You are on page 1of 8

Madhuri Baghel, ARPB, 2011; Vol 1(2

)
(Research Article)

ISSN 2250 - 0744

QSAR ANALYSIS OF A SERIES OF
VOL - 1(2) OCT-DECEMBER 2011

1-(2-(2,2,2-TRIFLUOROETHOXY) ETHYL-1 HPYRAZOLO[4,3-D]PYRIMIDINES: AS POTENT PHOSPHODIESTERASE 5 (PDE5) INHIBITORS
*M. Baghel Parul Institute of Pharmacy, Limda, Waghodia, Vadodara, Gujarat India.

ABSTRACT:
A series of recently synthesized 1-(2-(2,2,2-Trifluoroethoxy) ethyl-1 H-pyrazolo[4,3-d]pyrimidines as Potent Phosphodiesterase 5 (PDE5) inhibitor was subjected to Quantitative Structure Activity Relationship (QSAR) Analysis . QSAR investigation based on semiemperical AM1 calculations reveals that thermodynamic, electronic & topology parameters are responsible for PDE5 inhibitory activity. Attempt in correlating the derived physiochemical properties with the PDE5 inhibitory activity resulted in some statistically significant QSAR models with good predictive ability. Our study explored some interesting findings for the design of potent new class of PDE5 inhibitors.

Received on 12/12/2011 Revised on 20/12/2011 Accepted on 23/12/2011 *Corresponding author
Mrs. Madhuri Baghel Parul Institute of Pharmacy, Limda, Waghodia, Vadodara, Gujarat India. E-mailbanchhormadhuri@gmail.com

KEYWORDS: QSAR, PDE 5 inhibitors, Sildenafil, Erectile
Dysfunction.

www.arpb.info

Page 139

tadalafil & Verdenafil.2.0003 0.00065 PDE5 PIC50 (µM) -0. Recently Michael B.0744 INTRODUCTION Sildenafil.294296 -0.arpb.3262109 -0. no molecular modellingQSAR studies have been reported for this new lead structure.3137662 Sildenafil. Vol 1(2) (Research Article) ISSN 2250 . Greater deal of attention has been paid to the pyrazolo pyrimidine heterocyclic analogues as PDE5 inhibitors ever since the discovery of Comp 1 2 3 4 R1 Et n-Pr i-Pr CH2CF3 PDE5 IC50 (µM) 0. al. Table No 1: The PDE5 inhibitory activities of 1-(2-(2. Owing to our special interest & in continuation of our research effort towards QSAR analysis.3- d]pyrimidines are listed in Table No 1. ARPB. www.2.2-Trifluoroethoxy) ethyl-1 Hpyrazolo[4.2.3-d] pyrimidines as PDE 5 inhibitor3.. et. 2011. is successfully used in the treatment of erectile dysfunction (ED) and pulmonary arterial hypertension 1 MATERIALS & METHODS The PDE5 inhibitory activities of 1-(2-(2. cGMP interacts with protein kinase G which leads to the reduction of intracellular Ca+ levels and vasodilation2.Madhuri Baghel. T.3-d]pyrimidines (PAH). In spite of this spectacular & unprecedented rate of progress in this therapeutic area.2- trifluoroethoxy)ethyl)1H-pyrazolo-[4. Sildenafil acts by competitive inhibition of phosphodiesterase 5 (PDE5) . A. sold as Viagra and Revatio.00086 0.5’-monophosphate (cGMP) levels by converting cGMP to GMP. we attempt to subject this new series of molecule for an effective QSAR analysis. cGMP levels rise resulting in vasodilation of the vascular endothelium allowing for the treatment of ED and PAH. Brad A.2Trifluoroethoxy) ethyl-1 H-pyrazolo[4.2838588 -0.0004 0. have reported a novel series of 1-(2(2.info Page 140 . PDE5 regulates guanosine cyclic 3’. By inhibiting PDE5.

