revealed the appearance of a broad singlet at δ = 9.99 ppm assigned to the NH₂ group, which was exchangeable with D₂O, and the disappearance of NH₂ protons at δ = 11.64 and 8.65 ppm (92 K). The aromatic, pyrimidine and methyl protons were fully analyzed (see Experimental Section). In the ¹³C NMR spectra of 17–28, the resonances at the regions δ = 177.2–179.1 and 157.0–151.0 ppm were assigned to C-2 (C=S) and C-4 of the thiopyrimidine ring, whereas the resonances at the regions δ = 75.0–82.7 and 157.1–165.5 ppm were attributed to C-5 and C-6 of the same ring, respectively. The other carbon atoms of the pyrimidine and aromatic substituents were fully assigned (see Experimental Section), whereas all the synthesized compounds were confirmed from their ¹H and ¹³C HSQC NMR spectra [35]. Compound 27 was selected for further NMR experiments. The gradient HMBC [36] NMR spectrum of 27 revealed two ²J_C,H couplings: the aromatic carbon atom C-1′ at δ = 143.5 ppm showed a ²J_C,H coupling with 2′-H and 6′-H of the same ring at δ = 8.08 ppm, whereas the other one was observed between 5″-H of the pyrimidine ring at δ = 6.67 ppm and two methyl carbon atoms attached to C-4′ and C-6″ of the same ring at δ = 23.0 ppm. Furthermore, C-2 (C=S) of the pyrimidine ring at δ = 177.5 ppm showed two ³J_C,H couplings, one with N¹-Me protons at δ = 3.69 ppm and the other with N³-Me protons at δ = 3.89 ppm. In addition, a ³J_C,H coupling was observed between C-5″ of the pyrimidine scaffold at δ = 113.1 ppm and the two methyl carbon atoms attached to C-4″ and C-6″ of the same ring at δ = 23.0 ppm (Fig. 2).

Fig. 2:

Chemical shifts (δ, ppm) and J_C,H correlations in the HMBC NMR spectrum of 27.

Our procedure was modified by preparing some new platinum(II) and ruthenium(III) complexes. Thus, compound 5 was selected as a precursor for the synthesis of the new platinum(II) complex 29 in 48 % yield from the treatment of 5 with PtCl₂ in hot MeOH, aiming to examine its antiviral activity (Scheme 2). The structure of 29 was determined from the elemental analysis, ¹H, ¹³C NMR and IR spectra (see Experimental Section). In the ¹³C NMR spectrum of 5, carbon atoms 2 and 4–6 appeared at δ = 179.7, 164.4, 160.1 and 79.6 ppm, respectively. In the IR spectrum, υ(C=O) and υ(C=S) of the free ligand 5 were observed at 1630 and 1099 cm^–1, respectively, which shifted to 1649 and 1134 cm^–1 for complex 29. This indicated that the >C=O and >C=S groups of the ligand are involved in the bonding with metal ions. The υ(NH₂) observed at 3100 and 3400 cm^–1 in the spectrum of free ligand remained constant in the spectrum of the complex. This observation is in agreement with the data reported by Masouda et al. [37] and Parthasarathi et al. [38] concerning the metal complexes of some thiouracil derivatives. The spectrum of 29 also showed a low wavenumber band at 448 cm^–1, characteristic for a υ(Pt–O) coordinated bond, as expected [39], whereas a band at 706 cm^–1 was probably due to a pyrimidine-ring vibration [40]. These data suggested that the geometrical structure of 29 is a distorted square planar. The chelation via N and S donor atoms and the formation of a four-membered ring at 29 are in agreement with the X-ray data of the related complex Pt(η²-N,S-pymS)(η¹-S-pymS)(PPh₃)] reported by Lobana et al. [41].

Scheme 2:

Synthesis of the bis(4-amino-N³-methyl-6-oxo-2-thioxo-pyrimidin-1-yl)platinum(II) complex [Pt(ThPyrm)₂] (29).

The platinum complexes of some synthesized azothiopyrimidine analogs had been explored, aiming to study their stereochemical geometry in addition to evaluating their anti-HIV activity. Thus treatment of the azo ligands 17 and 21–23 with Pt(II)Cl₂ in refluxing MeOH gave, after purification, the desired Pt(II) complexes [(AzPyrm-Cl)PtCl₂] 30–33 in 55 %, 46 %, 45 % and 58 % yield, respectively (Scheme 3).

Scheme 3:

Synthesis of the bis(4-amino-N¹,N³-dimethyl-6-oxo-2-thioxo-pyrimidin-1-yl)Pt(II)Cl₂ complexes [(ADiThPyrm)₂PtCl₂] (29–33).

