Turk J Med Sci 31 (2001) 493-497 © TÜB‹TAK
Özlem TEM‹Z ARPACI QSARs of Some 5- or 6-Methyl-2-Substituted Benzoxazoles/Benzimidazoles against Candida albicans
Received: November 07, 2000 Abstract: QSAR analysis of some 5- or 6- using the BILIN statistical software. The methyl-2-substituted benzoxazoles/ resulting QSAR revealed that substitution at benzimidazoles was studied for the antifungal position Y with the CH2 group was significant activity against C. albicans using Hansch for the improved antifungal activity. analysis. Prediction for the lead optimization Moreover, hydrophobic properties of the in this QSAR analysis was described by the substituents at position R 2 are indicative for description of various hydrophobic, electronic, the antifungal activity against C. albicans. steric and structural parameters related to Key Words: Benzoxazole, Benzimidazole, Department of Pharmaceutical Chemistry, positions R 1, R 2, R 5, R 6, X, Y. The cross- Antifungal activity, QSAR, Hansch analysis Faculty of Pharmacy, Ankara University, validation method was also applied to the data Tando¤an 06100, Ankara - TURKEY set in order to prove the predictive power by
Introduction chemotherapy (5). New developments have included the Mycotic illnesses in humans are divided into three modification of existing drug molecules to eliminate groups: contagious skin and hair infections, toxicity and improve activity. In short, antifungal therapy noncontagious soilborne or airborne systemic infections has advanced rapidly in the last few years compared to and noncontagious foodborne toxemias. The responsible previous years and recommendations for treatment of organisms and methods of prevention and treatment fungal infections are likely to change in the near future as differ with each group. The prevalence of systemic fungal our understanding of fungal infections improves and new infections has increased significantly during the past antifungal therapies are discovered. decade. This increase is due to greater use of broad- Biologically active benzoxazoles have been known for spectrum antibiotics, immunosuppressive agents, central a long time and it was seen that position 2 is decisive for venous catheters, intensive care low birth weight infants, the biological activity, whereas position 5 determines the organ transplantation and the acquired immunodeficiency intensity of their activity (20-22). syndrome (AIDS) epidemic (1-13). Recently, we reported the synthesis and in vitro Significant antifungal chemotherapy began in 1903, antifungal activities of different 5- or 6-methyl-2- with the successful use of potassium iodide for the substitutedbenzoxazoles/benzimidazoles against the treatment of sporotrichosis. Except for the development fungus Candida albicans (23-25) (Figure 1). of flucytosine, there was little progress until the early 1970s and the development of the azole drugs. The R6 X current era, which is characterized largely by the YR2 modifications of azole drugs, began with miconazole and N ketoconazole and brought the agents fluconazole and R5 R1 itraconazole and ravuconazole, which can be given orally and have increased potency, decreased toxicity and a R1=H, Cl, F, NO2, CH3, OCH3 Y= —, CH2, CH2O, CH2S broader spectrum of activity (14-19). R2= H, Cl, Br, NH2, CH3, OCH3 X= O, NH R5=H, CH3 In the past 10 years there has been a major expansion R6= H, CH3 in the development of antifungal drugs, but there are still Figure 1. Antifungal active some benzoxazole and benzimidazole weaknesses in the range and scope of current antifungal derivatives against C. albicans.
