Chem Biol Drug Des 2010; 76: 340344

Research Article

2010 John Wiley & Sons A/S doi: 10.1111/j.1747-0285.2010.01019.x

Binding of Cordycepin Monophosphate to AMP-Activated Protein Kinase and its Effect on AMP-Activated Protein Kinase Activation
Zhanli Wang1, Xing Wang2, Kai Qu2, Ping Zhu2, Na Guo2, Ruiping Zhang2, Zeper Abliz2, Hui Yu3 and Haibo Zhu2*
1

College of Pharmaceutical Science, Zhejiang University of Technology, Hangzhou 310014, China 2 Key Laboratory of Bioactive Substances and Resources Utilization of Chinese Herbal Medicine, Ministry of Education & Key Laboratory of Natural Drugs Biosynthesis, Ministry of Health, Institute of Materia Medica, Chinese Academy of Medical Sciences & Peking Union Medical College, Beijing 100050, China 3 Department of Laboratory Medicine, the Affiliated Tenth People's Hospital, Tongji University, Shanghai 200072, China *Corresponding author: Haibo Zhu, zhuhaibo@imm.ac.cn These two authors contributed equally to this work.
It had been reported that cordycepin could activate AMP-activated protein kinase. One possible mechanism is that cordycepin mediated AMP-activated protein kinase activation by conversion into cordycepin monophosphate, which acts as an AMP analog to activate AMP-activated protein kinase. To conrm the aforementioned hypothesis, we investigate the binding of cordycepin monophosphate to AMP-activated protein kinase using molecular docking. The modeling results indicate that cordycepin monophosphate binds to AMPactivated protein kinase with high afnity. The hydrogen bonds provide attractive forces between molecules. Our results further identify the key residues contributing to the interaction. Also, the modeling results predict that cordycepin monophosphate and AMP would have similar binding modes with AMP-activated protein kinase. Further investigation of AMP-activated protein kinase activation in vitro provides the evidence that cordycepin monophosphate functioned as an AMP mimic to activate AMP-activated protein kinase. Key words: adenosine monophosphate (AMP), AMP-activated protein kinase, cordycepin, cordycepin monophosphate, molecular docking Received 24 November 2009, revised 25 June 2010 and accepted for publication 3 July 2010

vital metabolic pathways in the cell (1). Changes in AMPK activity have been shown to regulate energy consuming biosynthetic pathways, such as fatty acid and sterol synthesis, and adenosine triphosphate (ATP)-producing catabolic pathways, such as fatty acid oxidation. AMPK has therefore been proposed as a major therapeutic target for obesity and obesity-linked metabolic disorders such as hyperlipidemia. AMPK is a heterotrimeric complex consisting of a catalytic subunit (a) and two regulatory subunits (b and c). The c subunits of AMPK have been shown to contain AMP-binding sites (2). Regulation of AMPK activity is typically affected through the binding of AMP to allosteric sites. As AMPK represent potential drug targets for metabolic diseases, efforts were initiated to discover AMP mimics that bind to AMPbinding sites with high affinity and high enzyme specificity (3). Numerous investigators have focused on the nucleoside analogs that generate nucleoside monophosphates (NMPs) inside cells (4,5). For example, 5-aminoimidazole-4- carboxamide 1-bD-ribofuranoside (AICAR), a nucleoside analog discovered in the late 1980s, was one of the AMPK stimulators (6). AICAR mediated AMPK activation by conversion into ZMP (AICAR monophosphate), which acts as an AMP analog to activate AMPK without affecting the cellular AMP:ATP ratio. Cordycepin, a nucleoside analog 3-deoxyadenosine, is a bioactive compound present in traditional Chinese medicine Cordyceps. In recent years, many pharmacological properties of cordycepin have been reported, including antiviral (7), antifungal (8) and antitumor activity (9). Recently, cordycepin was shown to function as an activator of the AMPK pathway (10). One possible mechanism is that cordycepin mediated AMPK activation by conversion into cordycepin monophosphate, which acts as an AMP analog to activate AMPK. Moreover, a structural similarity has been noted among AMP, ZMP and cordycepin monophosphate (Chart 1). Therefore, we hypothesize that the aforementioned compounds and AMPK share common binding strategy and cordycepin monophosphate may interact with AMPK. According to our best knowledge, there are no studies evaluating the interaction between cordycepin monophosphate and AMPK. In this investigation, a docking simulation of the AMPK complexed with cordycepin monophosphate was performed. We examined whether cordycepin monophosphate and AMP have similar binding modes with AMPK. We also determined the AMPK activation by cordycepin monophosphate. Our results suggest that AMPK emerged as a possible target of cordycepin monophosphate.

