Two-Dimensional QSAR of Herpes Simplex Virus 2 Thymidine Kinase Inhibitors - Coursework Example

Two-Dimensional QSAR of Herpes Simplex Virus 2 (HSV 2) Thymidine Kinase Inhibitors Workshop assignment Quantitative structure activity relationship is a program used to accelerate drug discovery process, by predicts the activity of the compounds based on physico-chemical properties. Thus, the main purpose of this workshop is to identify a drug with high inhibitory activity more than Acyclovir. Herpes simples is virus disease affects either upper human body (HSV1) or genital area in lower human body (HSV2). Both can be treated with Acyclovir that is a competitive inhibitor of thymidine kinase. Hence, searching for other phenylguanine analogues (the core base of Acyclovir structure) could be helpful in treating HSV. Several parameters are implicated in QSAR model in a process to discover the new drug. These are lipophilicity, ionization and polarity, steric factors and molar refractory. According to this workshop the most important parameters are Pi (lipophilicity) and MR (molar refractory). As a result Hansch equation has been made and predictive inhibitory activity has done consequently. Based on actual and predictive activity Hansch equation was not accurate and applicable for the all phenylguanine analogues. In fact, the chosen of which chemical is good to be synthesized relays on the difference result of the predictive and actual activity. In addition, many factors affect the compound in therapeutic activity such as bioavailability, stability and solubility. Furthermore, because this model depends on mathematical calculation for an ideal process it is hard to be carried out in practical of biological test with same satisfaction level. In summary, Pi and MR were quietly successful chosen descriptors to initiate hansch equation which was relatively useful to predict activity. But its limitation makes it not very useful way in bio system. Introduction: In the field of pharmacology, medicine and biotechnology, drug discovery is the process in which new candidate medications get discovered. Historically, the process of discovering drugs involved identification of active ingredients by serendipitous discovery from traditional remedies. After this, chemical libraries involving small molecules, extracts and natural products were screened in whole organism or intact cells. This was done with a view to identify key substances, which have a desirable therapeutic effect. This process was called classical pharmacology. Since the discovery of the sequencing of human genome that could allow rapid synthesizing and cloning of large amounts of purified proteins, the use of high throughput screening involving large compound libraries has become common. The reverse pharmacology, which has been hypothesized to be modify diseases has progressively become a common practice. Modern drug discovery involves identifying screening hits, and optimizing these hits with a view to increase selectivity, affinity, metabolic stability, potency and oral bioavailability. After the compound that fulfill these requirements has been identifies, the process of developing the drug begins. This is done prior to clinical trials. This process, however, has been relatively slow. Quantitative Structural Activity Relationship model (QSAR) was established with a sole purpose of accelerating the drug discovery process. This model builds on practical models for predicting the quantities such as toxic potential and binding affinity of the existing molecules. The model is carried out by this computational research and is based on physico-chemical properties (Lill, 2007). Herpes Simplex is a disease that effects humans. It is caused by a virus. There are two types of herpes Simplex: HSV type 1 and HSV type 2. HSV1 affects the upper section of human body namely brain, mouth, and eyes. On the other hand, HSV2 affects the lower section of a human body namely the genital area. Acyclovir is the drug that is commonly used in treating herpes simplex. It does this through two mechanism. The first mechanism is by monophosphorylation. Monophosphorylation is a mechanism by which Acyclovir interacts with the enzyme as it acts as a competitive inhibitor to HSV thymidine kinase. In the second mechanisms, Acyclovir