493 QSARs of Some 5- or 6-Methyl-2-Substituted Benzoxazoles/Benzimidazoles against Candida albicans
An extrathermodynamic approach in the analysis of For the procedure of descriptor selection related to quantitative structure activity relationships (QSAR) has the activity among the candidate set of variables, forward been most widely and effectively used for theoretical drug step-wise multiple regression of elimination was applied. design. This method has also been called the Hansch During the development of the best fit model of approach and it assumes that the potency of a certain correlation equation, the minimum F value for entering biological activity exerted by a series of congeneric and removing the variables in the step-wise multiple compounds can be expressed in terms of a function of regression was taken to be 8.0, which is statistically various physicochemical (electronic, steric and significant at the 1% level of probability (31). hydrophobic) effects (26-29). This assumption is In order to judge the validity of the predictive power summarized in an equation as follows: of the QSAR, the cross-validation method is also applied f (biological activity) =f (electronic) +f (steric) +f to the original data set by removing a group of (hydrophobic) +[f (structural) +f (therotical)] compounds from the data in such a way that each If these functions could be formulated in an equation observation (compound) is deleted once and once only. showing certain effects favorable for the activity, For each reduced data set, a model is developed and the structural modifications that enhance such properties response values of the deleted observations are predicted would be expected to generate potent active compounds. from this model. The cross-validation method was also applied to the data set in order to judge the validity of Multiple regression analysis involving finding the best predictive power of the model and overall PRESS fit of a dependent variable (microbiological activity) to a (Predictive Residual Sum of Square) values were linear combination of independent variables (descriptors) calculated (32). by the least squares method was used (30). This is formally expressed as follows: Regression analysis and calculations were run on a PC using the BILIN statistical program package (33,34). In y=a0 +a1x1 +a2x2 +a3x3 + ……. + a nxn the equations, the figures in parentheses are the standard where y is related to the microbiological activity of a errors of the regression coefficients. For a given 2 compound, x 1, x 2, …… x n are the descriptor values equation, n is number of compounds, r denotes square related to the activity and a 0, a 1, a 2, ….. a n are the of the multiple correlation coefficients, F is the regression coefficients determined by the least square significance test and s represents the residual standard analysis. This equation is developed for each compound in deviation. our QSAR study. A congeneric set of 5- or 6-methyl-2- In this present study, QSAR analysis of some substitutedbenzoxazoles/benzimidazoles 1-25 were antifungal active 5- or 6-methyl-2- considered in this study. The candidate set of variables substitutedbenzoxazoles/benzimidazoles, 1-25 given in used in this analysis includes hydrophobic, electronic, Figure 1, against C. albicans were determined using steric and structural parameters. For the structural
Hansch analysis. variables I R6 was defined as 0 for possesing H, 1 for
possesing CH 3 at position R 6; I R5 was defined as 0 for
possesing H, 1 for possesing CH 3 at position R5; IyO was Experimental defined as 1 for possesing O atom; IyCH 2 was defined as
Multivariate Hansch analysis in QSAR has been most 1 for possesing CH 2 group; IyCH2O was defined as 1 for widely and effectively used for theoretical drug design possesing CH2O group; IyS was defined as 1 for possesing due to various physicochemical (electronic, steric and CH2S group at position Y and Ix was defined as 1 for hydrophobic) parameters and structural indicator possesing O atom, 0 for NH atoms at position X. The parameters being used together. The assumption can be hydrogen donating/accepting capabilities (H DONOR/HACCEPT) formulated as given in the equation below. of the substituents at R 1, R 2, R 5, R 6 were the indicator Σ Σ variables. log1/C= aiIi + bjXj + c where I is the structural indicator parameters and X is The screened physicochemical parameters in this i j π the physicochemical variables. QSAR study were for the hydrophobic effects, F (field
494 Ö. TEM‹Z ARPACI
effect), R (resonance effect) and σ as the electronic Results and Discussion influence and MR, MW, Es, Es-V, Verloop’s STERIMOL Predictions for the lead optimization in this QSAR parameters (L, B 1, B 4) for the steric interactions of the analysis of antifungal active 5- or 6-methyl substituents R 1, R 2, R 5 and R 6. Values for all candidate benzoxazoles/benzimidazoles were described by the physicochemical variables used in this QSAR study were description of various hydrophobic, electronic, steric and taken from the table of Hansch et al. (14). The values of structural parameters related to positions R 1, R2, R5, R6, the descriptors related to the activity among the X, Y. One leave out cross-validation method was also candidate test set of variables of the best results in the applied to the data set in order to prove the predictive QSAR analysis are shown in the Table. power by using the BILIN statistical software. Compounds
Table. Compounds and the parameters used in the best equation.
