James L. Thomas

Membrane solubilization by a hydrophobic polyelectrolyte: surface activity and membrane binding

We have previously observed that the hydrophobic polyelectrolyte poly(2-ethylacrylic acid) solubilizes lipid membranes in a pH-dependent manner, and we have exploited this phenomenon to prepare lipid vesicles that release their contents in response to pH, light, or glucose (Thomas, J. L., and D. A. Tirrell. Acc. Chem. Res. 25:336-342, 1992). The physical basis for the interaction between poly(2-ethylacrylic acid) and lipid membranes has been explored using surface tensiometry and fluorimetry. Varying the polymer concentration results in changes in surface activity and membrane binding that correlate with shifts in the critical pH for membrane solubilization. Furthermore, the binding affinity is reduced as the amount of bound polymer increases. These results are consistent with a hydrophobically driven micellization process, similar to those observed with apolipoproteins, melittin, and other amphiphilic alpha-helix-based polypeptides. The absence of specific secondary structure in the synthetic polymer suggests that amphiphilicity, rather than structure, is the most important factor in membrane micellization by macromolecules.

Ultrasound Mediates the Release of Curcumin from Microemulsions

Ultrasound is a powerful noninvasive modality for biomedical imaging, and holds much promise for noninvasive drug delivery enhancement and targeting. However, the optimal design of sound sensitive carriers is still poorly understood. In this study, curcumin, an important natural antioxidant and anticancer compound, was stably entrapped into microemulsion droplets with average size 20-35 nm. To release curcumin, low frequency (40 kHz) ultrasound at an intensity of 3.8 or 9.8 W/cm2 was applied to the microemulsions, using a probe sonicator. On insonation, much of the curcumin was released from the microemulsions and formed insoluble aggregates, as evidenced by decreased UV-vis absorption at 420 nm. The initial release rate (assayed by the rate of change of absorption) was as high as 0.11 microg/s (1.87%/sec) in phosphate buffered saline solution at neutral pH, but decreased at acidic pH. Interestingly, lower curcumin loading led to a more rapid release under insonation. Measurements of emulsion droplet size implicate droplet reorganization (fusion or fission) as an important contributing mechanism for the ultrasonic release of this compound. Although cargo in microemulsions is partitioned, rather than encapsulated (as in, for example, liposomes), these new results demonstrate that microemulsion carriers are feasible for some ultrasonic drug delivery applications.

Destabilization of fatty acid-containing liposomes by polyamidoamine dendrimers

Polyamidoamine dendrimers have been shown to be very effective at transporting DNA across cell membranes in transfection experiments. To investigate the membrane interactions with dendrimers that could contribute to this efficacy, the ability of dendrimers to permeabilize lipid vesicles in suspension has been studied. Vesicles were prepared containing the self-quenching, membrane impermeant dye calcein, and were treated with dendrimers of different sizes. Increase in fluorescence was attributed to release of the dye. Membranes containing dioleoyl phosphatidylethanolamine and stearic or oleic acid, lipids with a preference for non-lamellar phases, were very susceptible to disruption by dendrimers, with larger dendrimers being more effective than smaller. However, membranes containing lipids with a preference for the lamellar phase (either pure phosphatidylcholine, or phosphatidylcholine:phosphatidylserine) were largely unaffected. The concentration dependence of the permeabilization strongly suggests an aggregation-mediated mechanism for membrane disruption. Requenching measurements using cobalt citrate showed that permeabilization did not occur uniformly among the vesicles, but rather was all-or-none, with a subpopulation of vesicles responsible for essentially all of the dye release. This is also suggestive of an aggregation-induced response. Lastly, although osmotic forces are thought to play an important role in dendrimer-mediated transfection, we observed no effect of osmotic pressure and membrane tension on the efficacy of dendrimers in solution. It is likely that, in cells, dendrimers traverse cell membranes via endosomes, and the entrapment of the dendrimer itself within the endosomal vesicle may be a key factor in its ability to cause rupture.

Polymer-induced leakage of cations from dioleoyl phosphatidylcholine and phosphatidylglycerol liposomes.

The amphipathic polyacid, poly(2-ethylacrylic acid) (PEAA) has recently been shown to form fluctuating channels in patch-clamp measurements of phospholipid bilayers ¿J.C. Chung, D.J. Gross, J.L. Thomas, D.A. Tirrell, L.R. Opsahl-Ong, Macromolecules 29 (1996) 4636-4641.. To explore this phenomenon further, we have quantified the PEAA-mediated pH-dependent release of sodium and calcium ions from phospholipid vesicles. Permeability to calcium increases linearly with polymer concentration and exponentially with decreasing pH. Permeabilization of negatively charged phosphatidylglycerol (PG) liposomes occurs to a similar extent and with a similar pH dependence to that of zwitterionic phosphatidylcholine (PC) liposomes, implying that a charge neutral species is responsible for the leakage. The pH dependence of leakage shows that the cooperative protonation of from three to five carboxylate anions is required for permeabilization. Such neutralization could result in a neutral segment of polymer chain of sufficient length to span the bilayer.

Controlled Cell Deformation Produces Defined Areas of Contact between Cells and Ligand-Coated Surfaces

AbstractA method which allows precise control of the time of initiation and the area of contact of T cells with immobilized ligands has been developed. Cells are trapped in an asymmetric film that can be quantitatively thinned by reducing the films capillary pressure. Ligands adsorbed to the base of the apparatus are forced into close contact with the cells as the air–liquid interface is drawn down. Using interference microscopy and microbeads to indicate the film height, the amount of thinning can be controlled to within 1 μm. In this study, this system was used to produce contact areas of 182 and 356 μm2 between T cells and anti-CD3 coated surfaces. These contact areas were measured using fluorescent dye exclusion microscopy. This apparatus can be used for quantitative studies of T cell activation, as is reported in Patrick et al., J. Immunol. Method. 24:97–108, 2000.

