A procedure was designed to solve the molecular binding of the haloether sevoflurane to the Kv1. The procedure consisted of i an extensive production of docking solutions for the ligand-receptor interaction, ii clustering of docking solutions into binding sites along the receptor structure and iii estimation of binding affinities using the free-energy perturbation FEP method.
First completion of steps i through iii solved the ligand channel interaction for singly-occupied binding sites. Double occupancy of receptor sites were investigated by inputing the first generated ensemble of docked structures into another round of i through iii calculations. In detail, step i was accomplished by docking sevoflurane as a flexible ligand molecule against an MD-generated ensemble of membrane-equilibrated structures of the channel to properly handle the molecular flexibility of the protein receptor.
Docking calculations were restricted to the pore domain region of the channel, free from the membrane surroundings. Each of these volumes were treated as binding site regions in step iii calculations. A detailed description of the calculations is provided as Supplementary Text. Here, the main goal is to contribute a theoretical structure-based study of concentration-dependent binding of ligands against multiple saturable sites in membrane proteins. The work is illustrated in the context of binding of the general anesthetic sevoflurane to the well-understood open structure of the Kv1.
Our choice is justified as previous findings support that sevoflurane binds Kv1. Given that, we treated the system reservoir as a homogeneous aqueous solution despite its intrinsic inhomogeneity provided by the solvated lipid bilayer. The interaction sites spread over the TM region of the channel at the S4S5 linker, at the S6P-helix interface of adjacent subunits and at the extracellular face. Resolution of sevoflurane sites on Kv1.
Inset illustrates the MD-generated ensemble of channel structures considered for docking calculations. Shown is the ensemble average structure of the channel along with the set of centroid configurations of sevoflurane points determined from docking.
Centroid configurations of sevoflurane were clustered as a function of their location on the channel structure, that is at the S4S5 linker orange , at the S6P-helix interface blue and at the extracellular face green next the selectivity filter.
Inset shows a representative molecular structure resolved from docking. Voltage-sensor domains of the channel are not shown for clarity in b , c. From the docking ensemble, there are up to 3 12 occupancy states of the channel that might contribute to sevoflurane binding. To evaluate this quantitatively, we performed a series of FEP calculations to estimate the per site binding affinity for one and two bound ligands via equation 12 Table 1 , Supplementary Fig. As shown in Table 1 , binding constants or absolute binding free-energies for the individual sites are heterogeneous and take place under a diverse range ie.
There is however a clear decreasing trend of affinities involving sites respectively at the S4S5 linker, S6P-helix interface and extracellular face. Within this fraction, the most likely states involve single and double sevoflurane occupancy of the S4S5 linker as expected from the affinities reported in Table 1. State-dependent binding probabilities for different concentrations of sevoflurane at the reservoir. Strings for the four most likely states are highlighted.
As shown in Fig. Position-dependent binding probabilities for different concentrations of sevoflurane at the reservoir. Voltage-sensor domains are not shown for clarity. So far, our study supports that in average 0. It is informative to clarify further the dependence of the results on concentration changes of the ligand in the reservoir. In contrast, the probability density for sites at the selectivity filter remains negligible for all concentrations.
As shown in Table 1 , note for completeness that equilibrium constants for doubly-occupied sites are comparable to or even higher than estimates for one-bound ligand thus revealing important saturation effects in which one or two ligands can stably bind the channel at individual sites. The result is especially true for spots at the S4S5 linker. Sevoflurane sites identified from docking are therefore indistinguishable across the channel subunits implying that there might effectively be 3 distinguishable regions for ligand binding on the channel structure that is, S4S5 linker, S6P-helix interface and selectivity filter.
Table 2 shows average estimates and associated errors for sevoflurane affinities against each of these distinguishable regions. According to equation 18 , statistical errors reflect the structural heterogeneity across the channel subunits implicit in the calculations as a result of finite MD-sampling of the Kv1. Given that, Table 2 must provide us with statistically improved estimates when describing sevoflurane affinities to each of the distinguishable sites on Kv1.
Reduction to symmetry causes redistribution of ligand-channel probabilities without modifying its average properties. Strings for the five most likely states are highlighted. Membrane proteins are primary targets for a large fraction of small molecule drugs that likely bind the protein receptor through complex concentration and saturation effects.