The AMP-activated protein kinase (AMPK) has been proposed to act as a sensor of cellular energy status capable of regulating 340

Binding and Activating of Cordycepin Monophosphate to AMPK

A B C

Chart 1: The structures of (A) AMP, (B) ZMP and (C) Cordycepin monophosphate.

Materials and Methods

Agents Cordycepin used in this study was kindly provided by Professor Ping Zhu (Chinese Academy of Medical Sciences, Beijing, China). SAMS peptide was purchased from Millipore. ATP and AMP were from Sigma. Whatman p81 filter paper was obtained from Yingjun Technology (Beijing, China).

site. The docking result demonstrated that the predicted binding mode was in excellent agreement with the experimentally observed structure, with small Root-mean-square Deviation (RMSD) value between the heavy atoms (0.41 ) (data not shown).

Docking studies The binding conformations of cordycepin monophosphate bound to AMPK were modeled using AutoDock 3.05 (11) based on the published crystal structure of AMPK (PDB code: 2OOX) (12). For ligand, GasteigerMarsili partial charges were assigned, as implemented in AutoDockTools. We retained the 240th, 307th and 364th water molecules at the receptor-binding site for docking studies, as it was found to stabilize the interaction of the ligand molecule. Fifty independent docking runs were performed in ligand, and the resulting conformations clustered using a root mean-squared deviation criterion of 0.5 in x, y, z positional coordinates. The compound cordycepin was also docked to the AMPK active site using docking protocol as described earlier. The docked molecules were further scored with eHiTS_Score program, which contained the statistically derived empirical scoring function to capture the interactions between receptors and ligands (13).

Interactions between AMPK and cordycepin monophosphate To determine the interactions between AMPK and cordycepin monophosphate, a possible docking model of AMPK and cordycepin monophosphate is generated by the AutoDock 3.05 (Figure 1). In this model, we can see that cordycepin monophosphate located in the center of the binding pocket. We observed that hydrophobic interaction and hydrogen bond were the stabilizing force in this proteinligand interaction. The theoretical calculation indicated that compound formed hydrogen bonds with Arg139, Arg141, Ala196, Ser217, Ala218, Arg290, Asp308, the 240th water molecule and the 307th water molecule, respectively (Figure 2). Moreover, the binding mode of cordycepin monophosphateAMPK is consistent with that found in the crystal structure of AMPK complexed with AMP (Figure 3). A comparison and superposition of the docked structure of cordycepin with the crystal structure of AMP revealed a RMSD value of 0.99 (Figure 4).

Activation of AMPK protein Analyses of the activities of AMPK were carried out in typical assay conditions of a 30-ll reaction mixture containing 1 mM dithiothreitol (DTT), 50 mM MgCl2, 0.4 mM ATP (0.4 lCi of [c32P] ATP per reaction) and 90 lM SAMS peptide. The reaction was initiated by the addition of AMPK proteins treated with cordycepin or cordycepin monophosphate, incubated at 30 C for 20 min and terminated by the addition of 50 lL of 1% H3PO4. Particulate matter was then transferred to p81 filter paper and washed three times with 0.05% H3PO4. Radioactivity that had been incorporated in the SAMS peptide was determined by liquid scintillation counting in a Wallac Microbeta plate counter.

Results
Assessment of the Autodock ability to properly dock AMP into AMPK To validate the docking method, we first applied our technique to the rebuilding of the crystallographic AMPAMPK complex by docking the ligand extracted from the structures back into the binding Chem Biol Drug Des 2010; 76: 340344 Figure 1: The predicted three-dimensional structure of the AMPK complexed with cordycepin monophosphate. Different parts of AMPK are presented in different colors: the a subunit in blue, the b subunit in green and the c subunits in red. Compound cordycepin monophosphate is depicted in CPK representation. 341

Wang et al.

Figure 2: Key binding interactions of cordycepin monophosphate and AMPK. Hydrogen bonds are presented as dotted lines.

Figure 3: Key binding interactions of AMP and AMPK (PDB code: 2OOX). Hydrogen bonds are presented as dotted lines. Cordycepin was also docked into the binding pockets of AMPK (Figure S1). Furthermore, the eHiTS_Score program was used to evaluate relative affinities between AMPK and ligands. Lower docking score value indicates a higher binding affinity of a ligand. The docking score values obtained for AMP, cordycepin monophosphate and cordycepin were )2.153, )1.822 and )1.046, respectively, indicating that cordycepin monophosphate has relative higher binding affinity compared with cordycepin. monophosphate on AMPK. We found that cordycepin monophosphate can directly activate AMPK in vitro, and cordycepin monophosphate stimulates AMPK in a dose-dependent manner (Figure 5). Moreover, results indicated that cordycepin had no effects on AMPK activation (data not shown).