acts as a substrate to DNA polymerase by triphosphorylation (Gaudio, Richards, and Takahata, 2000). In this workshop, various meta-substituted phenylguanine analogues were tested at the University of Oxford for inhibitory activity against HSV 2 and HSV 1 thymidine kinase. The founded structure of HSV 1 and HSV 2 analogues was based on Figure 1. The chemical groups that was attached to these analogues at the analogues location X are the key causes of their differences. The key objective of this workshop based on 2D-QSAR is Identifying the molecular properties of chemical groups at a particular position X, which are essential for the determination of the inhibitory activity of analogues. It is also meant to develop the HSV thymidine kinase inhibitors through the quantitative models. Figure 01: The structure of the phenylguanine analogues considered in this study. Functionality in each of the analogues differs at position X. Methods: The sixteen phenylguanine analogues inhibitory activity alongside QSAR descriptors were placed in the Microsoft Excel file spreadsheet. Table 02 illustrates these descriptors. Table 02: The QSAR descriptors that was employed in the study LIPOPHILICITY Descriptors ELECTROSTATIC Descriptors STERIC Descriptors   Molar Refractivity (MR) F Molecular Weight (MW) R STERIMOL B Molar Refractivity (MR) STERIMOL L Molecular Volume (Å3) This practical session consisted of four sections. The first part was 1/ 2D-QSAR descriptor selection. This part involved downloading data analysis from Excel options. Then add-Ins was the next procedure and finally the Analysis ToolPak. To open HSV2 raw data, the research opened the file hsv.xlsx and selected HSV2 raw. The next steps were choosing Data >> Data Analysis >> Correlation. On the input range box the formula $C$1:$L$17 was added. The final step was clicking on column and Labels within the first row. The second part involved 2/ Model Formulation. Within the raw for HSV2, two new columns between C and D were added and named “ScPi and ScMR”. The formulae = (F2-AVERAGE ($F$2:$F$17))/STDEV($F$2:$F$17) was entered in cell D2 followed by copying it to the whole column. Similar procedure was repeated cell G2 though with some modifications being made on the formulae to make it compatible with the column. Following this, four unrelated chemical phenylguanine analogues chemical were picked and cut and pasted in raw 20-23. Regression equation was calculated using the procedure Data >> Data Analysis >> Regression. The input range was X and Y ($D$1:$E$13 and $C$1:$C$13). Then “Labels” and “Residuals” were clicked simultaneously. The formula below provided for the new QSAR equation. log (1/IC50) = (Coefficient 1 x QSAR Parameter 1) + (Coefficient 2 x QSAR Parameter 2) + Intercept. The third part involved 3/ Predictive Power. The estimation of the four chemicals was performed based on the above equation. A new column was inserted between C and D. The formula was entered basing on the regression equation. This was done with a view to predict the inhibitory activity of the compound. The last part involved 4/ Rational Design. This step was conducted to help understand the analogues that was supposed to be synthesized basing on the activity. In OtherSubs worksheet, all columns were deleted with the exception of Substituent, and Formula alongside the two selected QSAR parameters, which of important. The spreadsheet was then formatted. The E1 cell was named with ScPi with using the formulae =(C2-AVERAGE($C$2:$C$24))/STDEV($C$2:$C$24). It was then pasted in E2 for the column E. similarly this was done for column F with ScMR in F1 using =(D2-AVERAGE($D$2:$D$24))/STDEV($D$2:$D$24) as formulae in F2 for that column. Appropriate formula was entered in column G. This formula was important in predicting the inhibitory activities of all the un-synthesized analogues in the “OtherSubs” worksheet by using QSAR equation above. Finally, five phenylguanine analogues were recorded from the “OtherSubs” worksheet with the highest predicted activities. Results and Discussions: As shown in spreadsheet 1 in Excel file, Pi and MR are the most important parameters according to log1/IC50 (inhibitory activity) which are 0.79157858 and 0.7942243 respectively close to number 1. Thus they have been chosen to work in. Based on