Journal of Molecular Structure: THEOCHEM 729 (2005) 125–129 www.elsevier.com/locate/theochem
Farnesyltransferase: Theoretical studies on peptide substrate entrance— thiol or thiolate coordination? Se´rgio F. Sousa, Pedro A. Fernandes, Maria Joa˜o Ramos* REQUIMTE, Departamento de Quı´mica, Faculdade de Cieˆncias, Universidade do Porto, Rua do Campo Alegre, 687, 4169-007 Porto, Portugal Received 10 September 2004; accepted 14 October 2004 Available online 14 July 2005
Abstract During the last decade, farnesyltransferase enzyme (FTase) has established itself as a very promising target in anticancer research, with more than 100 patents describing farnesyltransferase inhibitors (FTIs) published since 2000. However, several crucial doubts in its catalytic mechanism still remain. Understanding the farnesylation mechanism of the natural substrates of FTase would allow the development of more specific and active inhibitors of this enzyme. One of the fundamental dilemmas which characterize the FTase mechanism is the coordination of the peptide substrate. It is known that peptide coordination results in the formation of a zinc thiolate; however, the exact protonation state of the peptide cysteine that reacts with zinc–thiol or thiolate-remains unknown. In this study, the potential energy surfaces for thiol and thiolate peptides entrance into the active-site zinc coordination sphere were determined. The results show that the Gibbs activation barrier for thiolate entrance is of around 20 kcal/mol, whereas for thiol entrance is of only 2.5 kcal/mol. Globally, the results demonstrate that the substrate peptide coordinates zinc as a thiol, subsequently losing a proton to give the thiolate bound minimum, and that a direct thiolate attack is not a kinetically competitive alternative. q 2005 Elsevier B.V. All rights reserved. Keywords: FTase; Carboxylate-shift; Zinc enzymes; B3LYP; ONIOM
1. Introduction Farnesyltransferase enzyme (FTase) catalyzes the farnesylation of protein substrates containing a typical-CAAX motif at the carbonyl terminus. In this characteristic motif, the C represents the cysteine residue that is farnesylated, A is an aliphatic amino acid, whereas X stands for the terminal amino acid residue, in general methionine, serine, alanine or glutamine [1]. Among the FTase substrates are the Ras family of proteins, nuclear lamins A and B, the g subunit of heterotrimeric G-proteins, centromeric proteins and several proteins involved in visual signal transduction [2–4]. Farnesyl diphosphate (FPP) is the typical isoprenoid farnesyl donor, and is the first substrate to coordinate the enzyme [5,6]. FTase is a zinc metalloenzyme, containing a single zinc ion per protein dimer [7] essential for its catalytic activity [8]. The enzyme also requires Mg2C for * Corresponding author. E-mail address:
[email protected] (M.J. Ramos).
0166-1280/$ - see front matter q 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.theochem.2005.03.022
maximal levels of activity, although this metal is present only in milimolar concentrations [8]. Interest in FTase inhibition arose primarily from the finding that farnesylation is absolutely required for oncogenic forms of Ras proteins to transform cells [9–11], given that mutant Ras proteins are implicated in about 30% of all human cancers [12–14]. As a consequence, farnesyltransferase inhibitors (FTIs) have emerged as a new class of extremely promising anti-cancer drugs. More than 100 patents describing FTIs have been published since 2000 [15], with some drugs already in advanced stages of clinical testing [15–18]. Presently, the development of FTIs as alternative drugs in the treatment of some diseases caused by pathogens, in particular, malaria [18,19] and African sleeping sickness [20], and as antiviral agents [21] is also underway. In spite of the very promising results obtained over the last few years in the development of FTIs [15–18], several key points in the FTase catalytic mechanism remain unexplained. A complete understanding of the farnesylation mechanism of the natural substrates of this enzyme would allow the rational design of more specific inhibitors, with increased activity and potential value in the treatment of
126
S.F. Sousa et al. / Journal of Molecular Structure: THEOCHEM 729 (2005) 125–129
cancer, malaria, sleeping sickness, or in the fight against infection caused by some viruses. In this article we analyze a crucial aspect in the FTase catalytic mechanism: the coordination of the peptide substrate. Previous studies focusing on the pH relation with peptide binding have demonstrated that peptide coordination results in the formation of a zinc thiolate [22]. However, the exact protonation state of the CAAX cysteine residue that reacts with zinc–thiol or thiolateremains doubtful. Our results successfully rule out a direct thiolate attack, establishing the thiol alternative as the only valid possibility, which implies the existence of a subsequent deprotonation step to generate a zinc thiolate.