Discussion
AMPK is a direct target of AMP and has emerged as a potentially key signaling intermediary in the regulation of changes in glucose and lipid metabolism (1416). ZMP acts as an AMP analog to activate AMPK. Cordycepin monophosphate also showed structural similarity with AMP. Therefore, we hypothesize that cordycepin Chem Biol Drug Des 2010; 76: 340344

The effects of cordycepin monophosphate on AMPK activation To determine whether cordycepin monophosphate stimulates AMPK, we investigated the effects of variety concentrations of cordycepin 342

Binding and Activating of Cordycepin Monophosphate to AMPK

Figure 4: Comparison of docked structures of cordycepin monophosphate (green) and AMP (yellow).

phosphate readily binds to AMPK with binding affinity very similar to that found with AMP. Moreover, the docking score values obtained for AMP and cordycepin monophosphate were )2.153 and )1.822, respectively, confirming that AMP and cordycepin monophosphate have similarly high binding affinity. In contrast, the value obtained for cordycepin was )1.046, indicating a decreased binding affinity compared with AMP and cordycepin monophosphate. It was well known that the loss of phosphate group might result in weaker binding energy that contributes to decreased affinity of ligand to AMPK (17). Compared with cordycepin monophosphate, cordycepin could not form hydrogen bonds with the positively charged phosphate-binding site of AMPK, and the interactions of cordycepin with AMPK were apparently weaker than that observed with cordycepin monophosphate. Therefore, we hypothesize that the loss of seven hydrogen bonds has significant effects on the binding potency of cordycepin. For the purpose of obtaining additional evidences that AMPK is activated by cordycepin monophosphate, the phosphorylation of SAMS peptide was used as indicator of AMPK activation. Analyses of the activities of AMPK provided the first experimental data supporting the concept and the preference for cordycepin monophosphate. Consistent with the molecular docking data, the result showed that cordycepin monophosphate was a potential activator of AMPK. The data support the concept that cordycepin mediated AMPK activation by conversion into cordycepin monophosphate.

Conclusions
In summary, evidence that cordycepin monophosphate not only bound to AMPK but also functioned as an AMP mimic was apparent from both docking results as well as analyses of the AMPK activation. The results reported give support to the idea that cordycepin mediated AMPK activation by conversion into cordycepin monophosphate.

Figure 5: Activation of AMPK by cordycepin monophosphate.

monophosphate may interact with AMPK. To test this assumption, we simulated the 3D structure of AMPK complexed with cordycepin monophosphate by molecular docking strategy. We obtained docking model of the complex between AMPK and cordycepin monophosphate. Docking result means that hydrogen bonds play an important role in the interactions. This binding mode is agreement with that of AMP, because local environment within AMPK-binding site is highly dependent on hydrogen bond interactions (17). Compared with AMP, the purine base moiety retains the three hydrogen bonds formed between the 6NH2 and 1N groups of cordycepin monophosphate and residues Ala196 and Ala218. Moreover, the phosphonic acid group retains the seven hydrogen bonds formed between the cordycepin monophosphate and residues in the positively charged phosphate-binding site. Ribose moiety retains the one hydrogen bond formed between the 2 hydroxyl group of cordycepin monophosphate and the 307th water molecule. Unlike the 3 hydroxyl group of AMP, ribose moiety of cordycepin monophosphate did not form hydrogen bond with the 364th water molecule. Previous studies showed that the hydrogen bonds between the 3 hydroxyl of AMP and residues contributed little to the overall binding affinity (18,19). The results suggested that the cordycepin monoChem Biol Drug Des 2010; 76: 340344

Acknowledgments
This study was supported by grants from National Natural Sciences Foundation of China (NSFC), Grant Number (30873063; 30873455; 30973527), the National 973 Fundamental Project of China, Grant Number (2009CB523004), Shanghai's Health Bureau Science Foundation for Youths, Grant Number (2008Y020), Natural Sciences Foundation of Beijing, Grant Number (7092068) and the Key Project of Youth Foundation of Institute of Materia Medica, Chinese Academy of Medical Sciences and Peking Union Medical College, Grant Number (2006QZH01). We are indebted to the grant from National S&T Major Project (2009ZX09303-003; 2009ZX09103-034).

References
1. Kahn B.B., Alquier T., Carling D., Hardie D.G. (2005) AMP-activated protein kinase: ancient energy gauge provides clues to modern understanding of metabolism. Cell Metab;1:1525.

343

Wang et al.

