Models competitive proton binding

With respect to the mechanisms of ion binding, model e places all charges of adsorbed ions in the fi plane, except the protons, which are placed in the surface plane. This corresponds to the original application of the triple-layer model by Davis et al. [4], In the more recent interpretation this practice would mean that all ions are considered as outer-sphere complexes. The consequence is that competition between electrolyte ions (A , C ) and the other ions placed in the fi plane can be made substantial (this is also at least partially the case for charge distribution, which has actually been mentioned in these early papers). 

It is probable that the negative charge induced by these three electrons on FeMoco is compensated by protonation to form metal hydrides. In model hydride complexes two hydride ions can readily form an 17-bonded H2 molecule that becomes labilized on addition of the third proton and can then dissociate, leaving a site at which N2 can bind (104). This biomimetic chemistry satisfyingly rationalizes the observed obligatory evolution of one H2 molecule for every N2 molecule reduced by the enzyme, and also the observation that H2 is a competitive inhibitor of N2 reduction by the enzyme. The bound N2 molecule could then be further reduced by a further series of electron and proton additions as shown in Fig. 9. The chemistry of such transformations has been extensively studied with model complexes (15, 105). 

To end up with a predictive pharmacophore model, it is necessary to start with reliable structural and biological data. First of all, it is important to have correct 3D structures of all compounds under study. Thus, atomic valences, bond orders, protonation state and stereochemistry have to be checked carefully. Also the consideration of different possible tautomers is necessary when the bioactive form is not exactly known. Another prerequisite is the existence of a similar binding mode of all ligands under study. Experimental data, from competition experiments or protein-ligand crystal structures, can clearly point out that the ligands interact with the same binding epitope in a similar way and not on distinct binding sites. 

Proton NMR titrations in DMSO-de/O.S water at 298 K were conducted to elucidate stability constants for a number of anions, added as their terabuty-lammonium salts. Fluoride affects the largest change on the proton resonance however an association constant was only calculated for 60a (>10 M ) therefore titrations were repeated in more competitive media (DMSO-d6/5% water). Both receptors showed selectivity for fluoride however 60a bound with a 1 1 binding stoichiometry (1360 M ) whereas the data for 60b could only be fitted to a 1 2 receptor/anion model (iCi = 940 andK2 = 21 M ). The crystal structures of 60a with chloride and fluoride are shown in Figs. 8 and 9 respectively. 

Further, this model is based on adsorption modeling (Section 11.1) thus, the binding of ions at each plane is treated as Langmuir competitive adsorption, with consideration for electrostatic interactions in the adsorption equilibrium constants as in Equations 11.16 and 11.17. Thus, the charge at the 0 plane is given by the balance between adsorbed protons and adsorbed hydroxide ions... 

The conclusion of Ward that Cl binds directly to the Zn at the active cleft has been questioned by Koenig and Brown (70). From a comparison of water proton and Cl relaxation data they argued that competition between Cl and water for a common Zn site, as also implied in Ward s model, does not occur. They Instead suggest that Cl may be bound to a cationic residue elsewhere on the protein. Nome et al. have reported that the Cl excess linewidth observed in carbonic anhydrase solutions is markedly affected by the addition of the anion Au(CN)2 (71). This anion binds very poorly to Zn and other metal ions. This result also indicates that Cl- is not directly coordinated to zinc. 

---