 |
Mission | Partners | Facilities | Research | People | Publications | Pictures | Calendar | Contact |
 |
|||||||||
|
Siricharn S. Jirapongphan, Juliusz Warzywoda, David E. Budil, Michael Walters, Michael P. Manning, and Albert Sacco, Jr. Nine of the top-ten selling drugs had chiral active ingredients, and have accounted for about $53.5 billion (US) in global sales in 2004 [Thayer, A.M., Trial Separations, Chem. Eng. News, 83(36), pp. 49-53, 2005]. Chiral compounds typically occur as the racemic (equimolar) mixture of enantiomers (mirror-image forms). For many chiral drugs (for example, Pfizer’s cholesterol-lowering drug Lipitor), only a single enantiomer possesses the desired therapeutic activity. The other enantiomer is often inactive or even harmful. Therefore, isolation of the desired enantiomer is necessary to obtain an effective and safe drug. However, not all chiral compounds can be resolved using the currently available chiral stationary phases employed in enantioselective chromatography. This is due, in part, to a lack of insight into the separation mechanism. Molecular modeling in combination with experimentation can provide a better understanding of the chiral separation. Newly developed supercage-based docking, molecular dynamics, and grand canonical Monte Carlo simulations have been performed on enantioseparation processes utilizing chiral-modified zeolite HY (i.e., zeolite HY containing (R,R)-hydrobenzoin, see Figure 1).
Figure 1: Preferential Hydrobenzoin Adsorption Type (Window Site). Space-Filling Model Represents Hydrobenzoin Chiral Modifier. Stick Model Represents Zeolite HY Supercage (Adapted from Jirapongphan et al., Chirality, 2007, 19:514-517). The simulations predicted an enantioselectivity of 2.6 for (S)- and (R)-valinol enantiomers compared to that of 1.4 measured experimentally in this laboratory. The simulations suggested a single binding mode in S diastereomeric complex, and two binding modes in R complex. In the binding mode that was common to both complexes, three types of interactions were identified to involve aromatic hydrogen bonding between amine and two phenyl groups, aromatic hydrogen bonding between hydroxyl and a phenyl group, and steric interactions (Figure 2).
Figure 2: Binding Motif Identified from Lowest Energy Configurations of S Diastereomeric Complex between (S)-Valinol and Chiral Modifier (Adapted from Jirapongphan et al., Chirality, 2007, 19:508-513). The additional binding mode in the R complex had the aromatic hydrogen bonding interactions between amine and two phenyl groups replaced by the weaker aromatic hydrogen bonding interactions between amine and a single phenyl group (Figure 3).
Figure 3: Binding Motif Identified from Lowest Energy Configurations of R Diastereomeric Complex between (R)-Valinol and Chiral Modifier (Adapted from Jirapongphan et al., Chirality, 2007, 19:508-513). These multiple binding modes in the R complex resulted in its lower stability. The single binding mode in the S diastereomeric complex was hypothesized to be the origin of enantioselective separation in the chiral-modified zeolite HY. A potential usefulness of the proposed mechanism has been illustrated in an enantioseparation of (S)- and (R)-phenylglycinol (a precursor of AIDS drugs) on the chiral-modified zeolite HY model. Both phenylglycinol and valinol have identical amine and hydroxyl groups, but the valinol isopropyl group is substituted by the phenyl group in the phenylglycinol (Figure 4).
Figure 4: Molecular Structures of Valinol (Top) and Phenylglycinol (Bottom) (Adapted from Jirapongphan et al., Chirality, 2007, 19:514-517). In the phenylglycinol system, for which the modeling predicted an enantioselectivity of 1.3, the observed binding modes were the same as those in the valinol system. This suggests that specific structural features (i.e., number of polar groups) generating the hypothesized binding modes are required in enantioseparation in the chiral-modified zeolite HY. |
