Biophysical characterization of mycobacterial model membranes and their interaction with rifabutin: Towards lipid-guided drug screening in tuberculosis

Lipid structure critically dictates the molecular interactions of drugs with membranes influencing passive diffusion, drug partitioning and accumulation, thereby underpinning a lipid-composition specific interplay. Spurring selective passive drug diffusion and uptake through membranes is an obvious solution to combat growing antibiotic resistance with minimized toxicities. However, the spectrum of complex mycobacterial lipids and lack thereof of suitable membrane platforms limits the understanding of mechanisms underlying drug-membrane interactions in tuberculosis. Herein, we developed membrane scaffolds specific to mycobacterial outer membrane and demonstrate them as improvised research platforms for investigating anti-tubercular drug interactions. Combined spectroscopy and microscopy results reveal an enhanced partitioning of model drug Rifabutin in trehalose dimycolate-containing mycobacterial membrane systems. These effects are apportioned to specific changes in membrane structure, order and fluidity leading to enhanced drug interaction. These findings on the membrane biophysical consequences of drug interactions will offer valuable insights for guiding the design of more effective antibiotic drugs coupled with tuned toxicity profiles.     

Shedding light on the puzzle of drug-membrane interactions: Experimental techniques and molecular dynamics simulations

Lipid membranes work as barriers, which leads to inevitable drug-membrane interactions in vivo. These interactions affect the pharmacokinetic properties of drugs, such as their diffusion, transport, distribution, and accumulation inside the membrane. Furthermore, these interactions also affect their pharmacodynamic properties with respect to both therapeutic and toxic effects. Experimental membrane models have been used to perform in vitro assessment of the effects of drugs on the biophysical properties of membranes by employing different experimental techniques. In in silico studies, molecular dynamics simulations have been used to provide new insights at an atomistic level, which enables the study of properties that are difficult or even impossible to measure experimentally. Each model and technique has its advantages and disadvantages. Hence, combining different models and techniques is necessary for a more reliable study. In this review, the theoretical backgrounds of these (in vitro and in silico) approaches are presented, followed by a discussion of the pharmacokinetic and pharmacodynamic properties of drugs that are related to their interactions with membranes. All approaches are discussed in parallel to present for a better connection between experimental and simulation studies. Finally, an overview of the molecular dynamics simulation studies used for drug-membrane interactions is provided.     

Thermodynamics and kinetics of joint action of antiviral agent tilorone and DMSO on model lipid membranes

Individual and joint action of two water-soluble drugs, DMSO and tilorone, on model l-α-dipalmitoylphosphatidylcholine (DPPC) membranes were studied in equilibrium and kinetic regimes by differential scanning calorimetry (DSC). For equilibrium experiments, the drugs were introduced during preparation of the model membrane. In kinetic studies, one of the drugs was added to the DPPC membrane already containing the other drug, and the effects of drug-membrane interactions were monitored in real-time regime. It was found that tilorone and DMSO had opposite effects on the membrane melting temperature, which were non-additive under joint introduction of these drugs. Analysis of kinetics of DSC profiles under drugs introduction allowed us to discriminate two processes in drug-membrane interactions with different characteristic times, i.e., drug sorption onto the membrane (minutes) and drug diffusion through stacks of lipid bilayers (hours). It was established that 0.1?mol% DMSO effectively enhanced membrane penetration for tilorone with the rate of tilorone diffusion being dependent upon the scheme of drugs administration. A model was proposed describing how sorption of a dopant onto lipid membrane could affect the membrane permeability for other dopants. Conditions were determined for enhancement of membrane permeability, as it was observed for DPPC/DMSO/tilorone system.     

