2,4-Diacyl(dialkoxycarbonyl)-3-R-5-hydroxy-5-methylcyclohexanones 1 (also referred to as cyclic β-hydroxy ketones of β-cycloketols) are promising building blocks for fine organic synthesis [1, 2]. Compounds 1 readily undergo heterocyclizations with various 1,2- and 1,3-binucleophiles to produce isoquinolines 2 [3–7], indazoles 3 [1, 8–12], 2,1-benzoxazoles 4 [1, 9–11], [1,2,4]triazolo[3,4-b]quinazolines 5 [13], pyrazolo[3,4-c]isoquinolines 6 [14], 4,5,6,7,8,9-hexahydropyrazolo[1,5-a]quinazolines 7 [15], and 6,7,8,8a-tetrahydropyrazolo[5,1-b]quinazolin-9(5H)-one derivatives 8 [16] (Scheme 1). Among the heterocyclic systems presented in Scheme 1, we focused on indazole derivatives 3. The synthesis of compounds like 3 was described for the first time in the early 20th century [1, 17, 18]. Since that time, numerous data have been accumulated on the synthesis of analogous indazoles from various β-cycloketols and substituted hydrazines [8–12, 19–22]; however, their transformations and functionalizations have been explored to a much lesser extent. It should also be noted that indazole fragment is the key structural unit of many therapeutically important compounds (for reviews, see [23–26]).

Scheme
scheme 1

1.

Full size image

In continuation of our studies in the field of β-cycloketols and their analogs [15, 27–30], we turned our attention to the regioselectivity of tosylation of indazoles 3. A priori, the tosylation of 3 could give rise to N1-, N2-, and/or 6-O-tosyl derivatives (Scheme 2). Regioisomeric N1- and N2-substituted products could be formed assuming the possibility of prototropic tautomerism of 3.

Scheme
scheme 2

2.

Full size image

In fact, indazoles 3a–3e reacted with tosyl chloride in anhydrous acetone in the presence of triethylamine as a base to afford exclusively N1-tosyl derivatives 4a–4e (Scheme 3). No O-tosylation products were detected; presumably, this reaction path is unfavorable for steric reasons. The structure of 4a–4e was confirmed by NMR spectra and X-ray analysis of 4a. It should be noted that both initial hydroxy ketone 1 [1] and its hydrazination products exist as mixtures of diastereoisomers with preferential syn orientation of the HO and CH3C(O) groups. According to the X-ray diffraction data, indazole 3a [31] and its hydrochloride [32] are racemates that crystallize in centrosymmetric space groups; at least two stereochemical configurations, (4S,5R,6S) or (4R,6R) were reported for these compounds. The presence of diastereoisomers of 4 is responsible for doubling of some signals in their NMR spectra.

Scheme
scheme 3

3.

Full size image

Figure 1 shows the structure of molecule 4a determined by X-ray analysis. It was identified as 1-[6-hydroxy-3,6-dimethyl-1-(4-methylbenzenesulfonyl)-4-phenyl-4,5,6,7-tetrahydro-1H-indazol-5-yl]ethanone. The principal crystallographic data are given in Table 1, and Table 2 contains the bond lengths and bond angles in molecule 4a. The cyclohexene ring has a half-chair conformation with the C5 and C4 atoms significantly deviating from the C6C7C2 plane [by 0.521(5) and 0.270(5) Å, respectively] and the C3 atom lying almost in that plane (the deviation of C3 from the C6C7C2 plane does not exceed 0.1 Å). The bond angles in the cyclohexene ring are as follows: C2C7C6 125.0(3), C7C6C5 109.5(3), C6C5C4 108.6(3), C5C4C3 113.2(3), C4C3C2 110.3(3), C3C2C7 123.7(3)°. Due to the presence of different substituents, the cyclohexene ring is characterized by the following torsion angles: C7C6C5C4 51.0(4), C6C5C4C3 –63.0(4), C5C4C3C2 38.7(4), C4C3C2C7 –6.0(5), C3C2C7C6 –2.7(6), C2C7C6C5 –21.0(5)°. The substituents on C3, C4, and C5 occupy equatorial positions but are oriented oppositely with respect to each other. The hydroxy group on C5 is axial. Similar orientations of substituents were reported previously [33, 34]. The pyrazole ring in molecule 4a is virtually planar, and deviations of atoms from the C7N1C1N2C2 plane ar within 0.001–0.011 Å. The bond lengths and bond angles in the pyrazole ring conform to the corresponding reference values: N2–C1 1.321(4), C1–C2 1.424(4), C2–C7 1.364(4), C7–N1 1.388(4), N1–N2 1.386(4) Å; 111.3(3)–105.1(3)° [35]. The tosyl substituent on N1 is axial, and the methyl group on C1 is equatorial; the S1 atom of the tosyl group deviates from the pyrazole ring plane by 0.502(5) Å upward, whereas deviation of the methyl carbon atom (C8) from that plane is insignificant [0.043(5) Å]. Molecules 4a in crystal are linked through intermolecular hydrogen bonds (Table 3, Fig. 2) to form dimers with a graph set descriptor of R22(8) [36].