The structures of 30–33 were assigned from the elemental analysis, ¹H, ¹³C NMR and IR spectra, since they showed similar patterns of aromatic and pyrimidine protons. In the ¹³C NMR spectra, the lower field resonances at δ = 178.2–177.4 ppm were assigned to C-2, whereas the resonances at δ = 154.8–153.2 ppm were attributed to C-4 of the pyrimidine ring, respectively. The resonances at δ = 84.1–77.8 and 158.5–156.7 ppm were assigned to C-5 and C-6 of the same ring, respectively. The aromatic carbon atoms were fully analyzed (see Experimental Section). The IR spectra of the free ligands 17 and 21–23 showed the characteristic υ(C=O) bands in the 1626, 1629, 1630 and 1641 cm^–1 region, respectively, which shifted to higher wavenumbers in the spectra of the metal complexes 30–33 (1656, 1675, 1653 and 1657 cm^–1, respectively) [41–43]. In addition, the bands at 1414, 1416, 1406 and 1414 cm^–1 were due to the azo group υ(N=N), respectively, which shifted to higher wavenumbers of the metal complexes in the 1420, 1421, 1425 and 1427 cm^–1 range, respectively, indicating the coordination of C=O with one nitrogen atom of the azo group with platinum(II). Furthermore, the bands of the free ligands of υ(C=S) vibrated at 1091, 1100, 1101 and 1122 cm^–1, which almost vibrated at the same wavenumbers for the complexes at 1088, 1104, 1100 and 1120 cm^–1, respectively. The wavenumbers at ∼800 cm^–1 are characteristic for the pyrimidine ring stretching. The band at 554 cm^–1 of complex 32 probably belongs to the Pt–N coordinated bond. In conclusion, the above spectral data suggested that the complexes 30–33 exist in square-planar geometries.

Owing to the biological importance of ruthenium complexes [44, 45], we focused our work on the synthesis of a new derivative of ruthenium(III) pyrimidine complex 34 in 43 % yield by treatment of 5 with RuCl₃·3H₂O in a 2:1 ratio (Scheme 4). The structure of 34 was confirmed by the elemental analysis, ¹H, ¹³C NMR and IR spectra. In the ¹³C NMR spectrum of 34, the low field frequency at δ = 182.8 ppm was assigned to C-2 (C=S), whereas C-4–C-6 appeared at δ = 156.6, 82.3 and 162.5 ppm, respectively (see Experimental Section). In the IR spectrum, the C=O at C-6 showed two bands: one band appeared at wavenumber 1646 cm^–1, which was higher than that of the starting material 5 (1630 cm^–1), indicating a coordination of the C=O group of 34 with Ru(III), whereas the other band vibrated at 1600 cm^–1 as a non-coordinated C=O group. Similarly, the C=S group showed two bands as well: one band vibrated at 1124 cm^–1, which is a higher wavenumber than that for 5 (1100 cm^–1), indicating the coordination of the C=S group at C-2 with Ru, whereas the second band at 1100 cm^–1 was attributed to the non-coordinated C=S group. The above data suggested that 34 exists as a distorted compressed octahedron.

Scheme 4:

Synthesis of the bis(4-amino-N¹-methyl-6-oxo-2-thioxo-pyrimidine)RuCl·H₂O complex [(AMThPyrm)₂RuCl·H₂O] (34).

Next, the treatment of 17 with RuCl₃·3H₂O (1:1 molar ratio) in refluxing MeOH afforded, after purification, the Ru(III) complex 35 (54 %) (Scheme 5). The ¹³C NMR spectrum showed two lower field resonances at δ = 177.8 and 163.2 ppm assigned to C-2 and C-6 of the pyrimidine ring, respectively, whereas the resonances at δ = 158.2 and 80.4 ppm were attributed to C-4 and C-5 of the same ring, respectively. The aromatic and methyl carbon atoms were fully assigned (see Experimental Section). The IR spectrum of ligand 17 revealed bands at 1626 and 1414 cm^–1 assigned to υ(C=O) and υ(N=N), respectively, which shifted to higher wavenumbers (1649 and 1134 cm^–1) in the spectrum of complex 35. These differences in wavenumbers indicate that the >C=O and >N=N groups of 17 are involved in the bonding with metal ions. The υ(C=S) vibration of 17 was observed at 1091 cm^–1, which remained constant in the spectrum of the complex at 1088 cm^–1. The above spectral data suggest the geometrical structure of Ru(III) complex 35 as a distorted octahedral one.

Scheme 5:

Synthesis of the [6-amino-5-((4-chlrophenyl)diazenyl)-N¹,N³-dimethyl-2-thioxo-pyrimidin-4-one)Ru(III)Cl₃·H₂O complex [(AzPyrm-Cl)RuCl₃·H₂O] (34).

2.2 Molecular modeling analysis

The azothiopyrimidine analog 26, owing to its anti-HIV activity, was selected for the molecular modeling study, which was performed to understand the binding mode of these analogs with the HIV-RT binding pocket (NIBP) (PDB code: 1DTT [49]). The Sybyl-X 1.1 software was used, and the results were visualized with Pymol [50].

The molecular modeling study of compound 26 showed a binding energy score of –8.0, which revealed a selective interaction with the amino acids of the RT HIV enzyme. As shown in Figure 3, the aromatic ring of the sulfonamide group of 26 fitted into an aromatic-rich subpocket surrounded by the aromatic side chains of Tyr179, and a hydrogen bond between the hydroxyl group of Tyr179 and the nitrogen atom of the azo group was observed. The sulfonamide moiety was located in a favorable position of the pocket of RT enzyme, leading to the formation of a hydrogen bond between the oxygen atom of the SO₂ group and the NH₂ of Lys101, in addition to the hydrogen bond between the NH₂ of the sulfonamide group and the oxygen atom of the carboxylic group of Thr132 of the RT enzyme. Overall, the combination of hydrophobic interaction and π stacking appears to govern the binding of 130 with HIV RT.

Fig. 3:

Docked conformation of 26 showing three hydrogen bonds: Lys101 with the oxygen atom of the SO₂ group, Thr132 with the NH group of the sulfonamide residue and OH of Tyr179 with the nitrogen atom of the azo group. In addition, a hydrophobic interaction was observed between the phenyl group of the benzenesulfonamido moiety and the aromatic ring of Tyr179 of the reverse transcriptase (RT) enzyme residues.

3 Experimental section