Anna Seelig

Lipid conformation in model membranes and biological membranes

Protein molecules in solution or in protein crystals are characterized by rather well-defined structures in which α-helical regions, β-pleated sheets, etc., are the key features. Likewise, the double helix of nucleic acids has almost become the trademark of molecular biology as such. By contrast, the structural analysis of lipids has progressed at a relatively slow pace. The early X-ray diffraction studies by V. Luzzati and others firmly established the fact that the lipids in biological membranes are predominantly organized in bilayer structures (Luzzati, 1968). V. Luzzati was also the first to emphasize the liquid-like conformation of the hydrocarbon chains, similar to that of a liquid paraffin, yet with the average orientation of the chains perpendicular to the lipid–water interface. This liquid–crystalline bilayer is generally observed in lipid–water systems at sufficiently high temperature and water content, as well as in intact biological membranes under physiological conditions (Luzzati & Husson, 1962; Luzzati, 1968; Tardieu, Luzzati & Reman, 1973; Engelman, 1971; Shipley, 1973). In combination with thermodynamic and other spectroscopic observations these investigations culminated in the formulation of the fluid mosaic model of biological membranes (cf. Singer, 1971). However, within the limits of this model the exact nature of lipid conformation and dynamics was immaterial, the lipids were simply pictured as circles with two squiggly lines representing the polar head group and the fatty acyl chains, respectively. No attempt was made to incorporate the well-established chemical structure into this picture. Similarly, membrane proteins were visualized as smooth rotational ellipsoids disregarding the possibility that protruding amino acid side-chains and irregularities of the backbone folding may create a rather rugged protein surface.

Blood-Brain Barrier Permeation: Molecular Parameters Governing Passive Diffusion

Abstract. 53 compounds with clinically established ability to cross or not to cross the blood-brain barrier by passive diffusion were characterized by means of surface activity measurements in terms of three parameters, i.e., the air-water partition coefficient, Kaw, the critical micelle concentration, CMCD, and the cross-sectional area, AD. A three-dimensional plot in which the surface area, AD, is plotted as a function of K−1aw and CMCD shows essentially three groups of compounds: (i) very hydrophobic compounds with large air-water partition coefficients and large cross-sectional areas, AD > 80 Å2 which do not cross the blood-brain barrier, (ii) compounds with lower air-water partition coefficients and an average cross-sectional area, AD≅ 50 Å2 which easily cross the blood-brain barrier, and (iii) hydrophilic compounds with low air-water partition coefficients (AD < 50 Å2) which cross the blood-brain barrier only if applied at high concentrations. It was shown that the lipid membrane-water partition coefficient, Klw, measured previously, can be correlated with the air-water partition coefficient if the additional work against the internal lateral bilayer pressure, πbi= 34 ± 4 mN/m is taken into account. The partitioning into anisotropic lipid membranes decreases exponentially with increasing cross-sectional areas, AD, according to Klw=const. Kaw exp(−ADπbi/kT) where kT is the thermal energy. The cross-sectional area of the molecule oriented at a hydrophilic-hydrophobic interface is thus the main determinant for membrane permeation provided the molecule is surface active and has a pKa > 4 for acids and a pKa < 10 for bases.

Structure-activity relationship of P-glycoprotein substrates and modifiers

The air-water partition coefficients, K(aw), highly correlated with the corresponding lipid-water partition coefficients, K(lw), and the critical micelle concentrations, CMC, were measured for 11 compounds for which the kinetic parameters of P-glycoprotein ATPase activation (Michaelis-Menten constant, K(m), and maximal velocity, V(max)) had been determined previously in inside-out vesicles of CR1R12 Chinese hamster ovary cells. In addition, the hydrogen bond donor patterns (type I and type II) relevant for substrate recognition by P-glycoprotein were determined from the energy-minimized three-dimensional structure of these compounds. A linear relation between the air-water partition coefficient, K(aw), and the inverse of the Michaelis-Menten constant, K(m), was observed such that K(m) x K(aw) approximately = 1. The maximal velocity, V(max), was shown to decrease with the number and strength of electron donor (hydrogen bond acceptor) groups in recognition patterns. If two substrates are applied simultaneously to P-glycoprotein the compound with the higher potential to form hydrogen bonds generally acts as an inhibitor. We conclude that partitioning into the lipid membrane is the rate-limiting step for the interaction of a substrate with P-glycoprotein and that dissociation of the P-glycoprotein-substrate complex is determined by the number and strength of the hydrogen bonds formed between the substrate and the transporter.

