Synthesis and structure evaluation of new complex butylarylpiperazin-1-yl derivatives

A series of arylpiperazine derivatives of 1,16-diphenyl-19-azahexacyclo-[14.5.1.02,15.03,8.09,14.017,21]docosa-2,3,5,7,8,9,11,13,14-nonaene-18,20,22-trione and 4,10-diphenyl-1H,2H,3H,5H-indeno[1,2-f]isoindole-1,3,5-trione was synthesized. The pharmacological profile of compound 4 at the 5-HT1A receptor was measured by binding assay. The title compounds were tested in cell-based assay against the human immunodeficiency virus type-1. The X-ray crystallographic studies of derivatives 2, 6, 7, 11, 19, and 20 were presented.

Using quantitative structure-activity relationship analysis, the ''rule of five'' scheme was worked out for orally administrated drugs (Lipinski et al., 1997;Kerns and Di, 2008). According to authors, the drugs that cross the blood-brain barrier are those of molecular mass lower than 450 u and of theoretical partition coefficient n-octanol/water (logP) being in the range of 1-4 or logD 7.4 1-3. The biological barrier permeability is also determined by the following important parameters: numbers of hydrogen bond donors and acceptors in the potential medicine's structure (HBD maximum 4 and HBA less than 6), polar surface area (PSA) correlated with them [expected value is less than 60-70 Å 2 (Oprea, 2002)], as well as compound's solubility (logS greater than 60 lg/cm 3 ). Proper drug permeability makes it possible to cross the barrier and to reach the regions of a drug's action.
In last two decades, a number of binding modes of longchain arylpiperazine derivatives to 5-HT 1A (Lewgowd et al., 2011;Nowak et al., 2006), 5-HT 2A (Klabunde and Evers, 2005;Bronowska et al., 2001), and 5-HT 7 (Kim et al., 2012;López-Rodríguez et al., 2003) receptors have been proposed. The ionic interaction between the protonated nitrogen of the piperazine ring of a ligand and Asp3.32 residue of the receptor (Nowak et al., 2006;Vermeulen et al., 2003;Roth et al., 1997) constituted a main essential interaction. The hydrophobic terminal imide or amide group, the hydrocarbon linker, and an aromatic ring bound to the piperazine moiety are placed in a hydrophobic pocket composed of aromatic and/or aliphatic amino acids side chains (Kim et al., 2012;Varin et al., 2010;Lepailleur et al., 2005). The flexible chain of N-(4-arylpiperazin-1-yl-alkyl)substituted derivatives can adopt one of the two main conformations: extended or bent. The results of geometry optimization (Lewgowd et al., 2011) proved that conformers with extended spacer are preferred in a solution, whereas in vacuum bent geometries predominate. Theoretical calculations determine minimum energy for extended linker conformations also in solid state and for complexes with a receptor (Siracusa et al., 2008). According to pharmacophore model of the 5-HT 1A receptor (Chilmonczyk et al., 1997;Bronowska et al., 2001), a folded conformer promotes high affinity for the 5-HT 1A receptor. It is known that ligand binding can lead to a change in the conformation of the receptor protein, however, also in the ligand itself . In addition, the role of the solvent molecules is quite difficult to explain-they can take part in a ligand-receptor H-bond formation, be involved in the process of a receptor activation or influence entropy effects (Pardo et al., 2007). This paper reports synthesis and biological activity of compounds purposely designed to combine the bulky hydrophobic imide ring with alkyl linker bearing different substituents. The collected group of arylpiperazine derivatives can be used for further investigations concerning ligand-5-HT receptor interactions. For this reason X-ray crystallographic studies of derivatives 2, 6, 7, 11, 19, and 20 were described. The molecular descriptors for selected arylpiperazine derivatives were presented. The pharmacological profile of the compound 4 was evaluated for its affinity to the 5-HT 1A receptor. It was reported, that cytotoxicity of aromatic, highvolume arylpiperazine derivatives is low (Filosa et al., 2007;Ananda Kumar et al., 2009), and they act as anti-HIV-1 agents (Yang et al., 2010), cytotoxicity and anti-HIV activity of selected derivatives were examined.

