Stepwise Aggregation of Cholate and Deoxycholate Dictates the Formation and Loss of Surface-Available Chirally Selective Binding Sites.

Bile salts are facially amphiphilic, naturally occurring chemicals that aggregate to perform numerous biochemical processes. Because of their unique intermolecular properties, bile salts have also been employed as functional materials in medicine and separation science (e.g., drug delivery, chiral solubilization, purification of single-walled carbon nanotubes). Bile micelle formation is structurally complex, and it remains a topic of considerable study. Here, the exposed functionalities on the surface of cholate and deoxycholate micelles are shown to vary from one another and with the micelle aggregation state. Collectively, data from NMR and capillary electrophoresis reveal preliminary, primary, and secondary stepwise aggregation of the salts of cholic (CA) and deoxycholic (DC) acid in basic conditions (pH 12, 298 K), and address how the surface availability of chirally selective binding sites is dependent on these sequential stages of aggregation. Prior work has demonstrated sequential CA aggregation (pH 12, 298 K) including a preliminary CMC at ca. 7 mM (no chiral selection), followed by a primary CMC at ca. 14 mM that allows chiral selection of binaphthyl enantiomers. In this work, DC is also shown to form stepwise preliminary and primary aggregates (ca. 3 mM DC and 9 mM DC, respectively, pH 12, 298 K) but the preliminary 3 mM DC aggregate is capable of chirally selective solubilization of the binaphthyl enantiomers. Higher-order, secondary bile aggregates of each of CA and DC show significantly degraded chiral selectivity. Diffusion NMR reveals that secondary micelles of CA exclude the BNDHP guests, while secondary micelles of DC accommodate guests, but with a loss of chiral selectivity. These data lead to the hypothesis that secondary aggregates of DC have an exposed binding site, possibly the 7α-edge of a bile dimeric unit, while secondary CA micelles do not present binding edges to the solution, potentially instead exposing the three alcohol groups on the hydrophilic α-face to the solution.

An Edge Selection Mechanism for Chirally Selective Solubilization of Binaphthyl Atropisomeric Guests by Cholate and Deoxycholate Micelles.

Combining micellar electrokinetic capillary chromatography (MEKC) and nuclear magnetic resonance (NMR) experimentation, we shed light on the structural basis for the chirally selective solubilization of atropisomeric binaphthyl compounds by bile salt micelles comprised of cholate (NaC) or deoxycholate (NaDC). The model binaphthyl analyte R,S-BNDHP exhibits chirally selective interactions with primary micellar aggregates of cholate and deoxycholate, as does the closely related analyte binaphthol (R,S-BN). Chiral selectivity was localized, by NMR chemical shift analysis, to the proton at the C12 position of these bile acids. Correspondingly, MEKC results show that the 12α-OH group of either NaC or NaDC is necessary for chirally selective resolution of these model binaphthyl analytes by bile micelles, and the S isomer is more highly retained by the micelles. With NMR, the chemical shift of 12ß-H was perturbed more strongly in the presence of S-BNDHP than R-BNDHP. Intermolecular NOEs demonstrate that R,S-BNDHP and R,S-BN interact with a similar hydrophobic planar pocket lined with the methyl groups of the bile salts, and are best explained by the existence of an antiparallel dimeric unit of bile salts. Finally, chemical shift data and intermolecular NOEs support different interactions of the enantiomers with the edges of dimeric bile units, indicating that R,S-BNDHP enantiomers sample the same binding site preferentially from opposite edges of the dimeric bile unit. Chirality 28:525-533, 2016. © 2016 Wiley Periodicals, Inc.

Determining sequential micellization steps of bile salts with multi-CMC modeling.