Understanding the molecular structure of ligand binding thus prompts new advances in experimental and theoretical fronts, justifying the work herein. We presented a theoretical approach based on docking and FEP to study concentration-dependent interactions of ligands to multiple saturable sites in membrane receptors. Here, our study relies on two underlying assumptions that i docking can faithfully describe ligand interactions at protein sites and that ii binding events are independent over multiple sites.
The generation of false positive hits is however a well documented drawback of docking algorithms as a result of limitations of the scoring function in describing ligand solvation energies and protein flexibility In this regard, the combination of extensive docking calculations against an ensemble of equilibrium receptor structures to handle protein flexibility and FEP calculations based on fine force-fields to accurately estimate solvation energies are critical aspects of the presented strategy to minimize such drawbacks In this regard, although not considered here, it might be important to integrate docking results from different algorithms involving different scoring functions in order to characterize the bound ensemble.
Still, thanks to the generality of the presented formulation, extension of the current approach to sampling techniques other than docking, including all-atom flooding-MD simulations 3 , 6 , 7 , 11 , might also be an important refinement in that direction manuscript in preparation. When compared to docking, flooding-MD applied to membrane protein has however the disadvantage of handling with full partition of the ligand into protein sites for which slow kinetics may reflect into high computational costs for sampling convergence.
In relation with assumption ii , it is true that for case specific systems, the additive PMF in equation 10 may be a severe approximation that will likely fail as soon as nearby sites are simultaneously occupied. Given that, elaboration of a proper treatment of site dependence in multiple binding events and evaluation of its usefulness will be highly welcome in future studies.
Before that and for certain systems, the formulated work based on equation 10 must therefore be seen as an 0-level approximation of more elaborate and still more complex descriptions of the binding constant.
The approach is illustrated here in the context of sevoflurane binding to Kv1. A detailed description of sevoflurane binding and its implications for Kv1. Still, we find it pertinent to discuss key results of the study. The model system was chosen as previous findings support that sevoflurane binds Kv1.
Specifically, sevoflurane shifts leftward the voltage-dependence of channel and increases its maximum conductance.
Overall, our calculations demonstrate that sevoflurane binds Kv1. From a physical-chemical point of view, spots at these channel regions are primarily hydrophobic pockets Supplementary Fig. S2 providing with favorable interaction sites for the uncharged sevoflurane molecule. In contrast to the aforesaid spots, sites nearby the selectivity filter of Kv1.
S2 that disfavors sevoflurane interaction as reflected in the free-energies shown in Table 1. The unfavorable binding free-energies for the singly-occupied site thus support that the non-negligible fraction of poses determined from docking Fig. Its is particularly worth of mention that our findings recapitulate independently very recent photolabeling experiments demonstrating that photoactive analogs of sevoflurane and propofol do interact at the S4S5 linker and the S6P-helix interface of Kv1.
In detail, Leu and Thr were found to be protected from photoactive analogs, with the former being more protected than the latter. As highlighted in Supplementary Fig. S3 , atomic distances of these amino-acid to bound sevoflurane molecules at the S4S5 linker and S6P-helix interface are found here to be respectively 5.
Such intermolecular distances imply their direct interactions with bound sevoflurane in agreement with the measured protective reactions. Besides that, our calculations also recapitulate the stronger protection of Leu in the sense that, relative to sites at the S6P-helix interface, the affinity of sevoflurane is found here to be higher at S4S5 linker given its stable occupancy by one or two ligands. In special, a single residue Gly at a critical pivot point between the S4S5 linker and the S5 segment underlines potentiation of Kv1.
When bound at the S4S5, sevoflurane is found here to be in proximity to that amino acid Supplementary Fig. The stable interaction of sevoflurane at the S4S5 linker of Kv1.
On the other hand, the unfavorable or absent interactions at the central cavity and next the selectivity filter of Kv1. Because sevoflurane induces potentiation rather than blocking of Kv1. Given the critical role of the S4S5 linker for the gating mechanism of the channel 22 , it is likely that sevoflurane-S4S5 interactions as found here are at the origins of the experimentally measured voltage-dependent component of anesthetic action.