Fig. 1.
figure 1

Structure of the molecule of 1-[6-hydroxy-3,6-dimethyl-1-(4-methylbenzenesulfonyl)-4-phenyl-4,5,6,7-tetrahydro-1H-indazol-5-yl]ethanone (4a) in crystal according to the X-ray diffraction data.

Full size image
Fig. 2.
figure 2

A fragment of crystal packing of compound 4a.

Full size image
Table 1. Some crystallographic parameters for compound 4a
Full size table
Table 2. Some bond angles and bond lengths in molecule 4a
Full size table
Table 3. Hydrogen bonds in the crystal structure of compound 4a
Full size table

In order to rationalize the observed regioselectivity of the tosylation of indazoles 3, we performed a theoretical study of their reactivity and relative thermodynamic stability of possible products. The calculations were carried out in the framework of density functional theory (DFT) using a widely known B3LYP hybrid functional [37, 38] and split-valence 6-31G(d,p) basis set implemented in GAMESS (US) software package. The optimized structures were visualized by ChemCraft. The ground-state energies were calculated after preliminary search for most stable conformations, followed by geometry optimization. Non-specific solvation was taken into account according to the CPCM model with acetone as solvent [39].

The completely regioselective tosylation of 3a at the N1 atom (Figs. 3, 4) is primarily determined by the kinetic factor, i.e., by a significant difference in the partial negative charges on the N1 and N2 atoms (Table 4). This difference arises from the involvement of the N2 lone electron pair (LEP) in the pyrazole aromatic system; in contrast, the LEP on N1 appears in the pyrazole ring plane and is not involved in conjugation. Presumably, the thermodynamic factor is not crucial here, since the energy difference between the isomeric N1- and N2-substituted compounds is as small as 4.8 kJ/mol. Conformational analysis of product 3a revealed two most stable conformers 3a-1 and 3a-2 (Fig. 4) with an energy difference of 0.06 kJ/mol in favor of the former, where the benzene rings are located at the same side of the tetrahydroindazole plane.

Table 4. Calculated [B3LYP/6-31G(d,p)] partial negative charges on the nitrogen atoms in molecule 3a
Full size table
Fig. 3.
figure 3

Optimized structure of 1-[6-hydroxy-3,6-dimethyl-4-phenyl-4,5,6,7-tetrahydro-1H-indazol-5-yl]ethanone (3a).

Full size image
Fig. 4.
figure 4

Most stable conformers of compound 4a with (a) syn and (b) anti orientation of the benzene rings.

Full size image

Compounds 4a–4e were analyzed in silico to predict their drug-likeness, ADMET (Absorption, Distribution, Metabolism, Excretion, Toxicity) parameters, biological activity, and possible targets with the aid OSIRIS Property Explorer [40], SwissADME [41], SwissTargetPrediction [42], Molinspiration Property Calculation Service [43], and AntiBac-Pred services [44]. OSIRIS Property Explorer was used to evaluate cLog P (lipophilicity), log S (solubility), TPSA (Topological Polar Surface Area), risks of side effects (such as mutagenic, oncogenic, and reproductive), drug-likeness, and drug score [40]. The structures were analyzed for the conformity to “Lipinskiʼs rule of five” (cLog P ≤ 5.0, MW ≤ 500, TPSA ≤ 140, number of H-bond acceptors ≤ 10, number of H-bond donors ≤ 5) [45–47]. The results are presented in Table 5.