2. Computational methods Calculations were performed on 119/120 atoms active ˚ cut around the site thiolate/thiol models, based on an 8 A zinc atom in the 1D8D structure [23], the crystallographic structure of FTase with a peptide substrate and an FPP ˚ ). The models analog with the highest resolution (2.00 A were first freely optimized. Subsequently, scans along the Zn–peptide thiolate/thiol sulphur distances were performed. The minima identifiable in the two potential energy surfaces (PES) were optimized freely. A model considering two different layers was used in all calculations (Fig. 1), following an ONIOM methodology [24,25] and using the program GAUSSIAN03 [26].
The high-level layer included the zinc ion, the acetate group of Asp297b, the methylthiolate group of Cys299b, the imidazole group of His362b, and the methylthiolate/ methylthiol group of the peptide’s substrate cysteine, comprising a total of 23/24 atoms (Fig. 1). This level was treated using density functional theory (DFT) with the B3LYP functional [27,28]. DFT calculations have been shown to give very accurate results for systems involving transition metals [29], particularly when using the B3LYP functional [30–32]. For zinc complexes, the superior accuracy of the B3LYP functional in comparison with Hartree–Fock and second-order Moller–Plesset perturbation theory has also been previously demonstrated [33]. The SDD basis set was used, as implemented in Gaussian 03. This basis set uses the small core quasi-relativistic Stuttgart/ Dresden electron core potentials (also known as StollPreuss, or simply SP) [34,35] for transition elements. For zinc, the outer electrons are described by a (311111/22111/ 411) valence basis specifically optimized for this metal and for the use with the SP pseudopotentials. C, N, and O atoms are accounted by a (6111/41) quality basis set, whereas S and H atoms are treated respectively by a (531111/4211) and a (31) quality basis sets. The high-performance of SP pseudopotentials in calculations involving transition metals compounds, particularly within closed-shell systems, has been previously demonstrated [36]. The outer layer considered four complete residues (Gly298b, Tyr300b, Asp352b, and Tyr361b), in addition to the remaining part of the zinc ligands Asp297b, Cys299b, His362b, and peptide’s cysteine not
Fig. 1. Farnesyltransferase enzyme. Representation of the 1D8D crystallographic structure (ternary complex) [23], with major emphasis being given to the model considered, and its division into two layers.
S.F. Sousa et al. / Journal of Molecular Structure: THEOCHEM 729 (2005) 125–129
included in the high-level layer, resulting in a total of 96 atoms (Fig. 1). In the treatment of the outer layer, the semiempirical method PM3 [37–39] was used. An evaluation of the applicability of PM3, AM1 and MNDO/d to the study of zinc complexes, and organometallic compounds containing Zn2C, clearly confirms PM3 as the most accurate of these three methods [40]. Moreover, it is known that the final energy is not very sensitive to the refinement of a correct geometry [41,42].
