Theses Doctoral

Osmophoresis of lipid vesicles in solute gradients

Gu, Yang

Lipid membranes are semipermeable, allowing for water transport across the membrane but rejecting polar solutes and large molecules. Differences in solute concentrations across a lipid membrane lead to differences in the osmotic pressure Δπ causing water to flow towards regions of higher osmotic pressure (higher solute concentration) at speeds of v=-LpΔπ , where Lp is the hydraulic conductivity of the membrane. When a lipid vesicle is placed in a concentration gradient, it is predicted to move to regions of lower concentration due to osmotic flows across the membrane. John Anderson labeled this motion "osmophoresis'' and proposed a theory that predicts the vesicle velocity as U = -½𝛼LpRTG, where 𝛼 is the vesicle radius, RT is the thermal energy, and G=∇C is the imposed concentration gradient. The first experimental demonstration of osmophoresis reported a 1 μm/s migration velocity of DMPC vesicles in a 10 mM/mm sucrose gradient created between two dialysis tubes. However, Anderson's theory predicts that osmophoretic velocity should be four order of magnitude smaller than the observed velocity. A central goal of this Dissertation is to resolve this discrepancy either by validating Anderson's theory or by identifying relevant physics it may lack. Our central conclusion is that Anderson's theory is likely correct and experimental reports of faster osmophoresis can be attributed to convective flows---particularly, density driven flows caused by the solute gradients.

To quantify osmophoresis, we need a steady concentration gradient and micron sized lipid vesicles which are predicted to move faster than smaller vesicles and can be observed by optical microscopy. In Chapter 1, I review methods for generating concentration gradients, including glass chambers for chemotaxis assays and microfluidic devices for sustained operation. In Chapter 2, I review the methods used in this Dissertation for producing giant lipid vesicles, which include film rehydration and emulsion templating techinques. Chapter 3 describes a quantitative investigation of the convective flows induced by solute gradients within microfluidic gradient generators. Solute gradients drive fluid motions due to combinations of gravitational body forces and diffusioosmotic surface forces. I quantify and model how these flows depend on solute type, concentration gradient, device height, and solution viscosity. I describe how undesired convective flows can be mitigated by adding thickening agents to increase the solution viscosity or by density matching between high and low concentration solutions.

Chapter 4 describes experiments to measure the osmophoresis of lipid vesicles in osmotic gradients while controlling for convective flows. I use density matched sugar solutions to create 315 mM/mm gradients within a commercial gradient generator (Dunn chamber) while limiting convective flows to less than 20 nm/s. I quantify the motions of lipid vesicles and tracer particles by optical microscopy and observe that both move in a common direction at speeds of 0-10 nm/s. According to Anderson's theory, the expected osmophoretic velocity of lipid vesicles in the solute gradient we applied is 4 nm/s assuming a membrane permeability is 100 μm/s. By contrast, the previously reported motion of DMPC vesicles at speeds of 1 um/s in a 10 mM/mm gradient could not be reproduced and was likely caused by inadequate control over gradient-induced convective flows.

One strategy to enhance vesicle motion in osmotic gradients is to increase the membrane permeability---for example, by addition of water channel proteins like aquaporins. The pursuit of giant vesicles with high permeability creates a challenge for measuring the membrane permeability owing to rapid vesicle swelling/shrinking in response to osmotic shocks. Chapter 5, I demonstrate how to use photo-initiated polymerization to create a light-induced osmotic shock to vesicles. Vesicle swelling rate in response to this osmotic shock can infer the membrane permeability when the reaction is fast enough. Chapter 6, I conclude how to set up a steady solute gradient with minimal convective flows to study the slow osmophoresis and highlight some specific future directions for osmophoresis, such as enhancing osmophoretic motion by increasing membrane permeability and designing a cell sorting microfluidic device to isolate living cells based on their size and permeability.


  • thumnail for Gu_columbia_0054D_16886.pdf Gu_columbia_0054D_16886.pdf application/pdf 7.55 MB Download File

More About This Work

Academic Units
Chemical Engineering
Thesis Advisors
Bishop, Kyle J.M.
Ph.D., Columbia University
Published Here
October 6, 2021