Binding free energy and kinetics: computation meets experiments
- Walter Rocchia (Italian Institute of Technology, Genova, Italy)
- Sara Bonella (CECAM EPFL, Switzerland)
- Andrea Cavalli (Istituto Italiano di Tecnologia and University of Bologna, Italy)
- Simone Meloni (Sapienza University of Rome, Italy)
- Vincenzo Barone (Scuola Normale Superiore Pisa, Italy)
- Chiara Cappelli (Scuola Normale Superiore Pisa, Italy)
Due to several questions, and in order to avoid ambiguities, we would like to clarify that the workshop will NOT take place at the Italian Institute of Technology but rather downtown Genoa!
IIT has agreed a rate with the following hotels:
in order to get the agreed rates, when you make your reservation please mention the agreement with the "FONDAZIONE ISTITUTO ITALIANO DI TECNOLOGIA" (IIT), which is supporting the workshop.
However, you might be able to find good or even better rates on some search engines such as booking.com or tripadvisor.it.
For the youngest, a few bed and breakfast near the workshop location follow:
The goal of this workshop is to survey the state of the art and foster progress in the field of bimolecular recognition, with particular focus on protein-ligand binding. This process is of key interest in biology, and has potentially great social and economical impact considering that it is at the basis of rational Drug Design and Discovery. We intend to bring together experts from the experimental and computational communities to share recent progress, recognize the main open challenges, and identify more precisely areas of complementarity/cooperation between experiments and simulations in view of pharmacologically interesting new applications. As an example, from an experimental perspective, it is often possible to characterize the final state of the binding process but a mechanistic description of the binding process as well as the identification of intermediate states can hardly be measured. Such a description could be (and in some cases is) provided by calculations. However, current progress in this area is limited by the fact that protein-ligand binding is a rare event, i.e. it can occur on timescales greater than those currently accessible via simulations. Thus, particular attention will be given to advanced computational techniques designed to reduce the time scale gap.
Specific attention will also be given to promoting interactions with industries. We will have scientists coming from Biotech and Pharmaceutical companies, who will describe how an early prediction of binding free energy and kinetic rate constants can benefit the drug discovery process.
To promote exchange of information among the different communities, ample time will be reserved for discussion both at the end of each talk and at the end of the sessions.
Rare events are defined as processes that occur on time-scales longer than those currently accessible by brute force simulations. This usually means up to tens of microseconds in classical molecular dynamics simulations and up to a few hundreds of picoseconds in ab initio simulations. Rare events can be described as transitions among states corresponding to local minima in the free energy surface of the system (metastable states). Typical questions addressed when studying rare events are the identification of metastable states of the system and the assessment of their relative stability, the determination of the microscopic mechanism of the process, and the evaluation of its rate. Addressing these questions would make it possible to characterize the most relevant aspects of a rare event and gain control of the phenomenon. A variety of methods have been developed to deal with each one of these questions. We divide them in three major categories [see 1 and refs therein for bibliography]: i) methods for reconstructing the free energy surface, such as Metadynamics , Temperature Accelerated Dynamics, Adiabatic Dynamics, Conformational Flooding, Temperature Accelerated Molecular Dynamics/Monte Carlo and Single Sweep, Alchemical Methods; ii) methods for identifying the reaction path, for example Nudged Elastic Band , Transition Path Sampling , Minimum Free Energy Path; iii) methods for computing the rate of a process/reaction like the already mentioned Transition Path Sampling, Milestoning  and Transition State Theory based methods .
While many interesting processes in different fields can be classified as rare events (chemical reactions, nucleation phenomena…) in this workshop we will focus on biophysical processes and in particular on protein-ligand binding. The affinity of a drug for a molecular target is considered to be, along with its pharmacokinetic and metabolic profile, a primary determinant of therapeutic effectiveness. For this reason, the affinity-related properties of new molecules – such as dissociation constants for receptor ligands or inhibition constants for enzyme inhibitors – play a key role in the selection of candidates for further development in the early stages of drug discovery.