3. Results and discussion In spite of the enormous attention that has been dedicated to the study of FTase over the last decade, several crucial aspects in the catalytic mechanism of this enzyme remain unexplained. The coordination of the peptide’s substrate CAAX cysteine to the active-site zinc coordination sphere is still subject to speculation. It has been proved that peptide coordination results in the formation of a zinc thiolate [22]. However, doubts remain on the protonation state of the CAAX cysteine that reacts with zinc–thiol or thiolate. These hypotheses result into alternative mechanisms for peptide entrance. The first hypothesis considers that the substrate enters the enzyme in the thiol form, and coordinates zinc through a weak zinc–sulfur bond. In a subsequent step the zinc-bound thiol loses a proton to a solvent molecule or, most probably, to a base in the active site, generating a tightly bound peptide thiolate. The pKa of an exposed cysteine is 8.3 [43]. At physiological pH, this value implies that 95% of the species are in the thiol form, whereas only 5% are ionized as a thiolate. Coordination to zinc would allow a decrease of around 2 pH units in the pKa, promoting a change from the zinc bound thiol to a thiolate [22]. Experimental studies have demonstrated the existence of an ionization in the ternary complex FTase$FPP$Peptide with a pKa of about 6.4 [22]. While this data is coherent with the existence of a deprotonation of the bound substrate thiol to a thiolate, it is also possible that the ionization is due to another residue of the enzyme. Studies of FTase with a substitution of the active site zinc by cadmium [44] have resulted in a further decrease in the pKa value. The decrease in approximately 1 pH unit is consistent with the expected increase in the binding affinity for the metal, as cadmium is more thiophilic than zinc. These results indicate that the ionization is dependent upon the characteristics of the metal, therefore, suggesting that in FTase the group that suffers the deprotonation interacts with zinc, significantly narrowing the number of possible candidates, and establishing a zinc bound thiol as the most likely alternative. The second mechanistic hypothesis for substrate entrance admits direct thiolate binding to the zinc ion. In solution, at physiological pH, thiol and thiolate exist in equilibrium, and even though the thiol form largely predominates, small percentages of thiolate are also present
127
(around 5%) [45]. SK is a much better nucleophile than SH, and is thus more reactive. Therefore, from a chemical point of view it makes sense that coordination to zinc proceeds through the thiolate form, driving the thiol/thiolate equilibrium in the deprotonation direction. We have analyzed both situations—thiol and thiolate entrance. We aimed to verify if a direct thiolate entrance could be a competitive alternative to an initial thiol coordination. The potential energy surfaces for both processes were compared. The models used were carefully planned in order to capture both the intrinsic characteristics of the zinc complex and the influence of the anisotropic enzyme environment on the zinc coordination sphere. This type of models has been used with success in the mechanistic study of FTase [46] and other enzymes [47–50]. The specific interactions of the enzyme environment with the metal coordination sphere were preserved by explicitly including all residues interacting directly with the zinc complex. From the analysis of both thiol and thiolate entrance into the active-site zinc coordination sphere, two minima connected by a transition state were observed (Fig. 2). In both scans one of the minima corresponds to a stationary state where the substrate peptide, represented by the cysteine residue, is non-zinc coordinated, whereas the other minimum is zinc bonded. Naturally, the Zn–S bond in the metal coordinated minimum is significantly shorter ˚ for thiolate against for the thiolate alternative (2.39 A ˚ for thiol). It is also important to note that in both 2.74 A protonation alternatives coordination to zinc resulted in a bidentante-monodentate change (carboxylate-shift) by zinc ligand Asp297b. The importance of this carboxylate-shift mechanism in FTase was only recently unveiled [46], and the mechanistic implications it may pose are still subject to thoughtful analysis and stringent discussion. Energetically, the observed trend with peptide entrance was very different for the two alternatives considered. The variation in the electronic energy with thiolate approach, represented in Fig. 2, demonstrates the existence of an energy barrier 20.1 kcal/mol high for direct thiolate coordination to the metal atom. For thiol entrance
Fig. 2. Electronic energy variation with peptide entrance for thiolate and thiol hypotheses. Values obtained with B3LYP/SDD//PM3.