R6 X YR2 N R5 R1
π Com. R1 R2 R5 R6 XY R2 IyCH2 MIC Obs. Cal. Residual No: µg/mL log1/C log1/C
1ClHCH3 HO 0 0 25 3.989 3.996 -0.007 2 OCH3 HCH3 HO 0 0 25 3.980 3.996 -0.015 3NO2 HCH3 HO 0 0 25 4.007 3.996 0.011 4ClClCH3 HO 0.71 0 25 4.046 4.024 0.023 5 CHH3 CH3 CH3 HO 0.56 0 25 3.977 4.018 -0.041 6 OCH3 OCH3 CH3 HO -0.02 0 25 4.032 3.995 0.037 7ClHHCH3 O 0 0 25 3.989 3.996 -0.007 8 OCH3 HHCH3 O 0 0 25 3.980 3.996 -0.015 9FHHCH3 O 0 0 25 3.958 3.996 -0.038 10 NO2 HHCH3 O 0 0 25 4.007 3.996 0.011 11 Cl Cl H CH3 O 0.71 0 25 4.046 4.024 0.023 12 CH3 CH3 HCH3 O 0.56 0 25 3.977 4.018 -0.041 13 OCH3 OCH3 HCH3 O -0.02 0 25 4.032 3.995 0.037
14HHCH3 HO CH2 0 1 12.5 4.251 4.298 -0.047
15 H Br CH3 HO CH2 0.86 1 12.5 4.383 4.332 0.051
16 H NH2 CH3 HO CH2 -1.23 1 12.5 4.280 4.250 0.030
17 H H H CH3 OCH2 0 1 12.5 4.251 4.298 -0.047
18HHCH3 HNHCH2 0 1 12.5 4.249 4.298 -0.049
19 H Cl CH3 HNHCH2 0.71 1 12.5 4.312 4.326 -0.014
20 H Br CH3 HNHCH2 0.86 1 12.5 4.382 4.332 0.049
21 H NH2 CH3 HNHCH2 -1.23 1 12.5 4.278 4.250 0.028
22HHCH3 HO CH2O 0 0 25 3.980 3.996 -0.015
23HHCH3 HO CH2S 0 0 25 4.009 3.996 0.013
24HHCH3 HNHCH2S 0 0 25 4.007 3.996 0.011
25 H Cl CH3 HNHCH2O 0.71 0 25 4.037 4.024 0.014
495 QSARs of Some 5- or 6-Methyl-2-Substituted Benzoxazoles/Benzimidazoles against Candida albicans
and parameters used in the best equation and observed, F: Fischer value (measure of the statistical calculated and residual values are given in the Table. The significance) plot of the observed and calculated C values of the Q2: squared cross-validation regression coefficient growth inhibitory activity of compounds 1-25 against C. albicans using the best equation is shown in Figure 2. s-PRESS: standard deviation of cross-validation predictions Calculated Q 2 and s-PRESS values show that the 5 predictive power of this QSAR model is significant. 4.8 Predictions for the lead optimization in this set of compounds can be summarized as follows: 4.6 - The activity contributions obtained in this QSAR 4.4 model show that the indicator structural parameter IyCH C (observed) 2 is significant for the activity. So substitution at position Y 4.2 with the CH 2 group is more important than with CH 2O
and CH 2S groups for the antifungal activity against C. 4 4.1 4.2 4.3 4.4 4.5 albicans. C (calculated) - The substitution at R 2 is found to be more Figure 2. Plot of the observed and calculated C values of the growth significant than R 1, R 5, R 6 and X in improving the inhibitory activity of compounds 1-25 against C. albicans. antifungal activity. - Hydrophobic properties of the substituents at The best equation and cross-validation results position R2 are more indicative than electronic and steric observed in this QSAR study are: properties of the antifungal activity against C. albicans. π log1/C=+0.0393 (±0.027) R2 +0.303 (±0.030)
