Lipophilicity plays a pivotal role in drug discovery. It serves as an essential proxy for various physical parameters associated with ADMET (absorption, distribution, metabolism, excretion, and toxicity). Lipophilicity reflects a compound’s ability to partition into cell membranes, making it a valuable metric for evaluating a pharmaceutical’s capacity to permeate the diverse cell membranes required for it to reach its site of action. Although lipophilicity has proven successful in the drug discovery process, it is crucial to recognize that the octanol phase, commonly used in these measurements, is a highly simplified model compared to the complex composition and chemical environment of cell membranes. Cell membranes are anisotropic, composed of numerous lipid species, each with its unique physicochemical properties that influence the spatial organization and dynamics of lipids within the lipid bilayer. In contrast, octanol presents an isotropic environment, and this discrepancy between octanol and cell membranes affects their interaction with drugs. In general, compounds exhibit a stronger affinity for cell membranes than for octanol, primarily due to enthalpic and entropic factors. The presence of charged species and steric effects further complicates the picture, with octanol accommodating bulky groups that might disrupt lipid packing within the bilayer.
In a recent study published in the peer-reviewed Angewandte Chemie International Edition Journal, a group of researchers from the University of Southampton (UK), led by Professor Bruno Linclau (now at Ghent University, Belgium) and Professor Philip Williamson and conducted by Dr. Zhong Wang, Hannah Felstead and Robert Troup, investigated the influence of aliphatic fluorination on membrane permeability, utilizing innovative solid-state NMR techniques to gain insights into this critical aspect of pharmaceutical development.
Fluorination of drug candidates is a well-established strategy to modify their biological and physical properties. It can lead to significant changes in lipophilicity, which has prompted extensive research into how fluorination motifs modulate lipophilicity. However, prior to this study, a detailed investigation into the effect of aliphatic fluorination on water-membrane partitioning had not been undertaken. In this context, the researchers aimed to explore whether the often subtle modifications in lipophilicity resulting from fluorination are mirrored in membrane partitioning. Specifically, they investigated whether 1-octanol, commonly used as a model for membrane permeability, remains valid when considering purely aliphatic fluorination modifications.
To address these questions, the authors selected closely related compound analogues with subtle differences in fluorination motifs. These compounds were then studied using aqueous solutions of multilamellar lipid vesicles, which mimic eukaryotic cell membranes. Three different compound series with varying steric demands and lipophilicity ranges were chosen for the study. One of the remarkable contributions of this research lies in the development of a convenient magic-angle spinning (MAS) 19F solid-state NMR methodology. This methodology allows the determination of the partitioning of fluorinated compounds into lipid vesicles without the need for vesicle separation. While traditional solution-state NMR methods rely on the detection of free compounds in solution, solid-state NMR methods are more challenging due to restricted motion in a membrane environment. In this context, 19F solid-state NMR emerged as an ideal tool because of the sensitivity of 19F chemical shifts to changes in electrostatic environments. The use of MAS and low-power proton decoupling enabled the researchers to obtain well-resolved spectra that discriminated between the free and membrane bound populations.
The authors systematically investigated the factors influencing the measured logKP , including compound concentration and bilayer hydration. Their results revealed that the mean logKP remained consistent across varying levels of hydration and compound concentration. This consistency provided not only excellent reproducibility but also ensured robust signal-to-noise ratios and manageable spectral acquisition times. They encompassed different sets of fluorinated aliphatics, each exhibiting specific variations in lipophilicity due to changes in fluorination motifs. The researchers observed that changes in lipophilicity were reflected in the corresponding water-membrane partition coefficients (logKP), with the strength of the correlation varied depending on the compound series under consideration.
For instance, the cyclopropylmethyl derivatives demonstrated an outstanding logKP – logKPw correlation, including comparisons between substrates with different fluorine stereochemistries, positions, and numbers. In contrast, the glycosides exhibited a moderate correlation, with variations in logKP arising from changes in epimers. The pentane-1,5-diol derivatives demonstrated excellent logKP – logKPw correlations when considered separately based on fluorination positions and motifs. Interestingly, the authors observed that octanol/water and liposome systems displayed differences in their ability to discriminate between lipophilicity. The correlation between logKPw and logKP showed variations depending on the compound series, indicating that octanol and the more structured bilayer environment responded differently to molecular shape and interactions.
One of the significant advantages of the solid-state NMR methodology employed by the team is its applicability to membranes of arbitrary complexity. This capability allows for the analysis of how properties such as headgroup size, charge, chain length, and saturation influence the partitioning of fluorinated compounds. Membranes in living organisms exhibit significant variations in lipid composition, a feature that is not reflected in octanol/water partition experiments. Therefore, the developed solid-state 19F NMR MAS methodology opens up opportunities to tailor fluorination motifs favoring partitioning into specific classes of membranes.
As a practical application of their methodology, the researchers investigated the influence of cholesterol, a common sterol in eukaryotic cell membranes, on the membrane partitioning of specific compounds. They found that increasing levels of cholesterol in lipid vesicles resulted in a reduction in the partitioning of the studied compounds into the membrane. This reduction aligned with earlier studies that observed decreased membrane partitioning for small molecules in the presence of cholesterol. The researchers attributed this effect to the disruption of acyl chain packing and interactions with cholesterol. Additionally, the authors demonstrated that when multiple compounds with well-resolved 19F chemical shifts are present in a mixture, their relative partitioning can be investigated. This measurement allows for the determination of ΔlogKp between compounds, which can be useful when studying complex lipid mixtures with unknown partial molecular volumes. The results showed excellent agreement between logKP values measured for individual compounds and those measured in equimolar mixtures, highlighting the robustness of the methodology. In conclusion, the research conducted by University of Southampton scientists presented valuable insights into the impact of aliphatic fluorination on membrane permeability, and demonstrated that the conventional octanol-water lipophilicity determination methodology is suitable to investigate lipophilicity changes by fluorination. Their innovative solid-state NMR methodology provides a powerful tool for studying the partitioning of fluorinated compounds into membranes of varying complexity. Additionally, the study advances our understanding of the complex interplay between lipophilicity and membrane permeability, ultimately aiding in the design of more effective and efficient pharmaceuticals.
Wang Z, Felstead HR, Troup RI, Linclau B, Williamson PTF. Lipophilicity Modulations by Fluorination Correlate with Membrane Partitioning. Angew Chem Int Ed Engl. 2023;62(21):e202301077. doi: 10.1002/anie.202301077.