128
S.F. Sousa et al. / Journal of Molecular Structure: THEOCHEM 729 (2005) 125–129
Fig. 3. Gibbs energy values (kcal/mol) for all minima and transition-state like structures. Values obtained with B3LYP/SDD//PM3. The Gibbs energies for initial thiol deprotonation in thiolate entrance, and final zinc bound thiol deprotonation in thiol entrance were calculated from experimental values (from pKa values 8.3 and 6.4) [22,43].
the correspondent barrier amounts to only 2.5 kcal/mol. These differences are rather conclusive. Our results show that direct thiolate coordination is not a kinetically competitive alternative for substrate coordination to the active-site zinc in FTase. Fig. 3 illustrates the Gibbs energy values for all minima and transition-state like structures. Previous studies in this type of systems have shown that the effect of zero point corrections, thermal, and entropic effects in the determination of Gibbs reaction and activation energies is generally rather small (less than 2 kcal/mol) [46]. Taking into consideration the large size of the models used, which would render the calculation of these quantities extremely time-consuming, and given the enormous differences in the values obtained for the two mechanistic possibilities analyzed, it was decided not to determine these corrections. This approach does not affect, at any level, the definitiveness of the conclusions derived from this study. Fig. 3 also takes into account the Gibbs energies for deprotonation of the initial cysteine thiol for direct thiolate entrance (in solution at physiological pH), determined from the experimental pKa of 8.3 [43], and for deprotonation of a zincbound thiol, according to the experimental pKa value of 6.4 suggested from a pH dependence study with FTase [22]. These results definitively place thiol coordination as the only valid alternative for peptide entrance.
4. Conclusions In this study the potential energy surfaces for peptide coordination to the active site zinc coordination sphere in FTase were determined. Peptide coordination is one of several mechanistic dilemmas which have been preventing a detailed understanding, at the atomic level, of the catalytic mechanism of this enzyme. The energy profiles for peptide entrance were analyzed for the two possibilities suggested in the literature for the protonation state of CAAX cysteine: thiolate and thiol. Globally, our results clearly demonstrate
that a direct thiolate attack, with an energy barrier of 20.1 kcal/mol, is not a kinetically competitive alternative to thiol entrance (barrier 2.5 kcal/mol). This conclusion definitively confirms that peptide coordination to FTase proceeds through the thiol form, which implies a subsequent deprotonation in situ of the zinc-bound thiol to the thiolate form. The identity of the active-site base which receives this proton remains unknown. More experimental and computational studies are necessary to explain the catalytic mechanism of this enzyme.
Acknowledgements We thank the FCT (Fundac¸a˜o para a Cieˆncia e a Tecnologia) for a doctoral scholarship for Se´rgio Filipe Sousa (SFRH/BD/12848/2003) and the NFCR (National Foundation for Cancer Research) Centre for Computational Drug Discovery, University of Oxford, UK, for financial support.
References [1] S.L. Moores, M.D. Schaber, S.D. Mosser, E. Rands, M.B. O’Hara, V.M. Garsky, M.S. Marshall, D.L. Pompliano, J.B. Gibbs, J. Biol. Chem. 266 (1991) 14603–14610. [2] W.R. Schafer, J. Rine, Annu. Rev. Genet. 26 (1992) 209–237. [3] J.A. Glomset, C.C. Farnsworth, Annu. Rev. Cell. Biol. 10 (1994) 181–205. [4] F.L. Zhang, P.J. Casey, Annu. Rev. Biochem. 65 (1996) 241–269. [5] D.L. Pompliano, M.D. Schaber, S.D. Mosser, C.A. Omer, J.A. Shafer, J.B. Gibbs, Biochemistry 32 (1993) 8341–8347. [6] E.S. Furfine, J.J. Leban, A. Landavazo, J.F. Moomaw, P.J. Casey, Biochemistry 34 (1995) 6857–6862. [7] W.J. Chen, J.F. Moomaw, L. Overton, T.A. Kost, P.J. Casey, J. Biol. Chem. 268 (1993) 9675–9680. [8] Y. Reiss, M.S. Brown, J.L. Goldstein, J. Biol. Chem. 267 (1992) 6403–6408. [9] J.F. Hancock, A.I. Magee, J.E. Childs, C.J. Marshall, Cell 57 (1989) 1167–1177.