Chemistry
All chemicals and solvents were purchased from Aldrich. Melting points were determined on an Electrothermal Digital Melting Point Apparatus and are uncorrected. The NMR spectra were recorded on a Bruker AVANCE DMX400 spectrometer, operating at 300 MHz ( 1 H NMR) and 75 MHz ( 13 C NMR). The chemical shift values are expressed in ppm relative to TMS as an internal standard. Mass spectral electrospray ionization (ESI) measurements were carried out on a Mariner Perspective-Biosystem instrument with TOF detector. The spectra were obtained in the positive ion mode with a declustering potential 140-300 V. Elemental analyses were recorded on a CHN model 2400 Perkin-Elmer. TLC was carried out using silica gel 60 F 254 with layer thickness of 0.25 mm (Merck) and the results were visualized using UV lamp at 254 nm. Column chromatography was carried out using silica gel 60 (200-400 mesh, Merck) and chloroform/ methanol (19.5:0.5 vol) mixture as eluent.

Biological assays
Cell-based assays Cell-based assays were performed at Dipartimento di Scienze e Tecnologie Biomediche, Università di Cagliari, Monserrato, Italy.

Test compounds
Compounds were dissolved in DMSO at 100 mM and then diluted in culture medium.

Cells and viruses
Cell line and viruses were purchased from the American Type Culture Collection (ATCC). The absence of mycoplasma contamination was checked periodically by the Hoechst staining method. Cell line supporting the multiplication of human immunodeficiency virus type-1 (HIV-1) was the CD4? human T-cells containing an integrated HTLV-1 genome (MT-4).

X-ray structure determination
Suitable crystals were mounted for measurements. The X-ray measurements were performed at 100(2) K on a KUMA CCD k-axis diffractometer with graphite-monochromated Mo Ka radiation (0.71073 Å ). The crystals were positioned at 62.25 mm from the KM4CCD camera. The data were corrected for Lorentz and polarization effects, additionally absorption corrections were applied. Data reduction and analysis were carried out with the Kuma Diffraction (Wrocław, Poland) programmes (Oxford Diffraction CrysAlis CCD and CrysAlis RED, 2001). The structures were solved by direct methods (Sheldrick, 1990) and refined by using SHELXL (Sheldrick, 1997) The refinement was based on F 2 for all reflections except for those with very negative F 2 . The weighted R factor, wR, and all goodness-of-fit S values are based on F 2 . The nonhydrogen atoms were refined anisotropically. The hydrogen atoms were located from a difference map and were refined isotropically. The atomic scattering factors were taken from the International Tables (Wilson, 1992

Results and discussion
Chemistry
All obtained compounds were purified by flash chromatography. Elemental analysis, mass spectrometry, 1 H NMR and 13 C NMR spectra confirmed the identity of the products. For compounds 2 and 11, also for hydrochlorides of 6, 7, 19, and 20 X-ray analyses were done.

Biology
Cytotoxicity and anti HIV-1 activity Title compounds were tested in cell-based assay against the human immunodeficiency virus type-1 (HIV-1), using Efavirenz as reference inhibitor. The cytotoxicity was evaluated in parallel with the antiviral activity.
None of tested compounds showed selective antiviral activity against HIV-1. However compounds 10 and 14 turned out cytotoxic for exponentially growing MT4 cells in the low micromolar range (CC 50 = 9 lM) ( Table 1).