00315 -0.0744 PDE5 PIC50 (µM) PDE5 R5 IC50 (µM) NH R 5 PDE5 IC50 (µM) Comp NR3R4 N PDE5 PIC50 (µM) 1 1 Me 0.Madhuri Baghel.info Page 141 .00979 -0.00035 -0.arpb.00392 -0. Vol 1(2) (Research Article) Comp R 2 ISSN 2250 .00648 -0.4569496 Et 0.2594793 Et 0. 2011.3997299 Et 0.00328 -0. PDE5 inhibitory activity data (IC50 value in µM) were converted to negative logarithm (pIC50) to get the linear relationship in the QSAR equation.2615125 7 Me 0.3954588 A dataset of 20 molecules have been taken from literature3.330374 8 Et 0.4155043 10 Et 0.00007 -0.2893575 Me 0.00014 -0.00296 -0. ARPB.2471725 6 Me 0.402556 5 5 Et 0.2406796 6 Et 0.00015 -0.3338142 3 3 Me 0.3217601 2 2 Et 0.00184 -0.00078 -0. The sketched 2D structures were converted to 3D structures and were subjected to energy minimization using www.00009 -0.3656064 4 4 Et 0.4977062 Et 0.00094 -0.4947157 9 Et 0.00952 -0.00101 -0.

13925. Quantitative model building was accomplished through sequential multiple regression analysis using the method of least square in val_stat software.F=0. 3 www. F=1.9279( ± 300.4 Vander Walls Force values of the activity. 788.std=1.161132( ± 36.36273e+008( ± 3.482685 performed.r=0.1801e+008)] n=14.90969)] + NON1. 1a: BA = [32.9 % confidence interval were not considered whilst deriving QSAR models. The statistical quality of the models was gauged by parameters like correlation 143. fisher’s Fvalue.001 kcal/mol and applying the theory of AM1 Hamiltonian using closed shell restricted wave function. electronic.7718)] + BEND [-0. Sum of squared prediction errors called predictive residual sum squares (PRESS) are calculated as the sum of squares of the 1 differences between predicted and observed 2 EHOMO = Energy of highest occupied molecular orbital BEND = Bend energy NON1. bootstrapping technique and randomization test were 0.4VDW [-1. optimized The 20 RESULT & DISCUSSION Several models were developed but for the sake of brevity only two statically significant models were discussed here. To ascertain the predictivity of the model. Vol 1(2) (Research Article) ISSN 2250 .Madhuri Baghel. root mean square (RMS) gradient to 0.info Page 142 .r^2=0.315331.4VDW ± 1.093( ± 154.5089e+009)] n=15. internal validation using Leave One Out (LOO) cross validation process. steric & topology parameters were calculated for the energy minimized structures4 and from geometrically MM2 server. variance=3305. as it is considerably smaller than its regression coefficient. descriptor is gauged from its standard error. The correlation matrix among the various predictor variables was examined regularly in order to avoid simple colinearity problem. ARPB.5 and those are not significant at 99.39( ± 1242. thermodynamic.422)] + BEND 1. r=0.24666e+008( r^2=0.101618( ± 0.variance=129. 3 [1.869)] + EHOMO [- coefficient (r) or coefficient of determination (r2). 2011.17)] + EHOMO1 [2 derivatives selected for the study were divided into training set of 15 compounds and test set of 05 compounds by random selection method. standard error of estimate (s).419622)] + NON1.4VDW = Non 1. fixing maximum iteration limit to 1000.74908.12438( ± [-8.arpb.355653. In order to cover wide range of physiochemical properties to describe the observed activity in a better way.19. Std=5.126489. The parameters having intercorrelation above 0.0744 The significance of individual semiemperical quantum mechanics module implemented on molecular orbital package (MOPAC) version.68872 Optimized model no.561543. Model No 1: BA = [-1184.

