# INTRODUCTION ore than ten years have passed since the research paper concerning the Rule of Five (RO5) by Lipinski et al. (2001) was published 1 . Assuming the evaluation of potentially orally active compounds, in the discovery setting, the RO5 predicted 1,2 that poor absorption or permeation was more likely when there were more than five hydrogen-bond donors (OH plus NH count), more than ten hydrogen -bond acceptors (O plus N atoms), the molecular weight (MW) was greater than 500 and the calculated log P value for the system octan-1-ol/water using CLOGP approach 3 was more than 5 or higher than 4.15 when applied Moriguchi MLOGP predictive method 4 , respectively. In other words, these physicochemical parameters were connected with acceptable aqueous solubility and intestinal permeability and comprised the first steps of oral bioavailability. The RO5 was M deliberately created to be a conservative predictor in an area where medicinal chemistry produced too many compounds with poor physicochemical properties 2 . Veber et al. 5 later suggested that the commonly applied MW cutoff at 500 did not itself significantly separate compounds with poor oral bioavailability from those with acceptable values. Taking into account the fact that molecular rigidity was a much more complex issue than the simple counting of rotatable bonds, their conclusions 5 also pointed out that the compounds which met only two criteria, i.e. ten or fewer rotatable bonds and the value of polar surface area (PSA) equal to or less than 140 Å 2 (or 12 or fewer hydrogen-bond donors and acceptors), would have a high probability of good oral bioavailability. Nevertheless, it is important to emphasize the limitations of these rules: (i) the RO5 applies only to the compounds which are delivered by the oral route, (ii) the RO5 applies only to the compounds which are absorbed by passive mechanisms, (iii) there are important exceptions (MW>500 and reduced molecular flexibility and constrained PSA; natural products), (iv) passing the RO5 is no guarantee that a compound is drug-like, (v) the RO5 says nothing about specific chemistry structural features found in drugs or non-drugs 2,6,7 . From chemical viewpoint, currently in silico investigated molecules, labelled as 8aB-8iB, could be regarded as ortho-/meta-/para-alkoxyphenylcarbamic acid-based derivatives as well as N-arylpiperazine -based structures (Figure). It was previously found out that some of them (all the investigated molecules were prepared and tested as the salts with hydrochloric acid) have shown relatively promising antimicrobial profile 8,9 . In general, the N-arylpiperazine moiety was regarded as so-called privileged structure 10 -it represented a class of the molecules capable of binding to multiple receptors with high affinity, inter alia by displaying key physicochemical characteristics that facilitated their ability to bind to them. The use of such fragment should allow the medicinal chemists to rapidly discover biologically active compounds across a broad range of therapeutic areas 11 . Following mentioned, the principal objective of current paper was to evaluate if these derivatives 8aB-8iB could potentially exhibit convenient physicochemical properties, which would be interactive applet of Molinspiration Cheminformatics software tool (Molinspiration Cheminformatics, Slovak Republic). The CLOGP 4.0 readouts were generated by using ChemBioDraw Ultra 11.0 program package (CambridgeSoft, USA). The outputs of log P calculated by Moriguchi method (MLOGP) were obtained by an interactive applet of Virtual Computational Chemistry Laboratory (VCChL) 13 . The ALOGP procedure of log P prediction was calculated using VCChL applet based on the approach developed by Ghose and Crippen 14 . The readouts of partition coefficient logarithms, which were experimentally estimated in octan-1-ol/buffer (pH=7.0) medium, were adopted from the research paper 15 of Sedlárová et al. (2007). The observed differences between calculated (CLOGP 4.0, MLOGP, miLogP 2.2 and ALOGP, respectively) and experimentally estimated (log P exp ) log P values were expressed by the Absolute Average Residual Sums (AARS). If the AARS output was in the range of 0.000-0.490, the MLOGP, CLOGP 4.0, miLogP 2.2 and ALOGP approach, respectively, was qualified as acceptable, while AARS value within the area of 0.500-0.999 indicated that respect predictor tool was viewed as disputable and, finally, if calculated AARS readout exceeded 1.000, it was then classified as an unacceptable predictive procedure 16 . Table : The