Table 5. Toxicity risks and physicochemical characteristics of compounds 4a–4e predicted by OSIRIS Property Explorer
Full size table

It is seen that the cLogP values of 4a–4e do not exceed 2.9, which suggests a good absorption and acceptable lipophilicity [45–47]. All compounds (except for 4b and 4e) are characterized by an acceptable solubility (log S). The molecular weights of 4a–4e do not exceed 500, which meets Lipinskiʼs rule of five. On the other hand, compound 4e was the only one that showed a positive drug-likeness value. All compounds 4a–4e exhibited a moderately high drug score (~0.3). Possible toxicological risks for the reproductive systems were predicted for all these compounds. Molinspiration Property Calculation Service predicted kinase inhibitory activity of 4a–4e as the most probable biological activity (Molinspiration bioactivity score –0.42 to –0.26).

The AntiBac-Pred computations suggest probable resistance of Staphylococcus simulans and Mycobacterium ulcerans toward the examined compounds [the confidence parameters (C) calculated as the difference between probable activity (PA) and probable inactivity (PI) range from 0.28 to 0.42 and from 0.18 to 0.36, respectively]. Antibacterial activities against Staphylococcus lugdunensis (C = 0.24–0.33) and Staphylococcus sciuri (C = 0.29–0.32 for 4a–4c) were predicted.

According to the SwissADME prediction, all compounds 4a–4e are characterized by high gastrointestinal absorption and the lack of BBB (blood–brain barrier) permeability, as well as by possible inhibitory activity against proteins of the cytochrome P450 family (CYP) (Table 6). As the SwissTargetPrediction data showed, the most probable targets of all compounds (except for 4d for which no appropriate targets were found) are a number of enzymes and class A G-protein-conjugated receptors. The bioavailability index for all compounds is equal to 0.55, in keeping with Lipinskiʼs rule of five [48].

Table 6. ADMET parameters and biological activity of compounds 4a–4e predicted by SwissADME and SwissTargetPrediction
Full size table

In summary, the tosylation of indazoles obtained from cyclic β-hydroxy ketones and hydrazine gives the corresponding N1-tosyl derivatives with complete regioselectivity. Theoretical study of the reactivity of the initial indazoles and thermodynamic stability of the tosylation products confirmed the predominant reaction direction and revealed the determining effect of the kinetic factor. In silico evaluation of biological activity of the synthesized N-tosylindazoles showed that these compounds meet the bioavailability criterion and are promising candidates for further in vitro and in vivo screening.

EXPERIMENTAL

The 1H and 13C NMR spectra were recorded on a Bruker AC-300 spectrometer (300 and 75 MHz, respectively) in CDCl3 or DMSO-d6 using the residual proton and carbon signals of the solvent as reference. The IR spectra were measured in KBr on a Varian 3600 FT-IR Excalibur Series spectrometer. The elemental analyses were obtained using a Carlo Erba 1106 CHN analyzer. The melting points were measured on a Koefler hot stage and are uncorrected. The purity of the isolated compounds was checked by TLC on Silufol UV-254 plates using acetone–hexane (1 : 1) as eluent; spots were visualized by treatment with iodine vapor or under UV light.

1-[4-Aryl-6-hydroxy-3,6-dimethyl-1-(4-methylbenzenesulfonyl)-4,5,6,7-tetrahydro-1H-indazol-5-yl)ethanones 4a–4e (general procedure). A mixture of tetrahydroindazole 3a–3e (50 mmol), 9.53 g (50 mmol) of p-toluenesulfonyl chloride, and 10 mL of triethylamine in 300 mL of anhydrous acetone was refluxed for 6–8 h (TLC). The mixture was cooled and diluted with 300 mL of cold distilled water, the resulting suspension was kept for 24 h, and the precipitate was filtered off and recrystallized from ethanol. Compounds 4a–4e were isolated as white finely crystalline powders.