膜上相互作用对平板双分子层脂膜电性质的影响

以平板双分子层脂膜作为生物膜的简单模型,建立用平板双分子层脂膜电性质研究药物－生物膜相互作用的方法。研究以具有典型特征的物质－表面活性剂、自由基、金属手性配合物与平板膜的相互作用引起膜电性质的规律性改变;重组人Ｂ型血红细胞膜与溶液中抗Ｂ单克隆抗体发生特异相互作用时,膜电阻快速下降,下降的速率与加入的抗体量成正相关。在研究发生在平板膜上的典型反应的基础上,通过对膜电性质的监测和分析,从而确认平板双分     

Acetylcholinesterase: A probe for the study of antiarrhythmic drug-membrane interactions

Structural consequences of antiarrhythmic drug interaction with erythrocyte membranes were analyzed in terms of resulting changes in the activity of membrane-associated acetylcholinesterase. When enzyme inhibitory effects of drugs were compared at concentrations producing an equivalent degree of erythrocyte antihemolysis, a number of distinct groupings emerged, indicating that the molecular consequences of drug-membrane interaction are not identical for all agents examined. Differences in drug-induced acetylcholinesterase inhibition in intact erythrocytes, erythrocyte membranes and a brain synaptic membrane preparation emphasized the role of membrane structural organization in determining the functional consequences of antiarrhythmic interaction in any given system. While the inhibitory actions of lidocaine, D-600 and bretylium in intact red cells were not altered by an increased transmembrane chloride gradient, enhanced enzyme inhibition by quinidine and propranolol was observed under these conditions. The diverse perturbational actions of these membrane-stabilizing antiarrhythmics observed here may be indicative of a corresponding degree of complexity in the mechanisms whereby substances modify the potential-dependent properties of excitable tissues.     

Interaction of membrane proteins and lipids with solubilizing detergents

Detergents are indispensable in the isolation of integral membrane proteins from biological membranes to study their intrinsic structural and functional properties. Solubilization involves a number of intermediary states that can be studied by a variety of physicochemical and kinetic methods; it usually starts by destabilization of the lipid component of the membranes, a process that is accompanied by a transition of detergent binding by the membrane from a noncooperative to a cooperative interaction already below the critical micellar concentration (CMC). This leads to the formation of membrane fragments of proteins and lipids with detergent-shielded edges. In the final stage of solubilization membrane proteins are present as protomers, with the membrane inserted sectors covered by detergent. We consider in detail the nature of this interaction and conclude that in general binding as a monolayer ring, rather than as a micelle, is the most probable mechanism. This mode of interaction is supported by neutron diffraction investigations on the disposition of detergent in 3-D crystals of membrane proteins. Finally, we briefly discuss the use of techniques such as analytical ultracentrifugation, size exclusion chromatography, and mass spectrometry relevant for the structural investigation of detergent solubilized membrane proteins.     

Different distribution of daunomycin in plasma membranes from drug-sensitive and drug-resistant P388 leukemia cells.

When the anthracycline daunomycin (DNM) is incorporated into isolated plasma membranes from P388 murine leukemia cells, the drug partitions between 'deep' and 'surface' membrane domains. Such domains have been characterized on the basis of: (1) fluorescence resonance energy transfer between 1,6-diphenylhexa-1,3,5-triene or 1-[4-(trimethylamino)phenyl]-6-phenylhexa-1,3,5-triene as energy donors, which are well known in their positioning within the membrane, and daunomycin as the energy acceptor, and (2) quenching of the fluorescence of the membrane-associated drug by the water-soluble quencher iodide. The distribution of DNM between the two plasma membrane domains is different depending on the cellular phenotype. Thus, in membranes from drug-sensitive cells, DNM is preferentially confined to 'surface' domains, while in membranes from drug-resistant cells, the drug distributes more homogeneously between 'surface' and 'deep' domains. Experiments using artificial lipid vesicles suggest that differences in the relative levels of certain lipids in the plasma membranes from drug-sensitive and drug-resistant cells, namely phosphatidylserine and cholesterol, are partly responsible for the observed differences in the distribution of DNM. Since drug-membrane interactions are important in anthracycline cytotoxicity, it is possible that our observations on a different membrane distribution of daunomycin, may be related to the different sensitivity to the drug exhibited by these cells.     

