A systematic study of the effect of lipid architecture on cytotoxicity and cellular uptake of cationic cubosomes

HYPOTHESIS

Lipid nanoparticles containing a cationic lipid are increasingly used in drug and gene delivery as they can display improved cellular uptake, enhanced loading for anionic cargo such as siRNA and mRNA or exhibit additional functionality such as cytotoxicity against cancer cells. This research study tests the hypothesis that the molecular structure of the cationic lipid influences the structure of the lipid nanoparticle, the cellular uptake, and the resultant cytotoxicity.

EXPERIMENTS

Three potentially cytotoxic cationic lipids, with systematic variations to the hydrophobic moiety, were designed and synthesised. All the three cationic lipids synthesised contain pharmacophores such as the bicyclic coumarin group (CCA12), the tricyclic etodolac moiety (ETD12), or the large pentacyclic triterpenoid "ursolic" group (U12) conjugated to a quaternary ammonium cationic lipid containing twin C12 chains. The cationic lipids were doped into monoolein cubosomes at a range of concentrations from 0.1 mol% to 5 mol% and the effect of the lipid molecular architecture on the cubosome phase behaviour was assessed using a combination of Small Angle X-Ray Scattering (SAXS), Dynamic Light Scattering (DLS), zeta-potential and cryo-Transmission Electron Microscopy (Cryo-TEM). The resulting cytotoxicity of these particles against a range of cancerous and non-cancerous cell-lines was assessed, along with their cellular uptake.

FINDINGS

The molecular architecture of the cationic lipid was linked to the internal nanostructure of the resulting cationic cubosomes with a transition to more curved cubic and hexagonal phases generally observed. Cubosomes formed from the cationic lipid CCA12 were found to have improved cellular uptake and significantly higher cytotoxicity than the cationic lipids ETD12 and U12 against the gastric cancer cell-line (AGS) at lipid concentrations ≥ 75 µg/mL. CCA12 cationic cubosomes also displayed reasonable cytotoxicity against the prostate cancer PC-3 cell-line at lipid concentrations ≥ 100 µg/mL. In contrast, 2.5 mol% ETD12 and 2.5 mol% U12 cubosomes were generally non-toxic against both cancerous and non-cancerous cell lines over the entire concentration range tested. The molecular architecture of the cationic lipid was found to influence the cubosome phase behaviour, the cellular uptake and the toxicity although further studies are necessary to determine the exact relationship between structure and cellular uptake across a range of cell lines.

Predicting the packing parameter for lipids in monolayers with the use of molecular dynamics

Lipid molecules form the backbone of biological membranes. Due to their amphiphilic structure, they can self-organize in a plethora of different structures when in contact with water. The type of self-assembled structure and its curvature depend on so-called shape factor or critical packing parameter, CPP, that can be derived knowing the molecular volume of a lipid (V), optimal surface area (a₀) and critical chain length (l_c) (see Intermolecular and Surface Forces by Jacob N. Israelachvili, Third Edition, 2011). The value of CPP allows not only to predict the type of self-assembled structure but also is a key factor for molecular interactions, which play a great role both in physiological and pathological conditions. The greatest difficulties arise when calculating the a₀ parameter, and although for some typical membrane lipids these values have been determined, there are a number of derivatives for which this parameter, and thus CPP, are unknown. The value of CPP allows not only to predict the type of self-assembled structure but also is a key factor for molecular interactions, which play a great role both in physiological and pathological conditions. So far, the determination of the packing parameter required the use of theoretical models with assumptions deviating from the physical conditions. Here we report a method based on molecular dynamics, which was applied to simulate lipid membranes consisting of cholesterol, oxysterols, sphingolipids, phosphatidylcholines, and phosphatidylethanolamines. For lipid molecules for which CPPs have already been determined, high compliance has been demonstrated. This proves that the method presented herein can be successfully used to determine packing parameters for other membrane lipids and amphiphilic molecules.

The effect of H3O+ on the membrane morphology and hydrogen bonding of a phospholipid bilayer

At the 2017 meeting of the Australian Society for Biophysics, we presented the combined results from two recent studies showing how hydronium ions (H₃O⁺) modulate the structure and ion permeability of phospholipid bilayers. In the first study, the impact of H₃O⁺ on lipid packing had been identified using tethered bilayer lipid membranes in conjunction with electrical impedance spectroscopy and neutron reflectometry. The increased presence of H₃O⁺ (i.e. lower pH) led to a significant reduction in membrane conductivity and increased membrane thickness. A first-order explanation for the effect was assigned to alterations in the steric packing of the membrane lipids. Changes in packing were described by a critical packing parameter (CPP) related to the interfacial area and volume and shape of the membrane lipids. We proposed that increasing the concentraton of H₃O⁺ resulted in stronger hydrogen bonding between the phosphate oxygens at the water-lipid interface leading to a reduced area per lipid and slightly increased membrane thickness. At the meeting, a molecular model for these pH effects based on the result of our second study was presented. Multiple μs-long, unrestrained molecular dynamic (MD) simulations of a phosphatidylcholine lipid bilayer were carried out and showed a concentration dependent reduction in the area per lipid and an increase in bilayer thickness, in agreement with experimental data. Further, H₃O⁺ preferentially accumulated at the water-lipid interface, suggesting the localised pH at the membrane surface is much lower than the bulk bathing solution. Another significant finding was that the hydrogen bonds formed by H₃O⁺ ions with lipid headgroup oxygens are, on average, shorter in length and longer-lived than the ones formed in bulk water. In addition, the H₃O⁺ ions resided for longer periods in association with the carbonyl oxygens than with either phosphate oxygen in lipids. In summary, the MD simulations support a model where the hydrogen bonding capacity of H₃O⁺ for carbonyl and phosphate oxygens is the origin of the pH-induced changes in lipid packing in phospholipid membranes. These molecular-level studies are an important step towards a better understanding of the effect of pH on biological membranes.