Itraconazole is a triazole drug widely used in the treatment of fungal infections, and it is in clinical trials for treatment of several cancers. However, the drug suffers from poor solubility, while experiments have shown that itraconazole delivery in liposome nanocarriers improves both circulation half-life and tissue distribution. The drug release mechanism from the nanocarrier is still unknown, and it depends on several factors including membrane stability against defect formation. In this work, we used molecular dynamics simulations and potential of mean force (PMF) calculations to quantify the influence of itraconazole on pore formation over lipid membranes, and we compared the effect by itraconazole with a pore-stabilizing effect by the organic solvent dimethyl sulfoxide (DMSO). According to the PMFs, both itraconazole and DMSO greatly reduce the free energy of pore formation, by up to ∼20 kJ mol −1 . However, whereas large concentrations of itraconazole of 8 mol% (relative to lipid) were required, only small concentrations of few mol% DMSO (relative to water) were sufficient to stabilize pores.In addition, itraconazole and DMSO facilitate pore formation by different mechanisms.Whereas itraconazole predominantly aids the formation of a partial defect with a locally thinned membrane, DMSO mainly stabilizes a transmembrane water needle by shielding it from the hydrophobic core. Notably, the two distinct mechanisms act cooperatively upon adding both itraconazole and DMSO to the membrane, as revealed by an additional reduction of the pore free energy. Overall, our simulations reveal molecular mechanisms and free energies of membrane pore formation by small molecules.We suggest that the stabilization of a locally thinned membrane as well as the shielding of a transmembrane water needle from the hydrophobic membrane core may be general mechanism by which amphiphilic molecules facilitate pore formation over lipid membranes at sufficient concentration.

The formation of pores over lipid membranes by the application of electric fields, termed membrane electroporation, is widely used in biotechnology and medicine to deliver drugs, vaccines, or genes into living cells. Continuum models for describing the free energy landscape of membrane electroporation have been proposed decades ago, but they have never been tested against spatially detailed atomistic models. Using molecular dynamics (MD) simulations with a recently proposed reaction coordinate, we computed potentials of mean force of pore nucleation and pore expansion in lipid membranes at various transmembrane potentials. Whereas the free energies of pore expansion are compatible with previous continuum models, the experimentally important free energy barrier of pore nucleation is at variance with established models. We trace the discrepancy to previously incorrect assumptions on the geometry of the transition state; previous continuum models assumed the presence of a membrane-spanning defect throughout the process whereas, according to the MD simulations, the transition state of pore nucleation is typically passed before a transmembrane defect has formed. A modified continuum model is presented that qualitatively agrees with the MD simulations. Using kinetics of pore opening together with transition state theory, our free energies of pore nucleation are in excellent agreement with previous experimental data.

Electric fields across lipid membranes play important roles in physiology, medicine, and biotechnology, rationalizing the wide interest in modeling transmembrane potentials in molecular dynamics simulations. Transmembrane potentials have been implemented with external electric fields or by imposing charge imbalance between the two water compartments of a stacked double-membrane system. We compare the two methods in the context of membrane electroporation, which involves a large change of membrane structure and capacitance. We show that, given that Ewald electrostatics are defined with tinfoil boundary conditions, the two methods lead to (i) identical potentials of means force (PMFs) of pore formation and expansion at various potentials, demonstrating that the two methods impose equivalent driving forces for large-scale transitions at membranes and (ii) to identical polarization of water within thin water wires or open pores, suggesting that the two methods furthermore impose equivalent local electric fields. Without tinfoil boundary conditions, effects from external fields on pore formation are spuriously suppressed or even removed. Together, our study shows that both methods, external fields and charge imbalance, are well suitable for studying large-scale transitions of lipid membranes that involve changes of membrane capacitance. However, using charge imbalance is technically more challenging for maintaining a constant transmembrane potential since it requires updating of the charges as the capacitance of the membrane changes.

Electric fields across lipid membranes play important roles in physiology, medicine, and biotechnology, rationalizing the wide interest in modeling transmembrane potentials in molecular dynamics simulations. Transmembrane potentials have been implemented with external electric fields or by imposing charge imbalance between the two water compartments of a stacked double-membrane system. We compare the two methods in the context of membrane electroporation, which involves a large change of membrane structure and capacitance. We show that, given that Ewald electrostatics are defined with tinfoil boundary conditions, the two methods lead to (i) identical potentials of mean force (PMFs) of pore formation and expansion at various potentials, demonstrating that the two methods impose equivalent driving forces for largescale transitions at membranes, and (ii) to identical polarization of water within thin water wires or open pores, suggesting that the two methods furthermore impose equivalent local electric fields. Without tinfoil boundary conditions, effects from external fields on pore formation are spuriously suppressed or even removed. Together, our study shows that both methods, external fields and charge imbalance, are well suitable for studying large-scale transitions of lipid membranes that involve changes of membrane capacitance. However, using charge imbalance is technically more challenging for maintaining a constant transmembrane potential since it requires updating of the charge imbalance as the membrane capacitance changes.