descriptors which characterized currently in silico inspected derivatives 8aB-8iB III. # RESULTS AND DISCUSSION Presently in silico calculated outputs revealed that almost all the inspected molecules completely met the RO5 (Table ). In detail, the substances 8aB-8iB have shown MW less than 500, nON=7 and nOHNH=2 as well. The calculation of PSA in a classical way, however, was rather time consuming, because of the necessity to The data, which characterized molecular weight (MW), hydrogen-bond acceptors count (nON), hydrogen-bond donors count (nOHNH), topological polar surface area 12 (TPSA), rotatable bonds count (nrotb) and the predicted logarithms of partition coefficient for the system octan-1-ol/water by applying miLogP 2.2 substructure approach, of investigated compounds 8aB-8iB were calculated by using generate a reasonable 3D molecular geometry and to determine the surface itself. Additionally, such calculations required specialized software to generate the 3D molecular structures and to determine the surface. The methodology for the calculation of used TPSA was described in details in the paper 12 . Briefly, such procedure was based on the summation of tabulated surface contributions of polar fragments (atoms regarding also their environment). These fragment contributions were determined by least squares fitting to the single conformer 3D PSA for 34,810 drugs from the World Drug Index. Topological polar surface area (TPSA) provided results of practically the same quality as the classical 3D PSA, the calculations, however, were two to three orders of magnitude faster. The values of TPSA for all studied compounds 8aB-8iB were 74.27 Å 2 . As suggested by Clark 17 , the criterion for poor absorption of PSA>140 Å 2 appeared to be an efficient method of computationally screening large numbers of compounds. Following given, all the studied derivatives fulfilled the requirement for good absoprtion. Palm and coworkers 18 found out that excellent correlation could be obtained between dynamic polar van der Waals´ surface areas (PSA d ) and Caco-2 permeabilities when evaluated a series of antagonists of ?-adrenergic receptors (ARs). However, the major drawback of such parameter, it was computationally expensive, made it inappropriate for large database screening. However, the RO5 violation was clearly indicated in terms of the rotatable bonds count for the compounds bearing meta-/para-propoxy or para-butoxy side chain, namely 8eB, 8hB and 8iB, respectively. Such descriptor was a widely used filter following the finding that greater than ten rotatable bonds correlated with decreased oral bioavailability in animal models 5 . Additionally, the same derivatives have shown the predicted values of log P by CLOGP 4.0 approach higher than 5. Following the AARS readouts, the CLOGP 4.0 method was considered unacceptable because of providing ARRS=1.067. The main disadvantage of using such relatively unconvenient (but required, actually) fragmental-based procedure for current investigation of homological series 8aB-8iB was that it did not take into account the position of alkoxy side chain attached to lipophilic aromatic ring -identical values were observed when in silico evaluated corresponding positional isomers (Table ). Similar situation was encountered when analyzing the data related to Moriguchi MLOGP. Given fragmental approach did not reflect the position of attached alkoxy side string, as expected. Surprisingly, calculated output of AARS=0.258 identified this in silico procedure as an acceptable tool for the prediction of log P. It was also documented that MLOGP>4.15 was generated for para-butoxy derivative (8iB) only. Ghose and Crippen proposed that the qualifying range of calculated log P for drug-like molecules was from -0.4 to 5.6 by fragmental ALOGP procedure. The mean ALOGP was set to 2.3 and the preferred interval was 1.3-4.1, as published in 14 . Following given, propoxy-and butoxysubstituted derivatives were out of the range. In addition, the AARS for ALOGP was 0.506 so this procedure was considered disputable. When inspecting the AARS readout assigned to miLogP 2.2, relatively conventient value was indicated (0.478). It seemed that, except for the MLOGP procedure, the miLogP 2.2 could be considered an acceptable prediction method for the in silico investigation of concerned compounds. From structural viewpoint, currently studied derivatives could be also regarded as arylcarbamoyloxyaminopropanols, i.e. a class of the compounds which act as the antagonists of ?