HYPOTHESIS: Bile salts exhibit complex concentration-dependent micellization in aqueous solution, rooted in a long-standing hypothesis of increasing size in bile aggregation that has historically focused on the measurement of only one CMC detected by a given method, without resolving successive stepwise aggregates. Whether bile aggregation is continuous or discrete, at what concentration does the first aggregate form, and how many aggregation steps occur, all remain as open questions. EXPERIMENTS: Bile salt critical micelle concentrations (CMCs) were investigated with NMR chemical shift titrations and a multi-CMC phase separation modeling approach developed herein. The proposed strategy is to establish a correspondence of the phase separation and mass action models to treat the first CMC; subsequent micellization steps, involving larger micelles, are then treated as phase separation events. FINDINGS: The NMR data and the proposed multi-CMC model reveal and resolve multiple closely spaced sequential preliminary, primary, and secondary discrete CMCs in dihydroxy and trihydroxy bile salt systems in basic (pH 12) solutions with a single model of one NMR data set. Complex NMR data are closely explained by the model. Four CMCs are established in deoxycholate below 100 mM (298 K, pH 12): 3.8 ± 0.5 mM, 9.1 ± 0.3 mM, 27 ± 2 mM, and 57 ± 4 mM, while three CMCs were observed in multiple bile systems, also under basic conditions. Global fitting leverages the sensitivity of different protons to different aggregation stages. In resolving these closely spaced CMCs, the method also obtains chemical shifts of these spectroscopically inaccessible (aka dark) states of the distinct micelles.

Sodium cholate aggregation and chiral recognition of the probe molecule (R,S)-1,1'-binaphthyl-2,2'-diylhydrogenphosphate (BNDHP) observed by 1H and 31P NMR spectroscopy.

Bile salt micelles can be employed as a pseudostationary phase in micellar electrokinetic capillary chromatography (MEKC) separations of chiral analytes. To improve MEKC separations of chiral analytes, a molecular level understanding of micelle aggregation in the presence of analyte is needed. Here, aggregation of sodium cholate has been observed by exploiting the presence of a model analyte molecule. The 31P and 1H nuclear magnetic resonance spectroscopy (NMR) chemical shifts of (R,S)-1,1'-binaphthyl-2,2'-diylhydrogenphosphate ((R,S)-BNDHP), a model analyte in chiral MEKC separations, are demonstrated to be very sensitive to the aggregation state of the bile salt sodium cholate. In addition to probing micellar aggregation, the NMR spectral resolution of enantiomeric species is also stronglycorrelated with chiral separations in MEKC. In this work, the aggregation of sodium cholate in basic solutions (pH 12) has been observed over the concentration range 0-100 mM. The primary critical micelle concentration (cmc) was found to be 14 +/- 1 mM for basic solutions of sodium cholate. In addition, a primitive aggregate is clearly observed to form at 7 +/- 1 mM sodium cholate. The data also show pseudo-cmc behavior for secondary aggregation observed in the regime of 50-60 mM cholate. Finally, the H5-H7 edge of BNDHP is shown to be sensitive to chirally selective interactions with primary cholate micelles.

Proton NMR assignments for R,S-1,1'-binaphthol (BN) and R,S-1,1'-binaphthyl-2,2'-diyl hydrogen phosphate (BNDHP) interacting with bile salt micelles.

We report proton chemical shifts for two model chiral analytes that are commonly used in the study of micellar electrokinetic capillary chromatography (MEKC), R,S-1,1'-binaphthol (1, BN) and R,S-1,1'-binaphthyl-2,2'-diyl hydrogen phosphate (2, BNDHP), in the absence and presence of monomers and micelles of sodium cholate and sodium deoxycholate. The analytes undergo fast exchange in and out of the micelles, which perturbs the analytes' chemical shifts, and which we use to resolve some resonances that are degenerate at both 300 and 600 MHz. Although BN and BNDHP are simple molecules, the proton assignments are only unambiguously established with the aid of the exchange with micelles, an attractive alternative to other methodologies such as the use of paramagnetic shift reagents which may also cause spectral distortions. We rely also upon 2D-NOE spectra of samples in the presence of micelles to perform these assignments. Recently published assignments, which were based upon 2D-COSY spectroscopy, appear to be in error and are corrected here. Finally, we note that these shifts are information-rich reporters on the nature of the interactions of these model analytes with the micelles.