Water Thermodynamics of Peptide Toxin Binding Sites on Ion Channels

Abstract

Peptide toxins isolated from venomous creatures, long prized as research tools due to their innate potency for ion channels, are emerging as drugs as well. However, it remains challenging to understand why peptide toxins bind with high potency to ion channels, to identify residues that are key for activity, and to improve their affinities via mutagenesis. We use WaterMap, a molecular dynamics simulation-based method, to gain computational insight into these three questions by calculating the locations and thermodynamic properties of water molecules in the peptide toxin binding sites of five ion channels. These include an acid-sensing ion channel, voltage-gated potassium channel, sodium channel in activated and deactivated states, transient-receptor potential channel, and a nicotinic receptor whose structures were recently determined by crystallography and cryo-electron microscopy (cryo-EM). All channels had water sites in the peptide toxin binding site, and an average of 75% of these sites were stable (low-energy), and 25% were unstable (medium or high energy). For the sodium channel, more unstable water sites were present in the deactivated state structure than the activated. Additionally, for each channel, unstable water sites coincided with the positions of peptide toxin residues that previous mutagenesis experiments had shown were important for activity. Finally, for the sodium channel in the deactivated state, unstable water sites were present in the peptide toxin binding pocket but did not overlap with the peptide toxin, suggesting that future experimental efforts could focus on targeting these sites to optimize potency.

Potency- and Selectivity-Enhancing Mutations of Conotoxins for Nicotinic Acetylcholine Receptors Can Be Predicted Using Accurate Free-Energy Calculations

“…Taken together, these results suggest that LvIA displaces more and higher-energy unstable waters when binding to the α3β2 nAChR than the α3β4 nAChR, which could account for why it is more potent in the former subtype. These findings are consistent with a previous study in which water thermodynamics was used to explain mutagenesis data for a variety of peptide toxins for different ion channels [ 12 ].…”

“…Some principles for improving the selectivity of these conotoxins via mutagenesis have begun to emerge [ 11 ], such as focusing efforts on loop 2 vs. loop 1. In addition, the locations and thermodynamics of water sites in the peptide toxin binding pockets of ion channels, as computed by inhomogeneous solvation theory implemented in the WaterMap algorithm, have recently been shown to explain structure–activity relationships (SAR) for bungarotoxin and the muscle-subtype nAChR [ 12 ]; however, accurately predicting selectivity-enhancing mutations is still an arduous process with much uncertainty that would benefit from new approaches [ 13 ].…”

Potency-Enhancing Mutations of Gating Modifier Toxins for the Voltage-Gated Sodium Channel NaV1.7 Can Be Predicted Using Accurate Free-Energy Calculations

“…For the first step of the workflow, we ran a WaterMap of ProTx-II bound to Na V 1.7 to identify residues suitable for mutagenesis as has been performed previously [ 20 ]. WaterMap placed a total of 101 water sites within 5 Å of the protein-protein interface atoms ( Figure 7 B).…”

“…The WaterMap simulations were performed as described previously [ 20 ] using the construct described above. As this WaterMap simulation was “holo,” ProTx-II was retained during the simulation.…”

“…Computational techniques have the potential to utilize these cryo-EM structures to both guide where and what mutations should be made on a GMT. A recent study used WaterMap—a method which predicts the location, occupancy, and thermodynamics of water molecules—to partially explain the potency of ProTx-II in terms of its displacement of waters from unstable water sites near VSDII [ 20 ]. In addition, computational relative binding free-energy methods (RBFE) hold great promise for predicting how a mutation to a GMT will affect its potency for Na V 1.7.…”