IyCH2 +3.996 (±0.017) Correspondence author: n=25; r=0.97; s=0.03; F=224.38; Özlem TEM‹Z ARPACI p<0.001; Q2=0.93; s-PRESS = 0.03 Ankara University, Faculty of Pharmacy, n: number of compounds Department of Pharmaceutical Chemistry, r: correlation coefficient Tando¤an 06100, Ankara - TURKEY s: standard deviation
References 1. T.C. Wu, “On the development of 5. J.R. Graybill, “The future of antifungal 9. J.N. Anker, N.L. Popele, P.J. Sauer, antifungal agents: Perspective of the therapy” Clin. Infect. Dis., 22 (2), 166- “Antifungal agents in neonatal systemic U.S. Food and Drug Administration” 168, 1996. candidiasis” J. Antimic. Chemother, (B), Clin. Infect. Dis., 19 (1), 54-58, 1994. 1391-1397, 1995. 6. C.E. Gonzalez, D. Venzon, S. Lee, B.U. 2. P.A. Pizzo, L.S. Young, “Limitations of Müeller, P.A., Pizzo, T.J. Walsh, “Risk 10. G.D. Gilfillan, D.J. Sullivan, K. Haynes, current antimicrobial therapy in the factors for fungemia in children infected T. Parkinson, D.C. Colemann, N.A. immunosuppressed host” Am. J. Med., with HIV: a case control study” Clin. Gow, “Candida dubliniensis: phylogency 76, 101-110, 1984. Infect. Dis., 23, 515-521, 1996. and putative virulence factors” Microbiology, 144, 829-838, 1998. 3. J.R. Graybill, “Future directions of 7. F. Meunier, “Targetting fungi: a antifungal chemotherapy” Clin. Infect. challenge” Am. J. Med., 29, 60-67, Dis., 14 (1), 170-181, 1992. 1995. 4. M.G. Rinaldi, “Laboratory evaluation of 8. D.W. Warnock, “Fungal complications antifungal agents: a brief overview” Clin. of transplantation: diagnosis, treatment Infect. Dis., 14 (1), 130-133, 1992. and prevention” J. Antimic. Chemother, 36, 73-90, 1995.
496 Ö. TEM‹Z ARPACI
11. M.A. Pfaller, S.A., Messer, S. Gee, S. 19. D.J. Sheehan, C.A. Hitchcock, C.M. 26. C. Hansch, “On the structure of Joly, C. Pujol, D.J. Sullivan, D.C. Sibley, “Current and emerging azole Medicinal chemistry” J. Med. Chem., Colemann, D.R. Soll, “ In vitro antifungal agents” Clin. Microbiol. Rev., 19(1), 1-6, 1976. susceptibilities of Candida dubliniensis 12, 40-70, 1999. 27. T. Fujita, “The extrathermodynamic isolates tested against the new triazole 20. W. G. Bywater, W. R. Colemann, O. approach to drug design” and echinocandin antifungal agents” J. Kamm, H. H. Merritt, “Synthetic Comprehensive Medicinal Chemistry, Clin. Microbiol., 37, 870-872, 1999. anticonvulsants, the preparation and Pergamon, Oxford, Vol. 4, pp. 497- 12. I. Polachek, J. Strahilevitz, D. Sullivan, properties of some benzoxazoles” J. 561, 1990. S. Donnelly, S. Salkin, D.C. Colemann, Ame. Chem. Soc., 67, 905-907, 1945. 