Mitoxantrone changes spectrin-aminophospholipid interactions

Understanding drug-membrane and drug-membrane protein interactions would be a crucial step towards understanding the action and biological properties of anthracyclines, as the cell membrane with its integral and peripheral proteins is the first barrier encountered by these drugs. In this paper, we briefly describe mitoxantrone-monolayer and mitoxantrone-bilayer interactions, focusing on the effect of mitoxantrone on the interactions between erythroid or nonerythroid spectrin with phosphatidylethanolamine-enriched mono- and bilayers. We found that mitoxantrone markedly modifies the interaction of erythroid and nonerythroid spectrins with phosphatidylethanolamine/phosphatitydcholine (PE/PC) monolayers. The change in Δπ induced by spectrins is several-fold larger in the presence of 72?nM mitoxantrone than in its absence: spectrin/mitoxantrone complexes induced a strong compression of the monolayer. Spin-labelling experiments showed that spectrin/mitoxantrone complexes caused significant changes in the order parameter measured using a 5′-doxyl stearate probe in the bilayer, but they practically did not affect the mobility of 16′-doxyl stearate. These results indicate close-to-surface interactions/penetrations without significant effect on the mid-region of the hydrophobic core of the bilayer. The obtained apparent equilibrium dissociation constants indicated relatively similar mitoxantrone-phospholipid and mitoxantrone-spectrin (erythroid and nonerythroid) binding affinities. These results might in part, explain the effect of mitoxantrone on spectrin distribution in the living cells.     

Interaction of the cholesterol reducing agent simvastatin with zwitterionic DPPC and charged DPPG phospholipid membranes

Simvastatin is a lipid-lowering drug in the pharmaceutical group statins. Interaction of a drug with lipids may define its role in the system and be critical for its pharmacological activity. We examined the interactions of simvastatin with zwitterionic dipalmitoyl phosphatidylcholine (DPPC) and anionic dipalmitoyl phosphatidylglycerol (DPPG) multilamellar vesicles (MLVs) as a function of temperature at different simvastatin concentrations using Fourier transform infrared (FTIR) spectroscopy and differential scanning calorimetry (DSC). The FTIR results indicate that the effect of simvastatin on membrane structure and dynamics depends on the type of membrane lipids. In anionic DPPG MLVs, high simvastatin concentrations (12, 18, 24 mol%) change the position of the CH₂ antisymmetric stretching mode to lower wavenumber values, implying an ordering effect. However, in zwitterionic DPPC MLVs, high concentrations of simvastatin disorder systems both in the gel and liquid crystalline phases. Moreover, in DPPG and DPPC MLVs, simvastatin has opposite dual effects on membrane dynamics. The bandwidth of the CH₂ antisymmetric stretching modes increases in DPPG MLVs, implying an increase in the dynamics, whereas it decreases in DPPC MLVs. Simvastatin caused broadening of the phase transition peaks and formation of shoulders on the phase transition peaks in DSC curves, indicating multi-domain formations in the phospholipid membranes. Because physical features of membranes such as lipid order and fluidity may be changed with the bioactivity of drugs, opposing effects of simvastatin on the order and dynamics of neutral and charged phospholipids may be critical to deduce the action mechanism of the drug and estimate drug-membrane interactions.     

Membrane–drug interactions studied using model membrane systems

The direct interaction of drugs with the cell membrane is often neglected when drug effects are studied. Systematic investigations are hindered by the complexity of the natural membrane and model membrane systems can offer a useful alternative. Here some examples are reviewed of how model membrane architectures including vesicles, Langmuir monolayers and solid supported membranes can be used to investigate the effects of drug molecules on the membrane structure, and how these interactions can translate into effects on embedded membrane proteins.     

