Intracellular targets constitute a vast opportunity for drug discovery but are often difficult to modulate using small molecules, and inaccessible for biological drugs which lack cell permeability. We have studied how cell permeability can be achieved in molecules that are significantly larger and structurally more complex than traditional drugs, and which therefore are promising ligands for challenging intracellular drug targets (Over & Matsson et al., Nat Chem Biol 2016). We have identified the ability to form transient, environment-dependent intramolecular bonds (e.g., intramolecular hydrogen bonds; IMHBs) to be a key feature in cell permeability among such larger-than-usual drugs. The hypothesis is that this property allows the molecules to alternately expose polar functional groups—enabling solvation in water-filled compartments such as plasma, interstitial fluids and the cytosol—and to shield that polarity when crossing the lipidic environment of the cell membrane (Matsson et al., Adv Drug Deliv Rev 2016).
We have now identified the most common IMHB motifs in crystal structures of drug molecules, and have synthesized series of compounds around each core motif that systematically vary in polarity, lipophilicity and IMHB strength. For each compound, we have measured aqueous solubility, octanol-water partition coefficients, and permeability across model lipid membranes and cultured cell monolayers.
Here we want to use all-atom molecular dynamics to simulate the partitioning and permeability behaviour of these model compounds in relevant lipid membranes. Free energy calculations and umbrella sampling will be used to determine the energetic landscape of the compounds during the water-membrane transition. This is expected to yield new detailed information about the energy barriers that limit drug permeation across cellular membranes. Specifically, the systematic design of the model compounds will allow us to study the influence of specific substructures and of the localized charge distribution in the IMHB motif on drug-membrane interactions. In a second step, we will use Markov State models to extract kinetic permeation rate constants from the molecular dynamics simulations. The simulated rate constants will be compared to values derived from permeability experiments using kinetic modelling, providing a molecular-level rationale for our experimental observations. Systematic simulation studies of these carefully designed molecules are expected to provide new insights into how dynamic intra- and intermolecular interactions affect the rate limiting steps of membrane permeability. The results will be used in future studies to rationally design drugs that use transient intramolecular hydrogen bonding to balance aqueous solubility, membrane permeability and affinity for intracellular drug targets.