S.F. Sousa et al. / Journal of Molecular Structure: THEOCHEM 729 (2005) 125–129 [10] J.H. Jackson, C.G. Cochrane, J.R. Bourne, P.A. Solski, J.E. Buss, C.J. Der, Proc. Natl Acad. Sci. USA 87 (1990) 3042–3046. [11] K. Kato, A.D. Cox, M.M. Hisaka, S.M. Graham, J.E. Buss, C.J. Der, Proc. Natl Acad. Sci. USA 89 (1992) 6403–6407. [12] S. Rodenhuis, Semin. Cancer Biol. 3 (1992) 241–247. [13] J.M. Dolence, C.D. Poulter, Proc. Natl Acad. Sci. USA 92 (1995) 5008–5011. [14] Y. Takai, T. Sasaki, T. Matozaki, Physiol. Rev. 81 (2001) 153–208. [15] C.Y. Huang, L. Rokosz, Expert. Opin. Therap. Patents 14 (2004) 175–186. [16] S. Ayral-Kaloustian, E.J. Salaski, Curr. Med. Chem. 9 (2002) 1003–1032. [17] J. Ohkanda, D.B. Knowles, M.A. Blaskovich, S.M. Sebti, A.D. Hamilton, Curr. Top. Med. Chem. 2 (2002) 303–323. [18] J. Wiesner, K. Kettler, J. Sakowski, R. Ortmann, A. Katzin, E. Kimura, K. Silber, G. Klebe, H. Jomaa, M. Schlitzer, Angew. Chem. Int. Ed. 43 (2004) 251–254. [19] D. Chakrabarti, T. Da Silva, J. Barger, S. Paquette, H. Patel, S. Patterson, C.M. Allen, J. Biol. Chem. 277 (2002) 42066–43073. [20] J. Ohkanda, F.S. Buckner, J.W. Lockman, K. Yokoyama, D. Carrico, R. Eastman, K. Luca-Fradley, W. Davies, S.L. Croft, W.C. Van Voorhis, M.H. Gelb, S.M. Sebti, A.D. Hamilton, J. Med. Chem. 47 (2004) 432–445. [21] B.B. Bordier, J. Ohkanda, P. Liu, S.Y. Lee, F.H. Salazar, P.L. Marion, K. Ohashi, L. Meuse, M.A. Kay, J.L. Casey, S.M. Sebti, A.D. Hamilton, J.S. Glenn, J. Clin. Invest. 112 (2003) 407–414. [22] K.E. Hightower, C.C. Huang, P.J. Casey, C.A. Fierke, Biochemistry 37 (1998) 15555–15562. [23] S.B. Long, P.J. Casey, L.S. Beese, Struct. Fold. Des. 8 (2000) 209–222. [24] S. Dapprich, I. Komaromi, K.S. Byun, K. Morokuma, M.J. Frisch, J. Mol. Struct. (Theochem) 461 (1999) 1–21. [25] T. Vreven, K. Morokuma, J. Comput. Chem. 21 (2000) 1419–1432. [26] M.J. Frisch, G.W. Trucks, H.B. Schlegel, G.E. Scuseria, M.A. Robb, J.R. Cheeseman, J.A. Montgomery, T. Vreven, K.N. Kudin, J.C. Burant, J.M. Millam, S.S. Iyengar, J. Tomasi, V. Barone, B. Mennucci, M. Cossi, G. Scalmani, N. Rega, G.A. Petersson, H. Nakatsuji, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai, M. Klene, X. Li, J.E. Knox, H.P. Hratchin, J.B. Cross, C. Adamo, J. Jaramillo, R. Gomperts, R.E. Stratmann, O. Yazyev, A.J. Austin, R. Cammi, C. Pomelli, J.W. Ochterski, P.Y. Ayala, K. Morokuma, G.A. Voth, P. Salvador, J.J. Dannenberg, V.G. Zakrzewski, S. Dapprich, A.D. Daniels, M.C. Strain, O. Farkas, D.K. Malik, A.D. Rabuck,