X-ray structural analyses
The crystal structures have been determined for three ''phencyclone'' derivatives 2, 6, and 7. Their main skeleton resembles buspirone, but have more bulky maleimide fragment and in the case of 2 there is no piperazine moiety (n-butyl chain is terminated by bromine atom). In structures 6 and 7, the aromatic fragment (p-chlorophenyl and o-fluorophenyl, respectively) is different from 2-pirymidinyl substituent in buspirone.
In all of these structures phenanthrene moiety forms a kind of ''roof'' over n-butyl chain, and phenyl rings are situated like ''wings'' directed outside (Fig. 2). In structures 6 and 7, the piperazine moiety adopts chair conformation. All compounds crystallize in monoclinic system without solvent with one molecule in an asymmetric unit. Unit cell contains 4 molecules related by inversion center (Fig. 3).
The crystal structure of 2 is stabilized by two kinds of short interactions between C-HÁÁÁO and C-HÁÁÁBr (Fig. 4). In 6 there are three types of C-HÁÁÁO contacts. The oxygen atom from maleimide moiety contacts with piperazine and phenanthrene fragments. Second one interacts with phenyl ring (Fig. 5). The structure of 7 shows similar C-HÁÁÁO interactions and there is an additional short C-HÁÁÁF contact (Fig. 6).
Two crystal structures based on ''Indanocyclone'' 11 and 19 are disordered. Compound 11 crystallizes without solvent in monoclinic P2 1 space group with two molecules in an asymmetric unit. The structure is a racemic twin in which one molecule is disordered. The disorder occurs in the n-butyl chain together with bromine atom and in the first phenyl ring of Indanocyclone. Two side phenyl rings are almost coplanar, the angle between mean best planes is 3.5°. There are three types of C-HÁÁÁO interactions between maleimide oxygens and the nbutyl chain, as well as the side phenyl ring, and between oxygen from Indanocyclone moiety and the side phenyl ring (Fig. 7).
Compound 19 crystallizes as a hydrochloride with one molecule of water in triclinic P-1 space group with one molecule in an asymmetric unit. Disorder occurs in first Indanocyclone phenyl ring and gives rise to pÁÁÁp stacking between disordered benzene and maleimide rings. Two side phenyl rings are tilted with respect to each other by 24.8° (Fig. 8). The n-butyl chain adopts cis conformation with dihedral angle N1-C28-C29-C30 equal to 55.6. The structure is stabilized by a set of N ? HÁÁÁClbonds between piperazine and chloride anions. There are two types of interactions between oxygens from maleimide moiety and C-H from butyl chain and Indanocyclone phenyl ring. Water molecule forms C-HÁÁÁO bonds with piperazine and Indanocyclone phenyl ring. There are also O-HÁÁÁClinteractions (Fig. 9).
Compound 20, an analog of NAN-190, crystallizes in triclinic P-1 space group as a hydrochloride with one molecule in an asymmetric unit. The imide moiety is almost planar. The piperazine ring adopts chair conformation (Fig. 10). The crystal structure forms layers  along a axis comprising of alternating molecules (Fig. 11). The structure is stabilized by N ? HÁÁÁClhydrogen bonds. In addition there are short contacts between chloride anion and C-H from the methoxy group, the butyl chain and the piperazine moiety. There are also interactions between oxygens from the imide fragment with C-H from piperazine and the methoxyphenyl ring (Fig. 12).

Conclusions
Compounds 6, 7, 19, and 20 fit well to the three-point pharmacophore model for 5-HT 1A receptor ligands (Chilmonczyk et al., 1997). Apart from an aromatic ring and the basic nitrogen of piperazine, localized in the distance of 5.2-5.7 Å from a centroid, authors have found the third point essential for a ligand-receptor interaction-the carbonyl oxygen, expected in the distance of 7.07 Å from the center of an aromatic ring and 4.3 Å from N4 piperazine atom. Intramolecular distances measured for a set of 5-HT 1A receptor ligands by Chilmonczyk et al. were in the range of 7.93-12.37 Å (CentroidÁÁÁO(1)), 3.95-7.16 Å (N(1)ÁÁÁO(1)), and 5.15-5.64 Å (CentroidÁÁÁN(1)). The values calculated for new arylpiperazine derivatives (6, 7, 19, and 20) are in agreement with the presented threepoint pharmacophore model (Table 2, Fig. 13). The distance between the center of the phenyl group and the imide oxygen (O1) is in the range of 8,13-11,89 Å . The measured distance of the protonated nitrogen (N1) and O1 atom is in the range of 4.06-6.66 Å . The value of centroid -N1 length is in a narrow range between 5.67 and 5.71 Å . Presented results suggest that compounds 6, 7, 19, and 20 could serve as potential 5-HT 1A receptor ligands. They also prove that similar molecular values can be estimated for the derivative 4. Although it is an exception from ''the rule of five,'' because of its high molecular weight, volume and logP, and low solubility logS (Table 3), the compound 4 possess moderate activity to the 5-HT 1A receptor.  Structural data obtained for a set of long-chain arylpiperazine derivatives can serve for further investigations concerning ligands activity to metabotropic 5-HT receptors.