-ARs. Side effects of these drugs, which are connected with their central nervous system (CNS) activity, are well-known; for instance lethargy, depressions, psychoses 19,20 or visual hallucinations 21 , respectively. Early assessment of the physicochemical properties of potentially active CNS drugs in terms of their ability to cross blood-brain barrier is extremely important. Additionally, CNS active drugs have shown notably fewer rotatable bonds count (five or less) than other drug classes 26 . Considering the log P data (log P exp ), Hansch and Leo 27 found out that blood-brain barrier penetration was optimal when mentioned readout was within the interval of 1.5-2.7. For the complexity of information, in paper 26 were summarized some essential attributes of successful CNS drugs. Some of them were as follows: MW<450, CLOGP<5, nOHNH<3, nON<7, nrotb<8, hydrogen bonds<8 and PSA<60-70 Å 2 , respectively. Kelder et al. 22 previously found out that non -CNS drugs transported passively and transcellurlarly needed a PSA of 120 Å 2 or less, whereas the drugs can be targeted to the CNS with a PSA less than 60-70 Å 2 . On the other hand, following the van de Waterbeemd research 23 , the cutoff for PSA cutoff for CNS penetration was set to 90 Å 2 or below and molecular weight cutoff of 450. Levin suggested the molecular weight cutoff 400 or lower 24 . It was also previously documented that PSA value was dependent upon hydrogen bonding and donating atoms 25,26 . Because the hydrogen bonding was primarily associated with oxygen (O) and nitrogen (N) moieties in a molecule, then, if the sum of the N and O atoms in the structure was five or less, the compound has shown a high probability of entering the CNS. Moreover, if following difference CLOGP -(N+O) was higher than 0, then the compound had a high probability of entering the CNS. Rather than trying to predict absorption-related quantities, researchers tried to find out general principles to distinguish drug-like from non-drug-like molecules by inspecting the drugs and non-drugs databases. Current article was focused on the in silico evaluation of some highly lipophilic N-arylpiperazine based compounds containing 2-hydroxypropan-1,3-diyl connecting chain and substituted alkoxyphenylcarbamoyloxy fragment as well, in terms of their drug -like properties. It seemed that majority of all the substances inspected successfully met the criteria which were previously pioneered especially by Christopher Lipinski in his famous Rule of Five and Daniel Veber, respectively. It could be assumed that focused compounds would potentially shown good absorption and permeation after an oral administration. In addition, following the molecular weight, PSA (TPSA), hydrogen-bond acceptors count, rotatable bonds count, CLOGP 4.0 readouts and estimated log P exp data as well, it could be hypothesized that all currently inspected derivatives would not be able to cross blood-brain barrier passively and consequently not to involve CNS -related side effects. :![Figure : General chemical structure of currently in silico investigated compounds 8aB-8iB; 8aB: R=2-OCH 3 , 8bB: R=2-OC 2 H 5 , 8cB: R=3-OCH 3 , 8dB: R=3-OC 2 H 5 , 8eB: R=3-OC 3 H 7 , 8fB: R=4-OCH 3 , 8gB: R=4-OC 2 H 5 , 8hB: R=4-OC 3 H 7 , 8iB: R=4-OC 4 H 9 .](image-2.png "Figure :") ![Currently in silico studied compounds 8aB-8iB, chemically 3-[4-(3-trifluoromethylphenyl)piperazin-1-yl]--2-hydroxypropyl (2-/3-/4-alkoxyphenyl) carbamates (Figure), were previously in vitro antimicrobially screened against Gram-positive Staphylococcus aureus ATCC 6538, Gram-negative Escherichia coli CNCTC 377/79 and Candida albicans CCM 8186, a yeast, respectively, as the salts with hydrochloric acid 8,9 , i.e. 8a-8i, chemically 1-[3-(2-/3-/4-alkoxyphenylcarba-moyloxy)-2hydroxypropyl]-4-(3-trifluoromethylphenyl)pi-perazinium chlorides. Although they contained stereogenic centre, they were prepared and tested as racemates. The letter "B" in the entry (Table) meant the "base".](image-3.png "") Following current in silico calculated outputs and previously estimated data 15 as well, the compounds under the study 8aB-8iB probably would not permeate cross blood-brain barrier so they would not probably involve the CNS side effects. ## IV. CONCLUSION * Experimental and computational approaches to estimate solubility and permeability in drug discovery and development settings CALipinski FLombardo BWDominy PJFeeney 10.1016/S0169-409X(00)00129-0 Advanced Drug Delivery Reviews 46 2001 * Lead and drug-like compounds: the rule-of-five revolution CALipinski 10.1016/j.ddtec.2004.11.007 Drug Discovery Today: Technologies 1 2004 * Calculating log P oct from structures AJLeo 10.1021/cr00020a001 Chemical Reviews 93 1993 * Molecular properties that influence the oral bioavailability of drug candidates DFVeber SRJohnson H.