this formulae “log (1/IC50) = (Coefficient 1 x QSAR Parameter 1) + (Coefficient 2 x QSAR Parameter 2) + Intercept” the new QSAR equation according to pi and MR parameters is Log (1/IC50) = (0.137*Pi) + (0.37*MR) + 5.7 As shown in table 2, R square is 0.65 which indicates the equation is satisfaction but it could be better as the range of R square is 0.6 – 0.95 (Gaudio, Richards, and Takahata, 2000). Table 2: Regression Statistics Regression Statistics Multiple R 0.811941 R Square 0.659248 Adjusted R Square 0.583525 Standard Error 0.393475 Observations 12 Hansch equation exemplifies that molar refractory (MR) is the most significant propriety as it has the highest number by mean of Bigger numbers bigger effect. In addition, lipophilicity (Pi) or hydrophobicity is required because the drug should enter the phosopholipid bilayer to reach the target effectively. Thus the degree of lipophilicity is required. The table below represents the actual activity and predictive activity of the four chosen chemical phenylguanine analogues. As it is displayed the predictive activity of Phenyl and Hydroxymethyl group are higher than the actual activity of them. In contrast, the actual activity of Thiol and Trifluoromethyl are greater than the predictive activity of the same compound. This result concludes that the model is effective for some compounds such as Phenyl and Hydroxymethyl at its predicted inhibitory activity. While table 4 illustrates that there were unacceptable values in measuring the predictive activity according to the residuals values. It means the closer residual value to zero the more acceptable one. Based on this theory, the last two compounds are rejected because they have high residuals numbers. Table 3: actual and predictive activity of the four analogues Compound formula Actual activity Predictive activity HYDROXYMETHYL -CH2OH 5.05 6.10 THIOL -SH 6.05 5.51 TRIFLUOROMETHYL -CF3 6.30 6.02 PHENYL -C6H5 6.64 7.02 Table 4: residual output RESIDUAL OUTPUT Observation Predicted HSV2 Log (1 / IC50) Residuals 1 5.046413 -0.25641 2 4.903133 -0.08313 3 5.216242 0.053758 4 5.245542 0.134458 5 5.643856 -0.19386 6 5.973215 -0.45322 7 5.759681 -0.14968 8 6.460245 -0.39024 9 6.192907 0.017093 10 6.12554 0.12446 11 6.00564 0.32436 12 5.827587 0.872413 In spite of predictive values, other parameters should be taken in account in determining drug’s therapeutic value such as bioavailability to know the extent of it that reaches the circulating system, and solubility of it as well as the volume of distribution. In fact stability of the compound is one of the important parameters as the more stable compound is the better one. Moreover, lipophilicity of the analogue should be considered. In searching for other phenylguanine analogues five compounds have been found with high predictive activity. Also their lipophilicity is high as well as molar refractory which both are important for increasing activity and therapeutic properties table 5. Table 5: predictive activity of five other substances Substituent Formula Pi MR SCPi ScMR Predictive activity PHENOXY -OC6H5 2.08 27.68 1.55 0.72 6.18 PHENYLAZO -N2C6H5 1.69 31.31 1.17 1.01 6.23 PHENYLSULPHONYL -SO2C6H5 0.27 33.20 -0.23 1.16 6.10 PHENYLSULPHONYLOXY -SO3C6H5 0.93 36.70 0.42 1.44 6.29 BUTYLAMINO -NHCH2CH2CH2CH3 1.16 24.26 0.65 0.45 5.95 Conclusion: To sum up, selection the best descriptor is the main purpose of this workshop. Despite of the benefits of having predictive activity, lipophilicity, and molar refractory of the module, it has some limitations. The model can predict only mathematical calculations for an ideal process, while in actual practice numbers of factors are involved. Thus, this model is not effective in biological test. References: 1. Drews J., (2000). Drug discovery: A historical perspective. Drug Discovery 287, 1960-1964. 2. Gaudio A., Richards W., and Takahata Y., (2000). QSAR and molecular graphics analysis of N2-phenylguanines as inhibitors of herpes simplex virus thymidine kinases. Journal of molecular graphics and modelling 18, 33-41. 3. Lill M., (2007). Multi-dimensional QSAR in drug discovery. Drug discovery today.12 (23/24), 1013-1017. 4. Johnson S., (2008). The trouble with QSAR (or how I learned so stop worrying and embrace fallacy). American chemical society 48, 25-26. Read More