28. C. Hansch, S. D. Rockwell, P.Y. Jow, A. “Recovery of Candida dubliniensis from 21. D. W. Dunwell, D. Evans, “Synthesis and Leo, E. Steller, C. Hansch, “Substituent non-human immunodeficiency cirus- antiinflammatory activity of some 2- constants for correlation analysis” J. infected patient in Israel” J. Clin. aryl-6-benzoxazoleacetic acid Med. Chem., 20 (2), 304-306, 1977. Microbiol., 38, 170-174, 2000. derivatives” J. Med. Chem., 20, 797- 29. R. Franke, “Theoretical Drug Design 13. A. Marot, L. Grimaud, S. Bouterige, V. 801, 1977. Methods”, “Theoretical Drug Design Apaire, D.J. Sullivan, R. Robert, “New 22. M. Pedini, G. De Meo, A. Ricci, L. Methods”, Elseiver, Amsterdam, pp. monoclonocal antibody specific for Bastianin, P. Jaquignon, “New 256, 1984. Candida albicans germ tube” J. Clin. heterocyclic derivatives of Microbiol., 38, 61-67, 2000. 30. K. C. Chu, “The quantitative analysis of benzimidazole with germicidal activity: structure-activity relationships” 14. D.A. Stevens, “Therapy for opportunistic II 2-(5’-nitro-2’furyl or 2’-thienyl) “Burger’s Medicinal Chemistry”, Wiley, fungal infections: past, present and benzimidazoles with different New York, pp. 393-419, 1980. future” Indian J. Cancer, 32, 1-9, 1995. substituents in the 5-position” Il Farmaco, 45(3), 303-312, 1990. 31. E. fiener, ‹. Yalç›n, E. Sungur, “QSARs 15. A. Espinel-Ingroff, S. Shadomy, “In vivo of some novel antibacterial and in vitro evaluation of antifungal 23. ‹. Yalç›n, E. fiener, T. Özden, S. Özden, benzimidazoles, benzoxazoles and agents” Eur. J. Clin. Microbiol. Infect. A. Ak›n, “Synthesis and microbiological oxazolopyridines against an enteric Dis., 8, 352-361, 1989. activity of 5-methyl-2-[p-substituted gram-negative rod: K. pneumoniae ” phenyl]benzoxazoles” Eur. J. Med. 16. J.W. Hiemenz, T.J. Walsh, “Lipid QSAR, 10, 223-228, 1991. Chem., 25, 705-708, 1990. formulations of amphotericin B: recent 32. S. Wold, “Validation of QSARs” QSAR, progress and future direction” Clin. 24. Ö. Temiz, ‹. Ören, E. fiener, ‹. Yalç›n, N. 10, 191-193, 1991. Infect. Dis., 22(2), 133-144, 1996. Uçartürk, “Synthesis and microbiological activity of some novel 33. H. Kubinyi, BASF AG ZHF/G-A30 17. K. Hata, H. Kimura, T. Toyosawa, T. 5- or 6-methyl-2-(2,4- 67045 Ludwigshafen, Germany. Nakamura, K. Katsu, “In vitro and in disubstitutedphenyl)benzoxazole vivo antifungal activities of ER-30346, 34. H. Kubinyi, O.-H. Kehrhahn, derivatives” Il Farmaco, 53, 337-341, A novel oral triazole with a broad “Quantitative activity relationships, 6. 1998. antifungal spectrum” Antimicrob. Nonlinear dependence of biological Agents Chemother, 40, 2237-2242, 25. ‹. Ören, Ö. Temiz, ‹. Yalç›n, E. fiener, A. activity on hydrophobic character, 1996. Ak›n, N. Uçartürk, “Synthesis and calculation procedures for the bilinear microbiological activity of 5(or 6) model” Arzneim.-Forsch (Drug Res.), 18. J.C. Fung, E. Huczko, B. Minassian, –methyl-2-substitutedbenzoxazole and 28, 598-601, 1978. D.P. Bonner, “In vitro activity of a new benzimidazole derivatives” Arzneim. oral triazole, ER-30306” Antimicrob. Forsch., 47, 1393-1397, 1997. Agents Chemother, 42, 313-318, 1998.
497