-YCheng BRSmith KWWard KDKopple 10.1021/jm020017n Journal of Medicinal Chemistry 45 2002 * A practical view of "druggability THKeller APichota ZYin 10.1016/j.cbpa.2006.06.014 Current Opinion in Chemical Biology 10 2006 * Lessons from natural molecules JClardy CWalsch 10.1038/nature03194 Nature 432 2004 * Methods for drug discovery: Development of potent, selective, orally effective cholecystokinin antagonists BEEvans KERittle MGBock RMDipardo RMFreidinger WLWhitter GFLundell DFVeber PSAnderson RSChang VJLotti DJCerino TBChen PJKling KAKunkel JPSpringer JHirshfield 10.1021/jm00120a002 Journal of Medicinal Chemistry 31 1988 * The combinatorial synthesis of bicyclic privileged structures or privileged substructures DAHorton GTBourne MLSmythe 10.1021/cr020033s Chemical Reviews 103 2003 * Fast calculation of molecular polar surface area as a sum of fragment-based contributions and its application to the prediction of drug transport properties PErtl BRohde PSelzer 10.1021/jm000942e Journal of Medicinal Chemistry 43 2000 * Virtual computational chemistry laboratory -design and description IVTetko JGasteiger RTodeschini AMauri DLivingstone PErtl VAPalyulin EVRadchenko NSZefirov ASMakarenko VTanchuk Yu VVProkopenko 10.1007/s10822-005-8694-y Journal of Computer-Aided Molecular Design 19 2005 * Atomic physicochemical parameters for three-dimensional structure-directed quantitative structure-activity relationships. I. Partition coefficients as a measure of hydrophobicity AKGhose GMCrippen 10.1002/jcc.540070419 Journal of Computational Chemistry 7 1986 * Study of lipophilicity of phenylcarbamic acid derivatives with substituted N-phenylpiperazine moiety ESedlárová IMalík FAndriamainty SKe?ké?ová JCsöllei Farmaceutický Obzor 76 2007 * What accuracy statistics BAKitchenham LMPickard SMacdonell GShepperd MJ 2001 * Simple method of calculating octanol/water partition coefficient IMoriguchi SHirono QLiu YNakagome YMatsushita Chemical and Pharmaceutical Bulletin 40 1992 * Antimicrobial activity of meta -alkoxyphenylcarbamates containing substituted N -phenylpiperazine fragment IMalík MBukovský FAndriamainty JGali?inová 10.1590/S1517-838220120003000016 Brazilian Journal of Microbiology 43 2012 * Antimicrobial effect of paraalkoxyphenylcarbamic acid esters containing substituted N-phenylpiperazine moiety IMalík MBukovský FAndriamainty JGali?inová 10.1590/S1517-83822013000200018.reallymeasure DOI: 10.1049/ip- sen:20010506 IEE Proceedings-Software Engineering 2013 44 * Rapid calculation of polar molecular surface area and its application to the prediction of transport phenomena. 1. Prediction of intestinal absorption DEClark 10.1021/js9804011 Journal of Pharmaceutical Sciences 88 1999 * Correlation of drug absorption with molecular surface properties KPalm KLuthman A.-LUngell GStrandlund PArtursson 10.1021/js950285r Journal of Pharmaceutical Sciences 85 1996 * Propranolol-induced depression and psychosis WAParker Clinical Pharmacy 4 1985 * Propranolol-induced psychosis DJMcgahan AWojslaw VPrasad SBlankenship Drug Intelligence and Clinical Pharmacy 18 1984 * Visual hallucinations and metoprolol FJSirois 10.1176/appi.psy.47.6.537 Psychosomatics 47 2006 * Polar molecular surface as a dominating determinant for oral absorption and brain penetration of drugs JKelder PD JGrootenhuis DMBayada LP CDelbressine J.-PPloemen 10.1023/A:1015040217741 Pharmaceutical Research 16 1999 * Estimation of blood-brain barrier crossing of drugs using molecular size and shape, and H-bonding descriptors HVan De Waterbeemd GCamenish GFolkers JRChretien OARaevsky 10.3109/10611869808997889 Journal of Drug Targeting 6 1998 * Relationship of octanol/water partition coefficient and molecular weight to rat brain capillary permeability VALevin 10.1021/jm00180a022 Journal of Medicinal Chemistry 23 1980 * Prediction of polar surface area and drug transport processes using simple parameters and PLS statistics TÖsterberg UNorinder 10.1021/ci000065l Journal of Chemical Information and Computer Sciences 40 2000 * Medicinal chemical properties of successful central nervous system drugs HPajouhesh GRLenz 10.1602/neurorx.2.4.541 NeuroRx, The Journal of the American Society for Experimental Neuro Therapeutics 2 2005 * Substituent constant for correlation analysis in chemistry and biology CHansch AJLeo 10.1002/jps.2600690938 1979 Wiley 339 New York