Such hypotheses have been raised also in the context of anesthetic action on bacterial sodium channels 7 , More than four samples are recommended for each concentration. Heat-dry the well plates in a dry bath incubator until all filter membranes are dry. This usually takes approximately 15 min. For each experimental measurement, subtract the cpm values of the groups.
Added activities [TA] could be measured using the standard samples. Total binding activities [TB] and non-specific binding activities [NSB] could be measured by the activities on the membranes of Plate A and B, respectively. Using Eqs. K d and B max could be calculated using a Scatchard plot, Woolf plot or the software.
Referring to the Results sheet for the regression analysis, the X - and Y -axis intercepts could be calculated. Referring to the results sheet for the regression analysis, the X - and Y -axis intercepts could be calculated. Create a new project file on Prism Motulsky For each point on the saturation-binding curve, enter the concentration of ligand into the X column and [SB] into the Y column.
Prism displays the best-fit values B max and K d for the binding parameters in the results sheets. The labeling yield of I-Nimotuzumab was The specific activity was Figure 2 A shows an example of a typical equilibrium saturation curve using the radiolabeled assay with increasing concentrations of I-Nimotuzumab.
In parallel comparisons from the same data obtained from the binding assays, the Scatchard plot, the Woolf plot and the saturation radioligand-binding curve gave similar estimates of the K d 7.
Cell sample preparation Culture receptor-overexpressed cells in corresponding medium under certain conditions. Collect the cells from the flasks or the plates. At least 10 5 cells are needed for each well. It takes 96 wells to test the binding of one ligand with its receptor. Wash the cell solution with 0. Cold protein L preparation mg of protein L is dissolved in or diluted with 2. However, it takes a large amount of ligand to finish the experiment. In general, the cold ligands should be times more concentrated than the radiolabeled ligand to block the receptors.
Fewer cold ligands could be used in low concentrations. Moreover, fewer concentrations for example, eight concentrations could conserve many ligands. This work was supported by National Natural Science Foundation of China , and This article does not contain any studies with human or animal subjects performed by any of the authors.
National Center for Biotechnology Information , U. Biophysics Reports. Biophys Rep. Published online Feb Author information Article notes Copyright and License information Disclaimer. Fan Wang, Email: nc. Corresponding author.
Received Nov 10; Accepted Dec This article has been cited by other articles in PMC. Abstract The reversible combination of a ligand with specific sites on the surface of a receptor is one of the most important processes in biochemistry.
Keywords: Equilibrium constant, Maximum density of receptors, Saturation binding assays, I labeling, Radioligand. Count the activity in the final recovered tube and calculate the specific activity. Table 1 Sample adding strategy in the typical well plate for the specific binding assay. Specific binding group No. Open in a separate window. Table 2 Sample adding strategy in the typical well plate for the non-specific binding assay.
Non-specific binding group No. Step 1—3 Preparation of the I-Protein L and cold ligands takes approximately 1 h. Step 4—6 Preparation of the receptor samples takes approximately 1 h. Step 7—13 The cell-binding assay usually takes 4—6 h, depending on how many samples are used. Step 13—16 Activity measurements and data analyses take approximately 1—2 h. Step 6 The presence of certain metal ions e. Binding buffer without these ions will result in low counts from the collected membrane.
Step 7 It is necessary to add the cells and buffer with multiple-channel pipettes to reduce the time of this step. It takes practice to become skilled at adding serial concentrations of radioligand and cold ligand. It is important to stay focused and patient.
Step 8 The non-specific binding assay requires a large quantity of the cold ligand. Usually the ligands are difficult to prepare or very expensive.
The Scatchard plot and the Woolf plot could be completed using fewer concentrations and fewer parallel samples. In total, about 20 samples are sufficient for fitting the linear regression. Table 3 Comparison of the results obtained by three plots. One advantage of fitting total binding only is that equations have been derived for fitting such data, even when a substantial fraction of the ligand binds, resulting in ligand depletion free concentration substantially less than the added concentration.
Prism offers models for fitting one or two sites. You can use choices in the Compare tab to compare the two fits. When comparing the fits to the one- and two-site models, use common sense as well as statistics. Don't accept a two site model, if one of the sites is only a tiny fraction of the total, or if its Kd is outside the range of radioligand concentrations you used in the experiment.
All rights reserved.
0コメント