Free energy calculations of ligand–receptor binding is a natural application of simulations in drug discovery [see ref. 7 and references therein]. Several approaches have been used. Grand canonical Monte Carlo simulations have been used to identify both potential ligands and their binding site(s) on the drug target. This method has proven successful in a few cases, for instance, in the design of novel nanomolar inhibitors of p38 kinase. High throughput molecular mechanics with Poisson–Boltzmann surface area (MM-PBSA) was used at Abbott Laboratories to directly estimate relative binding free energies for 308 ligands drawn from three representative drug discovery projects—the protease urokinase, the phosphatase PTP-1B, and the kinase Chk-1. Alchemical methods yield improved quantitative results at the cost of significantly more computation. The Jorgensen group’s work on non-nucleoside HIV reverse transcriptase (HIV-RT) inhibitors (NNRTI) is a notable recent example. Compound design decisions were based, in part, on calculated estimates of binding free energy differences, determined using free energy perturbation with Monte Carlo sampling, among various Cl-substituted test compounds . This optimization strategy has provided novel aminotriazines, possessing cellular EC50 values below 10 nM, effective against both wild-type HIV-RT and the resistant Tyr181Cys variant. Using the same method, a 5 uM virtual screening hit was transformed into a 55 pM inhibitor, apparently the most potent NNRTI reported to date . The results obtained thus far on HIVRT are quite encouraging, and the utility of this approach in other systems is an area of active investigation. Also, the process by which drugs bind to receptors has been studied in several systems. Benzamidine bound spontaneously to trypsin in MD simulations, achieving a good match to the crystal-structure–defined pose and revealing the binding pathway . The unbiased binding of kinase inhibitors and G protein–coupled receptor agonists and antagonists  has also been demonstrated. For example, the endogenous cannabinoid sn-2-arachidonoylglycerol was found to enter the binding pocket of a CB2 receptor homology model from the lipid bilayer. Notably, simulations of several beta-blockers and a beta-agonist binding to two beta-adrenergic receptors revealed where along the binding pathway dehydration of the ligand and receptor—long known to a major source of ligand affinity—occurs . The work further hinted that dehydration presents an unexpected kinetic barrier to binding, leading to suggestions on how ligand/receptor dehydration might be modulated to affect drug binding and unbinding kinetics.
Remarkably, from a Drug Discovery standpoint, it has recently emerged that drug binding kinetics, which is directly related to the length of time a drug spends in contact with its target, may be at least as important as the binding affinity [12-13]. The drug-target binding kinetics of new molecules, however, is seldom investigated, and only occasionally characterized retrospectively in post-launch analyses. At present, several factors limit the in silico and in vitro technologies needed to investigate this key property in a high-throughput manner, and the relationships between binding kinetics and chemical structures have not been explored in depth. Moreover, there has been no systematic investigation into the possibility of directly linking in vivo drug efficacy to the in silico predictions and in vitro evaluations of binding kinetics.
This Workshop is also announced on the CCL website
1 S. Bonella et al, European Physical Journal B, 85,97, 2012
2 A. Laio and M. Parrinello, Proceedings of the National Academy of Science, 99(20):1256212566, 2002.
3. H. Jonsson et al, Classical and Quantum Dynamics in Condensed Phase Simulations, World Scientific, 1998
4. C. Dellago et al, The Journal of Chemical Physics, 108(5):19641977, 1998.
5. A. Faradjian and R. Elber, The Journal of Chemical Physics, 120(23):1088010889, 2004.
6 E. Vanden-Eijnden and F. Tal, The Journal of Chemical Physics, 123(18):184103, 2005.
7 D. W. Borhani and D. E. Shaw, J Comput Aided Mol Des 26:1526, 2012.
8 Zeevaart JG et al., J Am Chem Soc 130:94929499, 2008.
9 Bollini M, et al., J Med Chem. 54 (24), pp 85828591 2011.
10 Buch I, Proc Natl Acad Sci USA 108:1018410189, 2011.
11 Dror RO, et al. Proc Natl Acad Sci USA, 108:1311813123, 2011.
12 Copeland, R.A.et al. Nat. Rev. Drug Discov. 5, 730, 2006.
13 Swinney, D.C. Nat. Rev. Drug Discov. 3, 801, 2004.