Glycogen synthase kinase-3β（GSK-3β）為一種蛋白質激酶，當GSK-3β活性過高，過度磷酸化下游蛋白分子，而導致細胞訊息傳遞異常，會引發疾病，如癌症、發炎反應以及阿茲海默症等。本研究利用分子嵌合計算的方式，並加上激酶抑制效果實驗尋找GSK-3β抑制劑。首先，從蛋白質資料庫中挑選16個胺基酸序列100%相同的複合體結晶結構，並且研究其再現性。再來選擇10個再現性較佳的結構進行交叉分子嵌合，藉此研究不同結晶構型時，能否正確的再現出蛋白質與小分子之間的結合模式。接著選擇交叉分子嵌合中，表現最佳的構型作為單一嵌合的蛋白質結構；交叉分子嵌合中，表現前五名的構型作為多重嵌合的蛋白質結構；並利用Q-SiteFinder計算活性中心體積精確度最高的前五名作為多重嵌合的蛋白質結構，同時不同測試評分函數及不同氫鍵限制位置對計算結果的影響。我們從Binding Database中208個已知Ki的GSK-3β抑制劑，利用Discovery Studio挑選出21個結構不相似的GSK-3β活性分子，再利用ZINC網站產生其對應的decoys，共有759個decoys，建立一套GSK-3β專用的標準測試資料庫，篩選的效率以Enrichment Factor（EF）為主要指標，測試不同評分函數的表現及有無氫鍵限制對篩選效率的影響。由計算的結果得到，最好的氫鍵限制及評分函數提升Enrichment Factor至26倍。選用Kinase_ChemScore評分函數及在ASP133骨幹上的O原子與VAL135骨幹上的H原子設定氫鍵的組合為最佳設定。因此我們利用這個設定進行高速虛擬篩選，並在ZINC及烏克蘭資料庫中進行篩選，從中找出較有潛力的小分子化合物，並且將這些化合物進行激酶抑制效果實驗，確認小分子化合物是否有抑制效果。在測試的10個化合物中，有5個化合物有抑制效果，並且將有抑制效果的分子再進一步給生物學家進行後續的生物實驗。

Glycogen synthase kinase-3β（GSK-3β）is an important protein target for treatment of several diseases. When the activity of GSK-3β is too high, it over-phosphorylates downstream proteins and causes some diseases such as cancer, inflammation, and Alzheimer’s Diseases. This study utilized docking computation and kinase assay to search new GSK-3β kinase inhibitors. First, we selected 16 crystal structures of GSK-3β kinase-ligand complexes which have 100% amino acid sequence homology in the Protein Data Bank to examine their reproducibility. Next, we chose 10 complexes which have good reproducibility to do cross docking, and investigate how ligand conformation can be regained when protein structures from different complexes were used. According to the result of cross docking, the best protein structure was used for later docking with single structure. The top five protein structures were used to do ensemble docking. In addition, the volume of binding sites were computed using Q-SiteFinder, and the top five protein structures which present high precision based on a volume computation criterion were chosen to do ensemble docking. Furthermore, we investigated how scoring functions and different H-bond constraints performed in our docking computation. For this purpose, we built a DUD-E benchmark set consisting of active and inactive compounds. For the former, Discovery Studio was used to selected 21 dissimilar compounds from the existing 208 active compounds with Ki values available in the Binding Database. Inactive decoy compounds of those 21 active compounds were generated using a ZINC program, which gave 759 decoy compounds. This compound set was used for benchmark based on the enrichment factor（EF）value for each docking protocol. The computed results showed that the best protocol is to use the scoring function: Kinase_ChemScore combined with the H-bond constraints on O atom of ASP133 and H atom of VAL135 on the backbone, which enhanced the EF up to 26. Moreover, we use this protocol to do high throughput virtual screening in the ZINC Database and the Enamine Database to find out some potential small molecular compounds, and suggested these compounds for kinase assay. Among the 10 compound undergone kinase assay, 5 compounds were identified to be active, and were recommended for further cell and animal assays.

1. Cohen, P., Protein Kinases—the Major Drug Targets of the Twenty-First Century? Nature Reviews Drug Discovery, 2002. 1: p. 309-315.
2. Xu, C., Kim, N.-G. and Gumbiner, B.M., Regulation of Protein Stability by Gsk3 Mediated Phosphorylation. Cell Cycle, 2009. 8: p. 4032-4039.
3. Taylor, R.D., Jewsbury, P.J. and Essex, J.W., A Review of Protein-Small Molecule Docking Methods. Journal of Computer-Aided Molecular Design, 2002. 16: p. 151-166.
4. Fang, J., Huang, D., Zhao, W., Ge, H., Luo, H.-B. and Xu, J., A New Protocol for Predicting Novel Gsk-3β Atp Competitive Inhibitors. Journal of Chemical Information and Modeling, 2011. 51: p. 1431-1438.
5. Frame, S. and Cohen, P., Gsk3 Takes Centre Stage More Than 20 Years after Its Discovery. Biochemical Journal, 2001. 359: p. 1-16.