4 VDW are negatively contributing to PDE5 inhibitory activity.std=0.arpb.0213)] + EHOMO parametric equation of Model 1 describes the PDE5 inhibitory activity of 15 analogs as a function of their electronic (EHOMO) & thermodynamic (BEND. r^2=0.0795875)] + NON1.79 The derived model explains 88. To validate the derived models parametric equation of Model 2 describes the PDE5 inhibitory activity of 15 analogs as a function of their electronic (EHOMO) and steric (CSEV) & topology (TVC) properties.2378 (± 271.1678)] + TVC5 [-42. 2c: BA = [115.49686.45 869.1305 (± 8.9482)] + EHOMO +EHOMO [-79.9732(± 7.0269)] + EHOMO prediction (Spress).r^2=0.054)] + EHOMO [4.581114( ± 16.8886( ± 20.31.941443.r=0.27124)] + CSEV [-1.881 (± 38.122184.5)] + CSEV4 [11. 2011.769966.28796 (± 4. The tri significance of the model. coefficient crossvalidated (Q2).01 in the randomized biological activity test revealed that the results are not based on chance correlation.r=0.925693.variance=132. [6.14345)] n=12.F=17. std=6.30255)] + TVC [0.2794(± 56.0268)] n=15. The models Q2 value > 0.4 VDW) Parameters.47946)] + CSEV [-1. The presence of outlier was further confirmed by 4 externally.41075 (± 1.std=0.01682)] + BEND [-0. standard error of prediction (SDEP).9654 The derived model explains 85. The tri variance=3654.856907.2864)] n=14.std=0.F=0.F=20.569245)] + TVC [2.variance=8.295131.variance=2.22358 (± 0. The satisfactory value of internal validation.71 024. F=1.4 supported the predictive ability and [12. biological activity of the test set molecules were predicted using models derived from training set.6595( ± 79. 1b: BA = [57.0744 Model no.13197)] + TVC [3.63 % of variance in observed activity.463969 Optimized model no.29501)] n=13.323 (± 72.variance=4.r^2=0.24687.00053 (± 4. EHOMO is contributing positively while BEND & NON 1.info Page 143 .23787e+008( ± 6. standard correlation deviation of [16. 2: BA = [-983.14943. NON 1.3706)] + CSEV [-2. r=0.r=0.6512( ± 146.F=10.93784 (± 2.25(± 1201.0322 (± 4. 118.349548.211156.r^2=0.Madhuri Baghel.r^2=0. CSEV = Connoly Solvent excluded volume TVC = Total valence connectivity 5 www.r=0.16)] Optimized model no. Vol 1(2) (Research Article) ISSN 2250 .05 226.std=1. 2a: BA = [81.877477.0416 Optimized model no.69 % of variance in observed activity.30142 (± 30. 2b: BA = [145.886315. bootstrapping squared correlation coefficient (r2 bsp) and chance correlation <0.2019 Optimized model no.4VDW [-1.01133e+007)] n=13.0514258( ± 0. ARPB.04509.143257.

4 VDW that electrophilicity is important for PDE 5 inhibitory activity.213 477 0. The electron withdrawing groups increases the HOMO energy and in turn increases the electrophilicity of the ligands.0744 Predicted. & 17 in Model 2. r2. calculated activity. r2.00 1 Q 0.000000 NON 1.info Page 144 .076848 0.2250 51 0. F-value.6486 75 0. Model 2 shows better r.4 VDW 1.169920 Model 2 BEND 1.arpb.320002 TVC Fig. EHOMO BEND NON 1.7194 36 Spress Sdep 0.Madhuri Baghel. Table 2: Validation parameters for Model 1 & Model 2.8411 06 2 0. Compound 6 and 17 were omitted stepwise upon deriving model 1 & compound 3. predicted and calculated activity for training molecules of Model 2 is given in figure 1-2. 2011. hence model 2 is better and significant for designing new analogues.9115 85 S boot 0.000000 0.438175 0. Z.000000 www. This in turn would increase the PDE5 inhibitory activity.value and descriptors of training and test set molecules for Model 2 is shown in Table No 4.000000 0. 1: Experimental versus Calculated activity for Training set molecules of Model 2 Both the models are statically significant in terms of r.00 1 <0. Model 1 EHOMO EHOMO CSEV TVC 1.2752 95 0.071 4711 chance <0.000000 0.504505 CSEV 1. higher Z-score value. F-value. Retention of these compounds in model lowers the stastical significance hence QSAR analysis was performed without it. 6. Table 3: Correlation Matrix among parameters for Model 1 & 2.000000 0.1837 53 Fig. r2bs and Q2 values. 2: Experimental versus Predicted activity for Training set molecules of Model 2 The Positive contribution of EHOMO suggests 1.534400 EHOMO 1. r2bs and Q2. The correlation between observed. Deletion of outlier showed improved r2 value and cross-validated Q2 value of the model. Mod r2 el boot 1 0. ARPB. Validation parameters of both the models are shown in Table No 2. Vol 1(2) (Research Article) ISSN 2250 .5 (shown in Table No 3). Cross correlations of the descriptors are lower than 0.3308 64 0.