[27] [28] [29] [30] [31] [32] [33] [34] [35] [36]
[37] [38] [39] [40] [41] [42] [43] [44] [45] [46] [47] [48] [49] [50]
129
K. Raghavachari, J.B. Foresman, J.V. Ortiz, Q. Cui, A.G. Baboul, S. Clifford, J. Cioslowski, B.B. Stefanov, G. Liu, A. Liashenko, P. Piskorz, I. Komaromi, R.L. Martin, D.J. Fox, T. Keith, A. Al-Lahan, C.Y. Peng, A. Nanayakkara, M. Challacombe, P.M.W. Gill, B. Johnson, W. Chen, M.W. Wong, C. Gonzalez, J.A. Pople, Gaussian, Inc., Pittsburg PA, 2003. C. Lee, W.T. Yang, R.G. Parr, Phys. Rev. B 37 (1988) 785–789. A.D. Becke, J. Chem. Phys. 98 (1993) 5648–5652. T. Ziegler, Chem. Rev. 91 (1991) 651–667. C.W. Bauschlicher, Chem. Phys. Lett. 246 (1995) 40–44. M.C. Holthausen, M. Mohr, W. Koch, Chem. Phys. Lett. 240 (1995) 245–252. A. Ricca, C.W. Bauschlicher, Theor. Chim. Acta 92 (1995) 123–131. U. Ryde, Biophys. J. 77 (1999) 2777–2787. M. Dolg, U. Wedig, H. Stoll, H. Preuss, J. Chem. Phys. 86 (1987) 866–872. D. Andrae, U. Haussermann, M. Dolg, H. Stoll, H. Preuss, Theor. Chim. Acta 77 (1990) 123–141. G. Frenking, I. Antes, M. Bohme, S. Dapprich, A.W. Ehlers, V. Jonas, A. Neubaus, M. Otto, R. Stegmann, A. Veldkamp, S.F. Vyboishchikov, Rev. Comput. Chem. 8 (1996) 63–144. J.J.P. Stewart, J. Comput. Chem. 10 (1989) 209–220. J.J.P. Stewart, J. Comput. Chem. 10 (1989) 221–264. J.J.P. Stewart, J. Comput. Chem. 12 (1991) 320–341. M. Brauer, M. Kunert, E. Dinjus, M. Klussmann, M. Doring, H. Gorls, E. Anders, J. Mol. Struct. (Theochem) 505 (2000) 289–301. P.A. Fernandes, M.J. Ramos, J. Am. Chem. Soc. 125 (2003) 6311–6322. P.A. Fernandes, M.J. Ramos, Chem. Eur. J. 9 (2003) 5916–5925. D.P. Wrathall, R.M. Izatt, J.J. Christensen, J. Am. Chem. Soc. 86 (1964) 4779–4783. M.J. Saderholm, K.E. Hightower, C.A. Fierke, Biochemistry 39 (2000) 12398–12405. P.A. Fernandes, M.J. Ramos, Chem. Eur. J. 10 (2004) 257–266. S.F. Sousa, P.A. Fernandes, M.J. Ramos, Biophys. J. 88 (2005) 483– 494. K. Morokuma, D.G. Musaev, T. Vreven, H. Basch, M. Torrent, D.V. Khoroshun, IBM J. Res. Dev. 45 (2001) 367–395. V. Pelmenschikov, P.E. Siegbahn, Inorg. Chem. 41 (2002) 5659–5666. M. Torrent, T. Vreven, D.G. Musaev, K. Morokuma, O. Farkas, H.B. Schlegel, J. Am. Chem. Soc. 124 (2002) 192–193. M. Kuno, S. Hannongbua, K. Morokuma, Chem. Phys. Lett. 380 (2003) 456–463.