1-[6-Hydroxy-3,6-dimethyl-1-(4-methylbenzenesulfonyl)-4-phenyl-4,5,6,7-tetrahydro-1H-indazol-5-yl]ethanone (4a). Yield 65%, mp 178°C, stereoisomer ratio ~3 : 1. 1H NMR spectrum (CDCl3), δ, ppm: major stereoisomer: 1.39 s (3H, CH3), 1.50 s (3H, CH3), 1.68 s (3H, CH3), 2.41 s (3H, CH3), 2.91–3.05 m (2H, 5-H, 7-H, overlapped), 3.29 d (1H, 7-H, 2J = 17.9 Hz), 3.70 br.s (1H, OH), 4.02 d (1H, 4-H, 3J = 10.6 Hz), 7.07 d (2H, Harom, 3J = 7.7 Hz), 7.28–7.33 m (5H, Harom), 7.86 d (2H, MeC6H4, 3J = 8.1 Hz); minor stereoisomer: 1.32 s (3H, CH3), 1.66 s (3H, CH3), 2.71 d (1H, 7-H, 2J = 16.7 Hz), 3.55 br.s (1H, OH), 4.09 d (1H, 4-H, 3J = 11.5 Hz), 7.80 d (2H, MeC6H4, 3J = 8.3 Hz). 13C NMR spectrum (CDCl3), δC, ppm: major stereoisomer: 13.3, 21.6, 28.3, 34.7, 37.4, 41.7, 62.7, 71.2, 119.4, 127.5, 128.0, 129.1, 129.9, 135.0, 140.1, 140.8, 145.2, 152.4, 162.2, 216.3; minor stereoisomer, 11.9, 28.2, 37.0, 41.6, 63.1, 71.3, 118.9, 127.4, 129.8, 139.7, 140.9, 145.6, 152.6, 162.1, 216.2. Found, %: C 65.70; H 6.06; N 6.37. C24H26N2O4S. Calculated, %: C 65.73; H 5.98; N 6.39.

1-[6-Hydroxy-3,6-dimethyl-1-(4-methylbenzenesulfonyl)-4-(4-methylphenyl)-4,5,6,7-tetrahydro-1H-indazol-5-yl]ethanone (4b). Yield 62%, mp 173°C. 1H NMR spectrum (DMSO-d6), δ, ppm: 1.28 s (3H, CH3), 1.34 s (3H, CH3), 2.02 s (3H, CH3), 2.22 s (3H, CH3), 2.37 s (3H, CH3), 2.81 d (1H, 5-H, 3J = 10.7 Hz), 3.15 br.s (2H, 7-H), 4.17 d (1H, 4-H, 3J = 10.7 Hz), 4.94 br.s (1H, OH), 6.99 d (2H, 4-C6H4, 3J = 7.5 Hz), 7.05 d (2H, 4-C6H4, 3J = 7.5 Hz), 7.42 d (2H, m-H, Ts, 3J = 7.8 Hz), 7.79 d (2H, o-H, Ts, 3J = 7.8 Hz). 13C NMR spectrum (DMSO-d6), δC, ppm: 12.9, 20.6, 21.1, 28.0, 30.7, 38.7, 38.9, 65.0, 69.4, 120.7, 127.3, 128.4, 129.2, 130.2, 134.6, 136.0, 137.9, 141.7, 145.4, 152.3, 210.1. Found, %: C 66.34; H 6.36; N 6.15. C25H28N2O4S. Calculated, %: C 66.35; H 6.24; N 6.19.

1-[6-Hydroxy-3,6-dimethyl-4-(4-methoxyphenyl)-1-(4-methylbenzenesulfonyl)-4,5,6,7-tetrahydro-1H-indazol-5-yl]ethanone (4c). Yield 59%, mp 184°C, stereoisomer ratio ~4 : 1. 1H NMR spectrum (CDCl3), δ, ppm: major stereoisomer: 1.37 s (3H, CH3), 1.52 s (3H, CH3), 1.71 s (3H, CH3), 2.40 s (3H, CH3), 2.89–3.02 m (2H, 5-H, 7-H, overlapped), 3.26 d (1H, 7-H, 2J = 18.1 Hz), 3.69 br.s (1H, OH), 3.78 s (3H, OCH3), 3.98 d (1H, 4-H, 3J = 10.7 Hz), 6.82 d (2H, m-H, 4-MeOC6H4, 3J = 8.3 Hz), 6.99 d (2H, o-H, 4-MeOC6H4, 3J = 8.3 Hz), 7.31 d (2H, m-H, Ts, 3J = 8.0 Hz), 7.85 d (2H, o-H, Ts, 3J = 8.0 Hz); minor stereoisomer, 1.31 s (3H, CH3), 1.69 s (3H, CH3), 1.86 s (3H, CH3), 2.68 d (1H, 7-H, 2J = 16.2 Hz), 3.54 br.s (1H, OH), 4.04 d (1H, 4-H, 3J = 11.1 Hz), 7.79 d (2H, o-H, Ts, 3J = 8.7 Hz). 13C NMR spectrum (CDCl3), δC, ppm: major stereoisomer: 13.4, 21.6, 28.3, 34.7, 37.4, 40.9, 55.1, 62.7, 71.2, 114.4, 119.7, 127.5, 129.9, 131.9, 135.0, 140.6, 145.2, 152.5, 158.8, 162.2, 216.5; minor stereoisomer: 11.8, 21.6, 28.4, 37.0, 40.7, 63.2, 71.2, 114.3, 119.1, 127.6, 129.0, 129.8, 132.6, 139.7, 142.3, 152.0, 158.6, 162.3, 216.4. Found, %: C 64.04; H 6.06; N 6.00. C25H28N2O5S. Calculated, %: C 64.08; H 6.02; N 5.98.