Mitoxantrone changes spectrin-aminophospholipid interactions

Understanding drug-membrane and drug-membrane protein interactions would be a crucial step towards understanding the action and biological properties of anthracyclines, as the cell membrane with its integral and peripheral proteins is the first barrier encountered by these drugs. In this paper, we briefly describe mitoxantrone-monolayer and mitoxantrone-bilayer interactions, focusing on the effect of mitoxantrone on the interactions between erythroid or nonerythroid spectrin with phosphatidylethanolamine-enriched mono- and bilayers. We found that mitoxantrone markedly modifies the interaction of erythroid and nonerythroid spectrins with phosphatidylethanolamine/phosphatidylcholine (PE/PC) monolayers. The change in delta pi induced by spectrins is several-fold larger in the presence of 72 nM mitoxantrone than in its absence: spectrin/mitoxantrone complexes induced a strong compression of the monolayer. Spin-labelling experiments showed that spectrin/mitoxantrone complexes caused significant changes in the order parameter measured using a 5'-doxyl stearate probe in the bilayer, but they practically did not affect the mobility of 16'-doxyl stearate. These results indicate close-to-surface interactions/penetrations without significant effect on the mid-region of the hydrophobic core of the bilayer. The obtained apparent equilibrium dissociation constants indicated relatively similar mitoxantrone-phospholipid and mitoxantrone-spectrin (erythroid and nonerythroid) binding affinities. These results might in part, explain the effect of mitoxantrone on spectrin distribution in the living cells.     

Derivative spectrophotometry as a tool for the determination of drug partition coefficients in water/dimyristoyl-L-alpha-phosphatidylglycerol (DMPG) liposomes.

The partition coefficients (K(p)) between lipid bilayers of dimyristoyl-L-alpha-phosphatidylglycerol (DMPG) unilamellar liposomes and water were determined using derivative spectrophotometry for chlordiazepoxide (benzodiazepine), isoniazid and rifampicin (tuberculostatic drugs) and dibucaine (local anaesthetic). A comparison of the K(p) values in water/DMPG with those in water/DMPC (dimyristoyl-L-alpha-phosphatidylcholine) revealed that for chlordiazepoxide and isoniazid, neutral drugs at physiological pH, the partition coefficients are similar in anionic (DMPG) and zwitterionic (DMPC) liposomes. However, for ionised drugs at physiological pH, the electrostatic interactions are different with DMPG and DMPC, with the cationic dibucaine having a stronger interaction with DMPG, and the anionic rifampicin having a much larger K(p) in zwitterionic DMPC. These results show that liposomes are a better model membrane than an isotropic two-phase solvent system, such as water-octanol, to predict drug-membrane partition coefficients, as they mimic better the hydrophobic part and the outer polar charged surface of the phospholipids of natural membranes.     

Membrane effects of trifluoperazine, dibucaine and praziquantel on human erythrocytes

Trifluoperazine (TFP) is a potent antipsychotic agent, dibucaine (DBC) is a local anaesthetic and praziquantel (PZQ) is a highly effective agent against schistosomiasis. The present work was conducted to (i) investigate the cytotoxic effects of TFP, DBC and PZQ on human erythrocyte membranes; and (ii) compare the alterations induced by the cationic drugs (TFP and DBC) with those induced by the uncharged compound (PZQ), in an attempt to have a better insight on the pathways of each drug-membrane interaction. The erythrocyte morphological alterations induced by sublytic concentrations of TFP, DBC and PZQ were evaluated by scanning electron microscopy and expressed quantitatively by the morphological index. Haemolysis and release of membrane lipids (phospholipids and cholesterol) produced by selected concentrations of TFP, DBC and PZQ, were compared with those resulting from the corresponding triple concentrations of each drug. Our results showed that the uncharged molecule of PZQ induces the same morphological alterations (stomatocytosis) as the cationic drugs TFP and DBC. Haemolysis was shown to vary with the drug used and to be concentration-dependent, with values approximately 10-fold more elevated for TFP and DBC than for PZQ, which revealed a maximum of 6% haemolysis for the highest concentration tested. Different concentration-response curves were obtained for lipid elution, although the profiles of cholesterol and phospholipids released were similar for all drugs. Nevertheless, at a fixed rate of 50% haemolysis, TFP induced a approximately 2-fold increment in the elution of cholesterol when compared with that produced by DBC (P<0. 05). The different effects induced by TFP, DBC and PZQ on erythrocyte morphology, haemolysis and lipid exfoliation are related to the physical and chemical characteristics of each compound. These results suggest that distinct cell membrane interaction pathways lead to drug-specific mechanisms of cytotoxicity.     