Local anesthetics and pressure: a comparison of dibucaine binding to lipid monolayers and bilayers

The binding of the local anesthetic dibucaine to monolayers composed of 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine was studied with a Langmuir trough at pH 5.5 (22 degrees C, 0.1 M NaCl). At this pH value only the charged form of the local anesthetic exists in solution. Charged dibucaine was found to be surface active and to penetrate into the lipid monolayer, with the hydrophobic part of the molecule being accommodated between the fatty acyl chains of the lipid. The dibucaine intercalation could be quantitated by measuring the expansion of the film area, delta A, at constant surface pressure, pi. At a given surface pressure, delta A increased with increasing dibucaine in the buffer phase. On the other hand, keeping the dibucaine concentration constant, the area increase, delta A, was strongly dependent on the surface pressure. The area increase, delta A, was large at low surface pressure and decreased with increasing surface pressure. A plot of the relative change in surface area, delta A/A, versus the surface pressure yielded straight lines in the pressure range of 25-36 mN/m for five different concentrations. The delta A/A vs. pi isotherms intersected at pi = 39.5 +/- 1 mN/m with delta A = O, indicating that charged dibucaine apparently can no longer penetrate into the monolayer film. By making judicial assumptions about the area requirement of dibucaine the monolayer expansion curves could be transformed into true binding isotherms. Dibucaine binding isotherms were constructed for different monolayer pressures and were compared to a bilayer binding isotherm measured under similar conditions with ultraviolet spectroscopy. The best agreement between monolayer and bilayer binding data was obtained for a monolayer held at a pressure of 30.7 to 32.5 mN/m, which can thus be considered as the bilayer-monolayer equivalence pressure. It is further suggested from this analogy that the binding of dibucaine does not change the internal pressure in the bilayer phase, at least not in the concentration range of physiological interest (0-2 mM dibucaine) but induces a lateral expansion. At higher molar ratios of cationic dibucaine to lipid, chi b, in the monolayer (chi b greater than 0.20) the area increase is larger than would be expected from the molecular dimensions of dibucaine. This is probably due to charge repulsion effects, which at still higher molar ratios (chi b greater than 0.6) lead to a micellisation. The pressure dependence of the intercalation of cationic dibucaine into lipid membranes may also be of relevance for the phenomenon of pressure reversal in anesthesia.

Neutron diffraction studies on phosphatidylcholine model membranes: II. Chain conformation and segmental disorder

Abstract In this paper neutron diffraction experiments on 1,2-dipalmitoyl- sn -glycero-3-phosphocholine selectively deuterated in the hydrocarbon chains are reported. The experiments were carried out in the gel phase L β′ and in the liquid crystalline phase L α . The labelled segments were assumed to have a Gaussian distribution in the projection on the bilayer normal and their mean positions were derived with an accuracy of ±1 angstrom unit from a fit to the observed structure factors. The values obtained in the L β′ phase confirm the model with chains in all trans configuration tilted with respect to the bilayer normal by an angle that increases with water content. From samples that were deuterated in both chains separately and studied at low water content it was seen that the chains of the molecule are out of step by as much as 1.5 carbon-carbon bond lengths. A constant width of the label distribution in the projection on the bilayer normal was observed for segments at the beginning and end of the chains. This is an additional indication for the chains being in the all trans state in L β′ phase. In the L α phase, the present experiments show that consecutive segments are well-separated in the profile. The whole chain region is shortened by a factor of ~0.75 compared to the L β′ phase. In contrast to the gel phase, the width of the label distribution is not constant over the entire region, but is found to be increased by more than a factor of two at the end of the chains. This complements the picture derived by deuterium magnetic resonance experiments, where order parameters and correlation times of segmental motions along the chains, which essentially determine the orientational disorder and angular fluctuations of the segments, were obtained.

Substrate recognition by P-glycoprotein and the multidrug resistance-associated protein MRP1: a comparison.

OBJECTIVES It has recently been suggested that substrate recognition patterns for human P-glycoprotein encoded by mdr1 consist of two electron donor groups with a spatial separation of 2.5 +/- 0.3 A (type I units) or three electron donor groups with a spatial separation of the two outer groups of 4.6 +/- 0.6 A (type II units) [Seelig 1998]. Since P-gp and the multidrug resistance-associated protein (MRP1) have overlapping substrate specificity, we screened the chemical structures of 21 compounds, previously tested as MRP1 substrates, for electron donor units. In addition, we searched the putative transmembrane domains (TMD 1-12) of P-gp and (TMD 6-17) of MRP1 for amino acid side chains having the potential to interact with the respective substrates. METHODS The three-dimensional structures of potential MRP1 substrates were modeled with a force-field approach and were then screened for electron donor units. Helical wheel projections of the 12 putative transmembrane domains of P-gp (1-12) and MRP (6-17) were analyzed for their content of amino acid residues with hydrogen bonding side chains, charged amino acid residues, and amino acid residues with pi-electron systems. RESULTS MRP1 recognizes compounds with type I and type II units. At least one electrically neutral together with either one negatively charged type I unit or two electrically neutral type I units are required for the compound to be bound and transported. Transport increases with increasing number of electron donor units. Compounds which carry exclusively electrically neutral type I units (P-gp substrates) are transported only weakly by MRP1, and compounds with cationic type I units (P-gp substrates) are not transported at all. An analysis of the putative transmembrane alpha-helices of MRP1 and P-gp reveals that the amino acid residues with hydrogen-bond donor side chains are arranged preferentially on one side of the helix and amino acid residues with inert (non-hydrogen-bonding) side chains on the other side. In the case of MRP1, the hydrogen-bonding face also contains several cationic residues whereas, in the case of P-gp, it contains clusters of amino acid residues with beta-electron systems. CONCLUSIONS We propose that P-gp and MRP1 recognize type I or type II units in chemical compounds having diverse structures, and that these transporters bind their substrates via hydrogen bond formation. Furthermore, we propose that transport of anionic substrates by MRP1 is facilitated by cationic amino acid residues present in the transmembrane helices of MRP1, whereas the transport of cationic substrates by P-gp is facilitated by a beta-electron slide guide.