1-[4-(Furan-2-yl)-6-hydroxy-3,6-dimethyl-1-(4-methylbenzenesulfonyl)-4,5,6,7-tetrahydro-1H-indazol-5-yl]ethanone (4d). Yield 57%, mp 136°C. 1H NMR spectrum (CDCl3), δ, ppm: 1.39 s (3H, CH3), 1.66 s (3H, CH3), 1.86 s (3H, CH3), 2.40 s (3H, CH3), 2.97 d (1H, 7-H, 2J = 17.9 Hz), 3.16–3.28 m (2H, 5-H, 7-H), 3.64 br.s (1H, OH), 4.20 d (1H, 4-H, 3J = 10.7 Hz), 6.15–6.16 m (1H, 3-H, Fu), 6.31–6.32 m (1H, 4-H, Fu), 7.31 d (2H, m-H, Ts, 3J = 7.9 Hz), 7.36–7.37 m (1H, 5-H, Fu), 7.85 d (2H, o-H, Ts, 3J = 7.8 Hz). 13C NMR spectrum (CDCl3), δC, ppm: 12.3, 21.6, 28.3, 33.7, 34.9, 37.3, 58.9, 71.0, 108.3, 110.5, 117.3, 127.6, 129.9, 135.0, 140.6, 142.3, 145.2, 152.0, 162.3, 215.9. Found, %: C 61.60; H 5.76; N 6.52. C22H24N2O5S. Calculated, %: C 61.67; H 5.65; N 6.54.

1-[4-(4-Chlorophenyl)-6-hydroxy-3,6-dimethyl-1-(4-methylbenzenesulfonyl)-4,5,6,7-tetrahydro-1H-indazol-5-yl]ethanone (4e). Yield 65%, mp 181°C, stereoisomer ratio ~3 : 1. 1H NMR spectrum (DMSO-d6), δ, ppm: major stereoisomer: 1.27 s (3H, CH3), 1.36 s (3H, CH3), 2.04 s (3H, CH3), 2.39 s (3H, CH3), 2.82 d (1H, 5-H, 3J = 11.2 Hz), 3.14 br.s (2H, 7-H), 4.24 d (1H, 4-H, 3J = 11.2 Hz), 5.00 br.s (1H, OH), 7.16 d (2H, ClC6H4, 3J = 8.3 Hz), 7.32 d (2H, ClC6H4, 3J = 8.3 Hz), 7.43 d (2H, m-H, Ts, 3J = 8.3 Hz), 7.78 d (2H, o-H, Ts, 3J = 8.3 Hz); minor stereoisomer: 1.18 s (3H, CH3), 1.82 s (3H, CH3), 2.00 s (3H, CH3), 2.61 d (1H, 7-H, 2J = 16.6 Hz), 2.77 br.s (2H, 7-H), 4.28 d (1H, 4-H, 3J = 11.0 Hz), 4.85 br.s (1H, OH), 7.76 d (2H, o-H, Ts, 3J = 8.3 Hz). Found, %: C 60.88; H 5.45; N 6.00. C24H25ClN2O4S. Calculated, %: C 60.94; H 5.33; N 5.92.

X-Ray analysis of compound 4a. The X-ray diffraction data for compound 4a were obtained at 296(2) K on a Bruker APEX-II automated three-circle diffractometer (MoKα radiation, λ 0.71073 Å, graphite monochromator, CCD detector, ω-scanning, 2θ = 49.42°). The structure was solved by the direct method using SHELXL-2014 [49] and WINGX [50] and was refined against F2 by the full-matrix least-squares method in anisotropic approximation for non-hydrogen atoms. Hydrogen atoms were placed in geometrically calculated positions (or their positions were determined by difference electron density synthesis) which were refined according to the riding model (Uiso = nUeq, n = 1.5 for methyl groups, n = 1.2 for other hydrogens). The molecular structures were plotted usng Platon [51] and Ortep-3 [52]. The coordinates of atoms and other crystallographic parameters of compound 4a were deposited to the Cambridge Crystallographic Data Centre (CCDC entry no. 1 874 578).