Enantiomers of phospholipids and cholesterol: A key to decipher lipid-lipid interplay in membrane

Most phospholipids constituting biological membranes are chiral molecules with a hydrophilic head group and hydrophobic alkyl chains, rendering biphasic property characteristic of membrane lipids. Some lipids assemble into small domains via chirality-dependent homophilic and heterophilic interactions, the latter of which sometimes include cholesterol to form lipid rafts and other microdomains. On the other hand, lipid mediators and hormones derived from chiral lipids are recognized by specific membrane or nuclear receptors to induce downstream signaling. It is crucial to clarify the physicochemical properties of the lipid self-assembly for the study of the functions and behavior of biological membranes, which often become elusive due to effects of membrane proteins and other biological events. Three major lipids with different skeletal structures were discussed: sphingolipids including ceramides, phosphoglycerolipids, and cholesterol. The physicochemical properties of membranes and physiological functions of lipid enantiomers and diastereomers were described in comparison to natural lipids. When each enantiomer formed a self-assembly or interacted with achiral lipids, both lipid enantiomers exhibited identical membrane physicochemical properties, while when the enantiomer interacted with chiral lipids or with the opposite enantiomer, mixed membranes exhibited different properties. For example, racemic membranes comprising native sphingomyelin and its antipode exhibited phase segregation due to their strong homophilic interactions. Therefore, lipid enantiomers and diastereomers can be good probes to investigate stereospecific lipid-lipid and lipid-protein interactions occurring in biological membranes.     

Effect of sodium bicarbonate as a pharmaceutical formulation excipient on the interaction of fluvastatin with membrane phospholipids

Excipients in the pharmaceutical formulation of oral drugs are notably employed to improve drug stability. However, they can affect drug absorption and bioavailability. Passive transport through intestinal cell walls is the main absorption mechanism of drugs and, thus, involves an interaction with the membrane lipids. Therefore in this work, the effect of the excipient NaHCO₃ on the interaction of the anticholesterolemic drug fluvastatin sodium (FS) with membrane phospholipids was investigated by ¹H NMR and FTIR spectroscopy. Sodium bicarbonate is often combined with fluvastatin for oral delivery to prevent its degradation. We have used model DMPC/DMPS membranes to mimic the phospholipid content of gut cell membranes. The results presented in this work show a 100% affinity of FS for the membrane phospholipids that is not modified by the presence of the excipient. However, NaHCO₃ is shown to change the interaction mechanism of the drug. According to our data, FS enters the DMPC/DMPS bilayer interface by interacting with the lipids’ polar headgroups and burying its aromatic moieties into the apolar core. Moreover, lipid segregation takes place between the anionic and zwitterionic lipids in the membranes due to a preferential interaction of FS with phosphatidylserines. The excipient counteracts this favored interaction without affecting the drug affinity and location in the bilayer. This work illustrates that preferential interactions with lipids can be involved in passive drug permeation mechanisms and gives evidence of a possible nonpassive role of certain excipients in the interaction of drugs with membrane lipids.     

The dissimilar interactions of the calcium antagonist flunarizine with different phospholipid classes and molecular species: a differential scanning calorimetry study

The influence of the class IV calcium antagonist flunarizine on the phase behaviour of different species of the major phospholipid classes of mammalian plasma membranes has been examined using differential scanning calorimetry. We show that it has the ability to substantially influence the phase behaviour of phospholipids. Flunarizine significantly influences the gel to liquid-crystalline transition temperature of phosphatidylserines whilst having little effect on those of the phosphatidylethanolamines tested. The liquid-crystalline to inverted hexagonal phase transition of phosphatidylethanolamines is, however, strongly induced by the presence of flunarizine. Examination of the effect of flunarizine on the phase behaviour of different phosphatidylcholine species revealed an acyl-chain dependent influence. Dissimilar results with phosphatidylcholines, phosphatidylethanolamines and phosphatidylserines reveal different locations and ionization states for the drug in the different phospholipid bilayers. These results not only indicate an essential role for the ionization state of the drug in determining drug-phospholipid interactions but also the role of the phospholipid in determining the ionization state of the drug and have important implications for drug-membrane interactions demonstrating that drug interaction with one phospholipid may bear no relation whatsoever to its interaction with another.     