2. Xiao B., Heath R., Saiu P., Leiper F.C., Leone P., Jing C., Walker P.A., Haire L., Eccleston J.F., Davis C.T. (2007) Structural basis for AMP binding to mammalian AMP-activated protein kinase. Nature;449:496500. 3. Hardie D.G., Hawley S.A. (2001) AMP-activated protein kinase: the energy charge hypothesis revisited. Bioessays;23:11121119. 4. Cool B., Zinker B., Chiou W., Kifle L., Cao N., Perham M., Dickinson R. et al. (2006) Identification and characterization of a small molecule AMPK activator that treats key components of type 2 diabetes and the metabolic syndrome. Cell Metab;3:403416. 5. Yamanaka G., Wilson T., Innaimo S., Bisacchi G.S., Egli P., Rinehart J.K., Zahler R., Colonno R.J. (1999) Metabolic studies on BMS-200475, a new antiviral compound active against hepatitis B virus. Antimicrob Agents Chemother;43:190193. 6. Gruber H.E. US Patent 5,658,889, August 19, 1997. 7. Mueller W.E.G., Weiler B.E., Charubala R., Pfleiderer W., Leserman L., Sobol R.W., Suhadolnik R.J., Schroeder H.C. (1991) Cordycepin analogues of oligoadenylate inhibit human immunodeficiency virus infection via inhibition of reverse transcriptase. Biochem Wash;30:20272033. 8. Sugar A.M., McCaffrey R.P. (1998) Antifungal activity of 3-deoxyadenosine (cordycepin). Antimicrob Agents Chemother; 42:14241427. 9. Okoshi R., Ozaki T., Yamamoto H., Ando K., Koida N., Ono S., Koda T., Kamijo T., Nakagawara A., Kizaki H. (2008) Activation of AMP-activated protein kinase induces p53-dependent apoptotic cell death in response to energetic stress. J Biol Chem;283:39793987. 10. Wong Y.Y., Moon A., Duffin R., Barthet-Barateig A., Meijer H.A., Clemens M.J., de Moor C.H. (2010) Cordycepin inhibits protein synthesis and cell adhesion through effects on signal transduction. J Biol Chem;285:26102621. 11. Morris G.M., Goodshell D.S., Halliday R.S., Huey R., Hart W.E., Belew R.K., Olson A.J. (1998) Automated docking using a Lamarckian genetic algorithm and empirical binding free energy function. J Comput Chem;19:16391662. 12. Townley R., Shapiro L. (2007) Crystal structures of the adenylate sensor from fission yeast AMP-activated protein kinase. Science;315:17261729. 13. Zsoldos Z., Reid D., Simon A., Sadjad B.S., Johnson A.P. (2006) eHiTS: an innovative approach to the docking and scoring function problems. Curr Protein Pept Sci;7:421435.

14. Ko H.J., Zhang Z., Jung D.Y., Jun J.Y., Ma Z., Jones K.E., Chan S.Y., Kim J.K. (2009) Nutrient stress activates inflammation and reduces glucose metabolism by suppressing AMP-activated protein kinase in the heart. Diabetes;58:25362546. 15. Gruzman A., Babai G., Sasson S. (2009) Adenosine monophosphate-activated protein kinase (AMPK) as a new target for antidiabetic drugs: a review on metabolic, pharmacological and chemical considerations. Rev Diabet Stud;6:1336. 16. Guo D., Hildebrandt I.J., Prins R.M., Soto H., Mazzotta M.M., Dang J., Czernin J., Shyy J.Y., Watson A.D., Phelps M., Radu C.G., Cloughesy T.F., Mischel P.S. (2009) The AMPK agonist AICAR inhibits the growth of EGFRvIII-expressing glioblastomas by inhibiting lipogenesis. Proc Natl Acad Sci U S A;106:12932 12937. 17. Erion M.D., Dang Q., Reddy M.R., Kasibhatla S.R., Huang J., Lipscomb W.N., van Poelje P.D. (2007) Structure-guided design of AMP mimics that inhibit fructose-1, 6- bisphosphatase with high affinity and specificity. J Am Chem Soc;129:15480 15490. 18. Reddy M.R., Erion M.D. (2001) Calculation of relative binding free energy differences for fructose 1, 6-bisphosphatase inhibitors using the thermodynamic cycle perturbation approach. J Am Chem Soc;123:62466252. 19. Reddy M.R., Erion M.D. (2007) Relative binding affinities of fructose-1,6- bisphosphatase inhibitors calculated using a quantum mechanics-based free energy perturbation method. J Am Chem Soc;129:92969297.

Supporting Information
Additional Supporting Information may be found in the online version of this article:

Figure S1. Key binding interactions of cordycepin and AMPK. Hydrogen bonds are presented as dotted lines.

Please note: Wiley-Blackwell is not responsible for the content or functionality of any supporting materials supplied by the authors. Any queries (other than missing material) should be directed to the corresponding author for the article.

344

Chem Biol Drug Des 2010; 76: 340344