6. Dokken, B.B., Sloniger, J.A. and Henriksen, E.J., Acute Selective Glycogen Synthase Kinase-3 Inhibition Enhances Insulin Signaling in Prediabetic Insulin-Resistant Rat Skeletal Muscle. American Journal of Physiology-Endocrinology and Metabolism, 2005. 288: p. E1188-E1194.
7. Martinez, A., Castro, A., Dorronsoro, I. and Alonso, M., Glycogen Synthase Kinase 3 (Gsk‐3) Inhibitors as New Promising Drugs for Diabetes, Neurodegeneration, Cancer, and Inflammation. Medicinal research reviews, 2002. 22: p. 373-384.
8. Ougolkov, A.V. and Billadeau, D.D., Targeting Gsk-3: A Promising Approach for Cancer Therapy? Future Oncology, 2006. 2: p. 91-100.
9. Lovestone, S., Reynolds, C.H., Latimer, D., Davis, D.R., Anderton, B.H., Gallo, J.-M., Hanger, D., Mulot, S., Marquardt, B. and Stabel, S., Alzheimer's Disease-Like Phosphorylation of the Microtubule-Associated Protein Tau by Glycogen Synthase Kinase-3 in Transfected Mammalian Cells. Current Biology, 1994. 4: p. 1077-1086.
10. Roher, A.E., Lowenson, J.D., Clarke, S., Woods, A.S., Cotter, R.J., Gowing, E. and Ball, M.J., Beta-Amyloid-(1-42) Is a Major Component of Cerebrovascular Amyloid Deposits: Implications for the Pathology of Alzheimer Disease. Proceedings of the National Academy of Sciences, 1993. 90: p. 10836-10840.
11. Perry, G., Friedman, R., Shaw, G. and Chau, V., Ubiquitin Is Detected in Neurofibrillary Tangles and Senile Plaque Neurites of Alzheimer Disease Brains. Proceedings of the National Academy of Sciences, 1987. 84: p. 3033-3036.
12. Wagner, U., Utton, M., Gallo, J.-M. and Miller, C., Cellular Phosphorylation of Tau by Gsk-3 Beta Influences Tau Binding to Microtubules and Microtubule Organisation. Journal of Cell Science, 1996. 109: p. 1537-1543.
13. Morris, G.M., Goodsell, D.S., Halliday, R.S., Huey, R., Hart, W.E., Belew, R.K. and Olson, A.J., Automated Docking Using a Lamarckian Genetic Algorithm and an Empirical Binding Free Energy Function. Journal of Computational Chemistry, 1998. 19: p. 1639-1662.
14. Ewing, T.J., Makino, S., Skillman, A.G. and Kuntz, I.D., Dock 4.0: Search Strategies for Automated Molecular Docking of Flexible Molecule Databases. Journal of Computer-Aided Molecular Design, 2001. 15: p. 411-428.
15. Friesner, R.A., Banks, J.L., Murphy, R.B., Thomas A. Halgren, Klicic|, J.J., Mainz, D.T., Repasky, M.P., Knoll, E.H., Mee Shelley, J.K.P., Shaw, D.E., Francis, P. and Shenkin, P.S., Glide: A New Approach for Rapid, Accurate Docking and Scoring. 1. Method and Assessment of Docking Accuracy. J. Med. Chem., 2004. 47: p. 1739-1749.
16. Verdonk, M.L., Cole, J.C., Hartshorn, M.J., Murray, C.W. and Taylor, R.D., Improved Protein–Ligand Docking Using Gold. Proteins: Structure, Function, and Bioinformatics, 2003. 52: p. 609-623.
17. Huang, N., Shoichet, B.K. and Irwin, J.J., Benchmarking Sets for Molecular Docking. Journal of Medicinal Chemistry, 2006. 49: p. 6789-6801.
18. Mysinger, M.M., Carchia, M., Irwin, J.J. and Shoichet, B.K., Directory of Useful Decoys, Enhanced (Dud-E): Better Ligands and Decoys for Better Benchmarking. Journal of Medicinal Chemistry, 2012. 55: p. 6582-6594.
19. Rikmenspoel, R. and Caputo, R., The Michaelis-Menten Constant for Fructose and for Glucose of Hexokinase in Bull Spermatozoa. J Reprod Fertil, 1966. 12: p. 437-444.
20. Cheng, H.C., The Power Issue: Determination of Kb or Ki from Ic50. A Closer Look at the Cheng-Prusoff Equation, the Schild Plot and Related Power Equations. Journal of pharmacological and toxicological methods, 2000. 46: p. 61-71.
21. Yung-Chi, C. and Prusoff, W.H., Relationship between the Inhibition Constant (Ki) and the Concentration of Inhibitor Which Causes 50 Per Cent Inhibition (I50) of an Enzymatic Reaction. Biochemical Pharmacology, 1973. 22: p. 3099-3108.