Kinetics of membrane micellization by the hydrophobic polyelectrolyte poly( 2-ethylacrylic acid)

Rates of pH-dependent micellization of multilamellar vesicles by the hydrophobic polyelectrolyte poly(2-ethylacrylic acid) (PEAA) have been measured turbidometrically. This polymer shows a strong ph-dependence in its affinity for phospholipid membranes, binding in increasing amounts as pH is lowered and ultimately solubilizing membranes to form mixed micelles (Tirrell, Takigawa and Seki (1985) Ann. N.Y. Acad. Sci. 446, 237). The rate of solubilization of dipalmitoylphosphatidylcholine (DPPC) vesicle suspensions by PEAA increases approximately linearly with reductions in pH below a threshold at pH 6.55. Interestingly, negatively-charged dipalmitoylphosphatidylglycerol membranes showed qualitatively similar behavior in the presence of PEAA, and incorporation of 10% or 20% dipalmitoylphosphatidic acid in DPPC membranes did not affect solubilization rates, demonstrating that membrane charge is not an important factor in determining micellization kinetics. Micellization of DPPC and dimyristoylphosphatidylcholine membranes occurs most rapidly at their respective gel-liquid crystalline transition temperatures (Tm); the rate enhancement is correlated with a peak in the temperature-dependent binding of a fluorescently-modified PEAA in slightly alkaline solutions in which no micellization is observed. The lateral compressibility of the membrane, which has a similar peak at Tm, is proposed to be an important determinant of the rate and extent of polymer adsorption, and consequently of the rate of micellization.

Single-Cell Measurements of Polyamidoamine Dendrimer Binding

AbstractThe binding of fluorescein-labeled polyamidoamine dendritic polymers (dendrimers) to the surface of individual fibroblast cells has been quantitatively measured using confocal fluorescence microscopy. Because the binding of these polymers is not strong, significant fluorescence from polymers in solution complicates the measurement, even using confocal optics. To overcome this difficulty, we have developed a new method employing a cell-entrapped marker dye to determine the extracellular volume in each voxel. This allows for a correct subtraction of fluorescence due to polymers in solution. We have found that, as expected, the larger dendrimers bound more tightly than the smaller generations, which may partially account for their increased efficacy in DNA transfection applications. The binding varied significantly from cell to cell, and an analysis of the cell surface area showed that this variability was not caused simply by variations in cell size.

Ultrasound-Induced Dissolution of Lipid-Coated and Uncoated Gas Bubbles

The 1.1 MHz ultrasound response of micrometer-scale perfluorobutane gas bubbles, coated with a mixture of 90 mol % saturated phospholipid (disteroylphosphatidylcholine, DSPC) or unsaturated phospholipid (dioleoylphosphatidylcholine, DOPC) and 10 mol % PEG-lipid, was studied by optical microscopy. Uncoated bubbles were also studied. Bubbles, resting buoyantly against the wall of a polystyrene cuvette, were exposed to brief pulses of ultrasound (∼200 kPa amplitude) at a repetition rate of 25 Hz; images of the bubbles were taken after every other pulse. The coating had little effect on the initial response: large (>10 μm diameter) bubbles showed no size change, while smaller bubbles rapidly shrank (or fragmented) to reach a stable or metastable diameter-ca. 2 μm for coated bubbles and 4 μm for uncoated bubbles. The coating had a significant effect on further bubble evolution: after reaching a metastable size, uncoated bubbles and DOPC-coated bubbles continued to shrink slowly and ultimately vanished entirely, while DSPC-coated bubbles did not change perceptibly during the duration of the exposure. Numerical modeling using the modified Herring equation showed that the size range in which DSPC bubbles responded does correspond well with the bubble resonance; the long-term stability of these bubbles may be related to the ability of the DSPC to form a two-dimensional solid at ambient temperature or to phase separate from the PEG-lipid.

Dependence of T cell activation on area of contact and density of a ligand-coated surface

An apparatus which allows precise control of the time of initiation and the area of contact of cells with immobilized ligands has been developed. Cells are trapped in an asymmetric film that can be quantitatively thinned, forcing the cells into close contact with ligands adsorbed on the base of the apparatus. Using microbeads to indicate the film height, the amount of thinning can be controlled to within 1 microm, producing known contact areas between cells and the ligand-coated surface. This system was used with anti-CD3-coated surfaces of different densities to examine the effect of ligand density on T cell activation, while keeping the number of ligands presented to the cells constant. T cell activation was observed individually in each cell as intracellular calcium mobilization. In these experiments both the percent of T cell activation and the rate of calcium rise were found to depend only on the number of anti-CD3 molecules presented and not on their density. This implies that the spacing between molecules is not important in the range studied, and suggests that receptor clustering to levels higher than dimers may not be necessary for induction of calcium mobilization by anti-CD3.

Affiliation in rats under stress

Rats tested in a two-compartment shuttlebox with another rat tethered in one compartment showed strong preference for the compartment with the other rat. Rats given the same test with the addition of intermittent shocks showed no preference for the other rat. This result indicates that fear does not necessarily lead to affiliation in rats and that pain-produced aggression (shocked rats often fought) is not necessarily reinforcing.