3: HOMO molecular orbital of most active molecule of series (com-16) Topological descriptor helps to differentiate the molecules according mostly to their size. 2011.437 -5.44806 2.20597 2.42343 1.697 -4.21875 EHOMO -6.0324 24.479516 0.788111 2. Vol 1(2) (Research Article) ISSN 2250 .641489 1.0035 -4.364 28.65759 2. not in contact with the solvent.150746 -1.988 -1.0744 Table 4: Observed.49 -5.94631 -0.92534 0.88251 3. Comp Calculated Activity -4.0921 -5.2267 29. predicted.96307 1.58903 -4.6547 32. suggest that increase in molecular branching & Page 145 .2751 -4.06591 -0.585275 -1.9202 28.05504 -0.684 -3.023 -6.113172 -0.382 -5.9435 -4. The moleculetopology index is calculated based only on the framework of a molecule.424 Descriptor BEND 29.491367 -2.5137 -4.41762 1. suggesting increase in the bulkiness of the substituents and molecular solvent excluded surface area is not conducive to the activity.9384 29.026 -6.Madhuri Baghel.99513 Z.419 -5.2732 31.395 -5.89783 -7. degree of branching.714845 Fig.67163 -6. 3 shows the HOMO orbital of the active molecule of the series (com-16).7493 28.024 -6.Value Loo Predicted Activity -2. represents the surface area that is www.898146 -0.6772 -5. The TVC is topology descriptor.603731 -2.459 -5. although some nonbond forces are important in QSAR or QSPR6.502 27.9165 27.2128 -5.01475 -0. The descriptor bears negative coefficient in the model.02 -5.099 30.5654 -4.4729 -4.4 VDW -3.363 -5.90265 -4.425 -5.373 -5.4595 -4.9837 31.77003 -3.25216 2.info shape. flexibility and overall Connolly’s solvent Excluded Volume.8823 -5.arpb.14837 -0. Molecular topology is a very useful tool for describing molecular structures and has been used for efficient analysis of QSAR data5.81326 -3.56119 1.734 -5.97 -5. ARPB.4715 1 2 4 5 7 8 9 10 11 12 13 14 15 16 18 19 20 -0.8531 -3.341789 -3.8131 29.415 -5.7159 3.7602 29.4446 -4.7122 30.3176 0.4082 29.98772 2. a steric descriptor.1851 NON 1. calculated activity and descriptors of training and test set molecules for Model 2. Positive coefficient in the model. Fig. contributing positively to activity.79301 0.366105 -1.70023 3.3246 -3.

B.0: For windows and Macintosh 5. phosphodiestrase 5 (PDE5) inhibitors. S. Beijing. Sildenafil in erectile dysfunction: A critical review. Rigatti and F. S. Med. J.arpb. S.). Res. A. Environmental Science Press of China. Soler. García-Domenech. Z. enzyme Find. Sandra. ARPB. F. 3. J. Chem 3D Version 10. 5.. Y. its Clin. Methods Exp. Phosphodiesterase-5 aspects.0 Cambridge soft. Gálvez. Bio. 2. K. Liansheng. Opin. 6. A. H.Madhuri Baghel. K. Michael.com. N. C. 32: 272–284 (1994). 4. 1-(2-ethoxyethyl)-1H-pyrazolo [4. and Patil. Chem. Curr. J. K. 19: 241 (2003). E. Lett. 20 : 3125-3128 (2010). 326 (1992. T. Kulkarni. Robert. J. Inf. Michael. V. Sci. REFERENCES 1. J. W. The finding of the study will be helpful in the design of potent PDE 5 inhibitors. Comput. Chem. Montorsi.3-d]pyrimidines as potent www. Topological approach to drug design. 2011. O. CONCLUSION QSAR studies provide deeper insight into the mechanism of action of compounds that ultimately becomes of great importance in modification of the structure of compounds. although the negative coefficient of CSEV suggest that too much bulkier substituents may appear to be detrimental to the PDE 5 inhibitory activity. Med. S. Zhengliang. P. Brad.0744 flexibility increases the activity. P. Jacobsen. A. Ying and R. 26: 789 (2004). R. Pharmacol. David. John. F. Vol 1(2) (Research Article) ISSN 2250 . Brain. de Julián-Ortiz and R. Molecular Connectivity and Molecular Structureactivity. W.info Page 146 . C.