22. Burlingham, B.T. and Widlanski, T.S., An Intuitive Look at the Relationship of Ki and Ic50: A More General Use for the Dixon Plot. Journal of Chemical Education, 2003. 80: p. 214.
23. Klebe, G., Virtual Ligand Screening: Strategies, Perspectives and Limitations. Drug Discovery Today, 2006. 11: p. 580-594.
24. Jones, G., Willett, P. and Glen, R.C., Molecular Recognition of Receptor Sites Using a Genetic Algorithm with a Description of Desolvation. Journal of molecular biology, 1995. 245: p. 43-53.
25. Kokh, D.B. and Wenzel, W., Flexible Side Chain Models Improve Enrichment Rates in in Silico Screening. Journal of Medicinal Chemistry, 2008. 51: p. 5919-5931.
26. Perola, E., Minimizing False Positives in Kinase Virtual Screens. Proteins: Structure, Function, and Bioinformatics, 2006. 64: p. 422-435.
27. Cheng, T., Li, X., Li, Y., Liu, Z. and Wang, R., Comparative Assessment of Scoring Functions on a Diverse Test Set. Journal of Chemical Information and Modeling, 2009. 49: p. 1079-1093.
28. Eldridge, M.D., Murray, C.W., Auton, T.R., Paolini, G.V. and Mee, R.P., Empirical Scoring Functions: I. The Development of a Fast Empirical Scoring Function to Estimate the Binding Affinity of Ligands in Receptor Complexes. Journal of Computer-Aided Molecular Design, 1997. 11: p. 425-445.
29. Cross, J.B., Thompson, D.C., Rai, B.K., Baber, J.C., Fan, K.Y., Hu, Y. and Humblet, C., Comparison of Several Molecular Docking Programs: Pose Prediction and Virtual Screening Accuracy. Journal of Chemical Information and Modeling, 2009. 49: p. 1455-1474.
30. Fogarty, J., Baker, R.S. and Hudson, S.E. Case Studies in the Use of Roc Curve Analysis for Sensor-Based Estimates in Human Computer Interaction. in Proceedings of Graphics Interface 2005. 2005. Canadian Human-Computer Communications Society.
31. Lobo, J.M., Jiménez‐Valverde, A. and Real, R., Auc: A Misleading Measure of the Performance of Predictive Distribution Models. Global Ecology and Biogeography, 2008. 17: p. 145-151.
32. Grünberg, R., Leckner, J. and Nilges, M., Complementarity of Structure Ensembles in Protein-Protein Binding. Structure, 2004. 12: p. 2125-2136.
33. Korb, O., McCabe, P. and Cole, J., The Ensemble Performance Index: An Improved Measure for Assessing Ensemble Pose Prediction Performance. Journal of Chemical Information and Modeling, 2011. 51: p. 2915-2919.
34. Korb, O., Olsson, T.S., Bowden, S.J., Hall, R.J., Verdonk, M.L., Liebeschuetz, J.W. and Cole, J.C., Potential and Limitations of Ensemble Docking. Journal of Chemical Information and Modeling, 2012. 52: p. 1262-1274.
35. Laurie, A.T. and Jackson, R.M., Q-Sitefinder: An Energy-Based Method for the Prediction of Protein–Ligand Binding Sites. Bioinformatics, 2005. 21: p. 1908-1916.
36. Ben Nasr, N., Guillemain, H.l.n., Lagarde, N., Zagury, J.-F.o. and Montes, M., Multiple Structures for Virtual Ligand Screening: Defining Binding Site Properties-Based Criteria to Optimize the Selection of the Query. Journal of Chemical Information and Modeling, 2013. 53: p. 293-311.
37. Carpenter, A.E., Jones, T.R., Lamprecht, M.R., Clarke, C., Kang, I.H., Friman, O., Guertin, D.A., Chang, J.H., Lindquist, R.A. and Moffat, J., Cellprofiler: Image Analysis Software for Identifying and Quantifying Cell Phenotypes. Genome Biology, 2006. 7: p. R100.
38. Rodems, S.M., Hamman, B.D., Lin, C., Zhao, J., Shah, S., Heidary, D., Makings, L., Stack, J.H. and Pollok, B.A., A Fret-Based Assay Platform for Ultra-High Density Drug Screening of Protein Kinases and Phosphatases. ASSAY and Drug Development Technologies, 2002. 1: p. 9-19.
39. Rodems, S.M., Hamman, B.D., Christina Lin, J.Z., Sundeep Shah, Heidary, D., Makings, L., Stack, J.H. and Pollok, B.A., A Fret-Based Assay Platform for Ultra-High Density Drug Screening of Protein Kinases and Phosphatases. ASSAY and Drug Development Technologies, 2002. 1: p. 9-19.
40. Lifetechnologies Corporation. Please refer to the homepage at: http://www.lifetechnologies.com/content/dam/LifeTech/migration/files/drug-discovery/pdfs.par.60256.file.dat/20130430%20ssbk%20customer%20protocol%20and%20assay%20conditions.pdf.
41. The ZINC database:http://zinc.docking.org/subsets/all-purchasable.