Bicontinuous Cubic Phase of Monoolein and Water as Medium for Electrophoresis of Both Membrane-Bound Probes and DNA - Oligonucleotides

Bicontinuous Cubic Phase of Monoolein and Water as Medium for Electrophoresis of Both Membrane-Bound Probes and DNA
Received July 29, 2005
In Final Form: January 26, 2006
Web Release Date: March 29, 2006
Nils Carlsson, Nima Sanandaji, Marina Voinova, and Bjàrn ?kerman*
Langmuir
ACS Publications
Copyright ? 2006 American Chemical Society
Department of Chemistry and Bioscience, and Department of Applied Physics, Chalmers University of Technology, SE-41296 Gàteborg, Sweden
Abstract:
Porous hydrogels such as agarose are commonly used to analyze DNA and water-soluble proteins by electrophoresis. However, the hydrophilic environment of these gels is not suitable for separation of important amphiphilic molecules such as native membrane proteins. We show that an amphiphilic liquid crystal of the lipid monoolein and water can be used as a medium for electrophoresis of amphiphilic molecules. In fact, both membrane-bound fluorescent probes and water-soluble oligonucleotides can migrate through the same bicontinuous cubic crystal because both the lipid membrane and the aqueous phase are continuous. Both types of analytes exhibit a field-independent electrophoretic mobility, which suggests that the lipid crystal structure is not perturbed by their migration. Diffusion studies with four membrane probes indicate that membrane-bound analytes experience a friction in the cubic phase that increases with increasing size of the hydrophilic headgroup, while the size of the membrane-anchoring part has comparatively small effect on the retardation.
Introduction
Electric-field-driven migration in porous gels is commonly used to separate DNA and proteins in order to measure a wide range of properties such as monomer sequence, molecular weight, and isoelectric point.1 Agarose and polyacrylamide gels are useful polymeric networks for such purposes because the electrophoretic transport through their aqueous pores of micrometer to nanometer dimensions causes size-separation by filtration (sieving).2 A large protein or DNA molecule migrates with lower velocity than a smaller one, and molecular mass can be determined by comparison with a set of homologous size standards. However, a severe limitation with the hydrophilic gels (containing typically 99% water) is that they are usually not compatible with important amphiphilic molecules such as native membrane proteins. Protein solubilization by use of detergents is one approach, but the appropriate surfactant has to be found for each protein more or less by trial and error3 and the surfactant tends to affect the gel velocity of the solubilized protein4,5 and hence the apparent mass compared to size standards. One solution may be to run electrophoresis of native proteins in planar lipid membranes, and separation of a peripheral-type of membrane proteins anchored to a surface-supported bilayer has indeed been demonstrated.6 However, for integral membrane proteins that span the membrane, lipid drag on the hydrophobic part is relative insensitive to protein size as predicted theoretically.7 Second, in the planar case, the aqueous drag contributes little to the total friction even when the hydrophilic part constitutes a major part of the protein,8 as expected from the viscosity of the lipid layer being about 100 times higher than for water.9 The electrophoretic velocity in a planar membrane can therefore not be expected to be sensitive to overall membrane protein size.
In an attempt to retain the possibility of membrane-bound migration but with an enhanced hydrophilic drag compared to planar membranes, we have explored the potential of running electrophoresis in a bicontinuous cubic phase formed by the lipid monoolein and water10 (Figure 1a). This liquid-crystalline structure contains a three-dimensional system of water-filled pores with a diameter of about 5 nm.11 It is created by a continuous and highly curved lipid bilayer, which is able to accommodate at least certain integral membrane proteins.11,12 Our two-part working hypothesis is that if (1) membrane-bound molecules can migrate along the bilayer then (2) the aqueous pores may be of the right size to cause sieving on their hydrophilic parts, analogous to hydrophil electrophoresis in conventional gels.
--------------------------------------------------------------------------------
Figure 1 Schematic structures of the porous media used in this work. (a) Diamond cubic phase (Q224), with the aqueous channels (grey circles) surrounded by lipid bilayers. (b) Schematic representation of a porous agarose network of randomly oriented gel fibers.
--------------------------------------------------------------------------------
We have taken the approach of investigating these two conditions individually, by separating the membrane-anchoring function and the hydrophilic parts in amphiphilic analytes. Previously, we have used hydrophilic oligonucleotides of well-characterized nanometer sizes13 comparable to hydrophilic parts of membrane proteins and showed14 that the aqueous pores indeed are of suitable size and nature to cause size-selective sieving. Here we study membrane-anchored fluorescent probes (Chart 1) with hydrophilic parts which are small compared to most membrane proteins. With the aim to establish the mode of migration (rather than separation), we show that the continuous nature of the lipid phase allows membrane probes to be transported by electrophoresis in the same cubic phase that causes sieving of oligonucleotides. As control for conventional hydrogels, we use a hydroxyethylated agarose gel (Figure 1b) which has similar average pore size as the monoolein cubic phase.
--------------------------------------------------------------------------------
Chart 1. Membrane Probes DiI-C18-(5)-Ds (DiI), DiOC18 (DiO), BODIPY-C5-HPA (C5-HPA), and [Ru(phen)2Me2Dppz]2+
Complete article may be viewed online.