Relative Protein-Ligand Binding Constants

The cost of a molecular dynamics (MD) free energy study depends very much on both the system and the goal of the study. If the goal is to reproduce qualitatively an experimental number and interpret it in terms of microscopic interactions, and if the systems of interest (e.g., native and mutant protein) are very similar, then only limited conformational sampling will be needed in most cases, and a few short runs with a small model may suffice. 

Alchemical transformations have also been applied to the challenging case of G protein-coupled receptors (GPCRs), for which little structural information is available experimentally at the atomic level. Starting from a template of a seven-helix 

These examples show that for difficult cases, and especially when a prediction is being made, a large number of simulations may be necessary. Today, the continuing increase in computer power has made such multiple simulations possible in a reasonable time frame. Several other recent studies illustrate the scope of molecular dynamics free energy for molecular recognition problems they include studies of nucleic acids [13], proteins [14-16], and methodological studies of convergence and precision [17, 18]. Several recent reviews provide additional examples [19, 20]. 

---

One asset of mass spectrometry in protein science is that ESI and MALDI [11, 75] can introduce noncovalent complexes to the gas phase [12, 76, 77]. If one can assume that the gas-phase ion abundances (peak intensities) for the complex, apo protein, and ligand are directly related to their equilibrium concentrations in solution, the relative and absolute binding affinities can be deduced [78-81]. Extended methods are now available that also make use of the intensity of the complex and the protein at high ligand concentration to determine binding constants [78, 82-84]. 

Figure 6.6 Spectrophotometric titration of the binding of inositol hexaphosphate (IHP) to methemoglobin (Methb). The complex has an increased absorbance at 512 and 649 nm, and no increases at 640,618,588, and 599 ran. The concentration of methemoglobin (20 fjM) is about 14 times higher than the dissociation constant of 1.4 f.iM for the complex. The intersection of the slope of the increase in absorbance with the maximum value gives the stoichiometry (1, in this case). Note that this simple procedure cannot be used if the protein is not initially present at such a high concentration relative to the dissociation constant, since the assumption is that all the added ligand is bound to the protein for the early additions.

---