Lipid membranes with grafted polymers: physicochemical aspects

Membranes grafted with water-soluble polymers resist protein adsorption and adhesion to cellular surfaces. Liposomes with surface-grafted polymers therefore find applications in drug delivery. The physicochemical properties of polymer-grafted lipid membranes are reviewed with mean-field and scaling theories from polymer physics. Topics covered are: mushroom-brush transitions, membrane expansion and elasticity, bilayer-micelle transitions, membrane-membrane interactions and protein-membrane interactions.     

Cooperativity of phospholipid reorganization upon interaction of dipyridamole with surface monolayers on water

Results from various surface sensitive characterization techniques suggest a model for the interaction of the piperidinopyrimidine dipyridamole (DIP)--known as a vasodilator and inhibitor of P-glycoprotein associated multidrug resistance of tumor cells--with phospholipid monolayers in which the drug is peripherally associated with the membrane, binding (up to) five phospholipids at a time. These multiple interactions are responsible for a very strong association of the drug with the lipid monolayer even at exceedingly low concentrations (approximately 0.2 mol%). Electrostatic interactions and hydrogen bonding are likely involved in the binding of DIP to DPPC. Cooperative effects among the lipids are invoked to explain the macroscopically measurable changes of lipid monolayer properties even when only one out of 100 DPPC molecules is directly associated with a DIP molecule. A reversal of the observed changes upon drug association with the membrane as the DIP concentration surpasses a threshold concentration (c(crit)approximately 0.5 mol%) may be explained by cooperativity in a different context, the self-aggregation of drug molecules. With its implications for the interaction of DIP with phospholipid films, this work provides a first approach to the explanation of the high sensitivity of cell membranes to piperidinopyrimidine drugs on a molecular level.     

Combining molecular modeling with experimental methodologies: mechanism of membrane permeation and accumulation of ofloxacin

The interaction between ofloxacin, as a model drug of the fluoroquinolone class, and biomembranes was examined as the possible initial step in a transmembrane diffusion process. Dipalmitoylphosphatidylcholine was used for the preparation of biomembrane models. The influence of environmental conditions and protonation on molecular physicochemical behavior, and hence on the membrane interaction, was investigated by differential scanning calorimetry (DSC). This technique has been shown to be very effective in the interpretation of interactions of drug microspeciations with biomembranes. These findings suggest that the interaction occurred owing to ionic and hydrophobic forces showing how the passage through the membrane is mainly favored in the pH interval 6–7.4. It was demonstrated that a pH gradient through model membranes may be responsible for a poorly homogeneous distribution of ofloxacin (or other related fluoroquinolones), which justifies the in vivo accumulation properties of this drug. DSC experiments, which are in agreement with computational data, also showed that the complexing capability of ofloxacin with regard to Mg⁺⁺ or Ca⁺⁺ may govern the drug entrance into bacterial cells before the DNA Girase inhibition and could ensure the formation of hydrophobic and more fluid phospholipid domains on the surface of the model membrane. These regions are more permeable with regard to various solutes, as well as ofloxacin, allowing a so-called ‘self-promoted entrance pathway’. The combination of experimental methodologies with computational data allowed a further rationalization of the results and opened new perspectives into the mechanism of action of ofloxacin, namely its interaction with lipid bilayers and drug–divalent cation complex formation, which might be extended to the entire fluoroquinolone class. Ofloxacin accumulation within Escherichia coli ATCC 25922 was measured as a function of time. Also in this example, the environmental conditions influenced ofloxacin penetration and accumulation. The in vitro experiments, reported here, show that a suitable balance of hydrophilic and hydrophobic fluoroquinolone properties needs to occur for there to be increased drug permeation.