Halogenation of Drugs Enhances Membrane Binding and Permeation

Halogenation of drugs is commonly used to enhance membrane binding and permeation. We quantify the effect of replacing a hydrogen residue by a chlorine or a trifluoromethyl residue in position C‐2 of promazine, perazine, and perphenazine analogues. Moreover, we investigate the influence of the position (C‐6 and C‐7) of residue CF3 in benzopyranols. The twelve drugs are characterized by surface activity measurements, which yield the cross‐sectional area, the air–water partition coefficient, and the critical micelle concentration. By using the first two parameters (AD and Kaw) and the appropriate membrane packing density, the lipid–water partition coefficients, are calculated in excellent agreement with the lipid–water partition coefficients measured by means of isothermal titration calorimetry for small unilamellar vesicles of 1‐palmitoyl‐2‐oleoyl‐sn‐glycero‐3‐phosphocholine. Replacement of a hydrogen residue by a chlorine and a trifluoromethyl residue enhances the free energy of partitioning into the lipid membrane, on average by ΔGlw≈−1.3 or −4.5 kJ mol−1, respectively, and the permeability coefficient by a factor of ∼2 or ∼9, respectively. Despite exhibiting practically identical hydrophobicities, the two benzopyranol analogues differ in their permeability coefficients by almost an order of magnitude; this is due to their different cross‐sectional areas at the air–water and lipid–water interfaces.

High Relaxivity for Monomeric Gd(DOTA)-Based MRI Contrast Agents, Thanks to Micellar Self-Organization

A new amphiphilic GdIIIchelate, which is capable of forming micelles in aqueous solution (see diagram), has been synthesized. Due to this self-aggregation, the compound has a long rotational correlation time and, consequently, has high proton relaxivities that thus far have only been obtained with macromolecular complexes.

Quantification and characterization of P-glycoprotein-substrate interactions.

It is generally accepted that P-glycoprotein binds its substrates in the lipid phase of the membrane. Quantification and characterization of the lipid-transporter binding step are, however, still a matter of debate. We therefore selected 15 structurally diverse drugs and measured the binding constants from water to the activating (inhibitory) binding region of P-glycoprotein, K(tw(1)) (K(tw(2))), as well as the lipid-water partition coefficients, K(lw). The former were obtained by measuring the concentrations of half-maximum activation (inhibition), K(1) (K(2)), in living NIH-MDR-G185 mouse embryo fibroblasts using a Cytosensor microphysiometer, and the latter were derived from surface activity measurements. This allowed determination of the membrane concentration of drugs at half-maximum P-glycoprotein activation (C(b(1)) = (0.02 to 67) mmol/L lipid), which is much higher than the corresponding aqueous concentration (K(1) = (0.02 to 376) microM). Moreover we determined the free energy of drug binding from water to the activating binding region of the transporter (DeltaG degrees (tw(1)) = (-30 to -54) kJ/mol), the free energy of drug partitioning into the lipid membrane (DeltaG degrees (lw) = (-23 to -34) kJ/mol), and, as the difference of the two, the free energy of drug binding from the lipid membrane to the activating binding region of the transporter (DeltaG degrees (tl(1)) = (-7 to -27) kJ/mol). For the compounds tested DeltaG degrees (tl(1)) was less negative than DeltaG degrees (lw) but varied more strongly. The free energies of substrate binding to the transporter within the lipid phase, DeltaG degrees (tl(1)), are consistent with a modular binding concept, where the energetically most efficient binding module comprises two hydrogen bond acceptor groups.

Deuterium magnetic resonance studies of phospholipid bilayers

Abstract L-α-dipalmitoyl lecithin is selectively deuterated at two different chain positions. The residual quadrupole splittings of the corresponding phospholipid bilayers are measured by means of deuterium magnetic resonance and evaluated in terms of the segmental order parameters. The results are briefly compared with other esr and nmr investigations of lipid bilayers.