AmericanChemicalSociety.com

Work was completed on a project that used NMR spectroscopy to characterize the binding of the chiral compound 1,1’-binaphthyl-2,2’-diyl hydrogen phosphate (BNP) to five molecular micelles with chiral dipeptide headgroups.  Molecular micelles have covalent linkages between the surfactant monomers and are used as chiral mobile phase modifiers in electrokinetic chromatography. NOESY analyses of (S)-BNP:molecular micelle mixtures showed that in each solution the (S)-BNP interacted predominately with the N-terminal amino acid of the molecular micelle’s dipeptide headgroup.  NOESY spectra were also used to generate group binding maps for (S)-BNP:molecular micelle mixtures.  In these maps, percentages are assigned to the (S)-BNP protons to represent the relative strengths of their interactions with a specified molecular micelle proton.  All maps showed that (S)-BNP inserted into a chiral groove formed between the molecular micelle’s dipeptide headgroup and hydrocarbon chain.  In the resulting intermolecular complexes, the (S)-BNP protons nearest the analyte phosphate group were found to point toward the N-terminal Hα proton of the molecular micelle headgroup.  Finally, pulsed field gradient NMR diffusion experiments were used to measure association constants for (R) and (S)-BNP binding to each molecular micelle.  These K values were then used to calculate the differences in the enantiomers’ free energies of binding, Δ(ΔG).  The NMR-derived Δ(ΔG) values were found to scale linearly with EKC chiral selectivities from the literature.

Work continued on a project that used NMR relaxation experiments to investigate the motional dynamics of chiral molecules bound to chiral molecular micelles.  In this project, spin-lattice and spin-spin relaxation times were measured for each of a chiral molecule’s protons when the molecule was bound to a molecular micelle.  These relaxation times were in turn used to calculate correlation times for each proton in the molecule.  The correlation time is the time required for the proton to rotate through one radian.  It was hypothesized that the chiral molecule’s reorientational motion would slow when it bound to the much larger molecular micelle and the resulting increase in the molecule’s proton correlation times could be used as a probe of chiral recognition.  Previous work had shown that when the chiral molecule 1,1’-binaphthyl-2,2’-diyl hydrogen phosphate (BNP) bound to a molecular micelle containing a single amino acid headgroup, the BNP correlation times increased as expected.  In addition, the correlation times for each BNP proton in the bound state were found not to differ significantly.  This result suggested that the bound BNP molecule still rotates in a near-isotropic fashion when bound to the molecular micelle.  In contrast, when BNP was bound to a molecular micelle with a chiral dipeptide headgroup, the BNP protons nearest the binding site had longer correlation times and the BNP protons farther from the binding site had shorter correlation times.  In other words, anisotropic motion was detected for BNP bound to this molecular micelle.  This summer, two additional dipeptide-terminated molecular micelles were investigated to determine if the motion of the bound BNP molecule was always anisotropic when it interacted with a molecular micelle containing a dipeptide headgroup.  Relaxation experiments with poly(sodium N-undecanoyl-L-valylalainate) and poly(sodium N-undecanoyl-L-leucylleucinate) showed results nearly identical to those obtained with the dipetide-terminated molecular micelle studied previously.  The BNP protons that were closest to the phosphate group or the point of attachment between the polymer and analyte had statistically different correlation times than those protons on the opposite side of the molecule.  These results suggested that the motional behavior of BNP bound to all dipeptide-terminated molecular micelles was similar and that the NMR relaxation method employed can be used to identify the atoms that become less mobile as a result of association with the molecular micelle.

Finally, a new project was initiated that sought to use NMR spectroscopy to study the binding of two chiral β-blocker drugs (atenolol and propranolol) to a chiral surfactant. β-blocker drugs hinder the effects of adrenaline and reduce the force with which the heart contracts.  The surfactant investigated is sold by Waters, Inc. as EnantioSelect, and like the molecular micelles discussed above is used as a chiral selector in electrokinetic chromatography.  The long term goal of this project is to again use NMR spectroscopy to characterize the intermolecular interactions and thermodynamic factors responsible for chiral recognition in electrokinetic chromatography.  All four diastereometric drug:surfactant complexes (S:S, S:R, R:R, and R:S) were studied.  NMR diffusion experiments with atenolol showed that the binding constants were largest for the S:S (39.2±1.8) and R:R (33.9±1.7) drug:surfactant diasteriomers.  The binding constants for the S:R and R:S diastereomers were, respectively 31.5±1.6 and 17.9±2.2. In contrast, binding constants for the propranolol:surfactant diastereomers were all very large and did not differ significantly.  Two-dimensional ROESY experiments were used to establish spatial relationships between drug and surfactant atoms in the bound state.  This structural data was then used in molecular modeling calculations to generate molecular structures and energies for the drug:surfactant complexes. In the atenolol studies, the molecular modeling energies were consistent with the association constant measurements, in that the lowest energy complexes had the largest binding constants.  We are currently refining the modeling calculations and using the structures of the minimum energy complexes to identify specific intermolecular interactions that lead to the differences observed in the binding affinities.  Next summer we plan to use the same methodology to extend this study to other chiral molecule:surfactant complexes.