Prof. Elizbar Elizbarashvili, D.Sc Technical University of Georgia 77 Kostava Street, Tbilisi, 0175 Georgia Email: eli

1. COMPUTER AIDED PREDICTION OF BIOLOGYCAL ACTIVITY OF NOVEL HETEROCYCLIC COMPOUNDS V. Kurtskhalia*, T. Matitaishvili, I. Lagvilava, E. Elizbarashvili Technical University of Georgia, Faculty of Chemical technology and metallurgy INTRODUCTION EXPERIMENTAL SYNTHESIS Today only the special chemical compounds undergo testing on biological activity, because the screening is very long-time and expensive process. To these special compounds commonly belong drugs; emulsifiers; test-modifiers; colorants and pigments used in food technology, body-care, etc. But the number of chemicals, to which humans have direct contact in everyday life, is rather larger and each of them, especially organic ones, has some biological activity1. Moreover, the number of existing potentially biological active compounds is expected to increase from about 500 to about 5000–10,000 in each years. Thus, experimental evaluation of potential biological activity is becoming more complicated2. Computer-aided structure–activity relationship analysis and molecular modeling are widely used now by pharmaceutical and non- pharmaceutical chemists to discover new lead compounds and optimize their structure and properties. However, the majority of available approaches are focused on a single macromolecular target, or on the only known pharmacological/biochemical action, and/or compounds from the same chemical series. Most (Q)SAR methods are focused on a single biological activity, whereas in reality each compound has both main and side pharmacological effects. Therefore, in the current report PASS C&T has been used for calculation of biological activity of novel heterocyclic systems. All of the chemicals used were of commercial grade and were further purified by recrystallisation and redistilled before use. The solvents used were spectroscopic grade. 1H NMR spectra were obtained by use of BRUKER WM-300 (300MHz). The IR spectra were obtained on a Thermo Nikolet spectrometer scanning between 4,000-400 cm-1 using KBr pellets. UV-Vis absorption spectra were measured in CF-26 spectrometer. Elemental analyses were performed using Heraeus CHNO-Rapid analyzer. Melting points were determined by Electrothermal 9100. For molecular modeling studies, structures were generated with the aid of Chem3D Ultra-9.00 and HyperChem-v.6.02, Wolfram Reserach Mathematics 6.0 software. Lone pairs of electrons and hydrogen atoms were added where appropriated. The equilibrium geometries of compounds were located using MM+ (for HyperChem) and MM2 (for Chem3D) functional set. In the next step, RHF calculation (semiempirical AM1 method, self-consistent field of Hartree - Fock) were performed and bond length, angles, torsion angles and partial charges have been calculated. Calculations were performed on a Intel (R) Core2 (TM) CPU 6600@ 2.4 GHz Pentium IV computer with 2 MB RAM. General procedure of preparation of aldehydes. To a stirred solution of bisphenol in ethanol (150 mL) was added the solution of sodium hydroxide (3.2 g, 80 mmol) in water (20 mL). The reaction mixture was heated up to 80 °C. Then, chloroform (9 mL, 0.11 mol) was added dropwise and the reaction mixture was boiled for a period of 1 hour. The excess of ethanol and chloroform was distilled. Hydrochloric acid was added subsequently until pH 5-6. The residue was dissolved in the minimal amount of diethyl ether and equal amount of saturated solution of Na2S2O5 was added. The mixture was kept for 24 hours and the precipitated solids were filtered off. The solid was dissolved in water and treated with sulfuric acid (10%). The yellow crystals were obtained by filtration and dried on air. General procedure of preparation of macrocyclic polyazomethines 1. Into the four-necked flask fitted with two dropping funnels, mechanical stirrer and reflux condenser 2-propanol (100 mL) was placed and heated until boiling. The solution of aldehydes (1.0 mmol) in 2-propanol (10 mL) and freshly obtained hydrazine hydrate (2.2 mmol, 50%) were added from different dropping funnels simultaneously for a period of 30 min. The reaction mixture had been heated for 1 hour and precipitated pale yellow crystals were isolated. ABSTRACT Today only the special chemical compounds undergo testing on biological activity, because the screening is very long-time and expensive process. To these special compounds commonly belong drugs; emulsifiers; test-modifiers; colorants and pigments used in food technology, body-care, etc. But the number of chemicals, to which humans have direct contact in everyday life, is rather larger and each of them, especially organic ones, has some biological activity. The one way to avoid of this problem is prediction of biological activity by special software. In the current report PASS C&T has been used for calculation of biological activity of novel heterocyclic systems. CONCLUSIONS The polyazomethines 1 were obtained in two steps from corresponding bis-phenols with consecutive conversions of carbonylation under Reimer- Tiemann conditions and condensation with hydrazine hydrate. The satisfactory purity of aldehydes have been achieved by purification via corresponding bisulphite derivatives (see scheme 1)3-5. Pyridone and dipyridone containing heterocyclic compounds 2-4 have been synthesized from corresponding amino anthraquinones with acylation, cyclization and aminolysis reactions. The spectral investigation of obtained compounds are in good agreemnet with calculated one. The posibility of formation of 1-4 is also confirmed with quantum-chemical calculation. ESTIMATION TABLE 1. Spectrum of Pharmacological Activity Predicted for Compound 1 A potential pharmacological activity of compounds 1-4, was estimated using a computer program PASS C&T (Prediction of Activity Spectrum for a Substance), which is capable of predicting some types of activity of a given compound upon analysis of its structural formula. The prognosis is based on the structure-activity relationships established by analysis of the data for more than 10,000 compounds forming the learning sample set. The PASS system either predicts the possible type of pharmacological activity or indicates a possible mechanism of the biological action. The activity prognosis has the form of a probability of the corresponding manifestations or their absence, since the learning sample set contains data on both definitely active and definitely inactive compounds. An analysis of the prognosis suggests the possible biological activity of compounds 1-4 studied. The experimental investigation showed that compounds 1-4 exhibited weak antibacterial activity with respect to both the Gram-positive and Gram-negative species (table 1). AIM The purpose of this work was to study the possibility of synthesizing the novel heterocyclic systems 1-4and to investigate the spectral and pharmacological properties of the synthesized compounds. Scheme 1. Synthesis of macrocyclic polyazomethines 1. CONTACT Figure 1. Geometrically optimized model of 1 Prof. Elizbar Elizbarashvili, D.Sc Technical University of Georgia 77 Kostava Street, Tbilisi, 0175 Georgia Email: elizbarashvili@gtu.ge Phone: (995 91) 191 -723 Website: http://www.gtu.ge/katedrebi/dep33/eliz/index.htm REFERENCES 1. Poroikov V., Filimonov D. (2001) Rational Approaches to Drug Design, Eds. H.-D. Holtje, W.Sippl, Prous Science, Barcelona, 403-407 2. Geronikaki A., Lagunin A., Poroikov V., Filimonov D., Hadjipavlou-Litina D., Vicini P. (2002). 13 (3/4), 457-471. 3. Elizbarashvili E., Matitaishvili T., Topuria Kh. Journal of Brazilian Chemical Society. 2007, 18, 6, 1254-1258. 4. Lagvilava I., Matitaishvili T., Iardalashvili I., Elizbarashvili E. Collection of Czechoslovak Chemical Communications, 2008, 74, 3, 409-418. 5. E. N. Elizbarashvili, I. V. Lagvilava, Sh. A. Samsoniya. Chemistry of Heterocyclic Compounds. 2005, 12, 1868-1869 Figure 2. Charge distribution in 1