Ca-Mediated Electroformation of Cell-Sized Lipid Vesicles

Cell-sized lipid giant unilamellar vesicles (GUVs) are formed when lipid molecules self-assemble to construct a single bilayer compartment with similar morphology to living cells. The physics of self-assembly process is only generally understood and the size distribution of GUVs tends to be very polydisperse. Herein we report a strategy for the production of controlled size distributions of GUVs by a novel mechanism dissecting the mediation ability of calcium (Ca) on the conventional electroformation of GUVs. We finely construct both of the calcium ion (Ca2+) and calcium carbonate (CaCO3) mineral adsorption layers on a lipid film surface respectively during the electroformation of GUVs. It is found that Ca2+ Slip plane polarized by alternating electric field could induce a pattern of electroosmotic flow across the surface, and thus confine the fusion and growth of GUVs to facilitate the formation of uniform GUVs. The model is further improved by directly using CaCO3 that is in situ formed on a lipid film surface, providing a GUV population with narrow polydispersity. The two models deciphers the new biological function of calcium on the birth of cell-like lipid vesicles, and thus might be potentially relevant to the construction of new model to elucidate the cellular development process.

The Removal of the CaCO 3 Mineral Layer from GUVs. The GUVs with CaCO 3 mineral layer attached was firstly grown in the electroformation chamber. For a complete removal of the mineral layer around GUVs, a 50 mM sucrose solution (5 ml) at pH 2 (adjusted by 0.1 M HCl) was exchanged into the chamber at 6 rmp by a syringe pump for two times. After that, a low frequency (5 Hz) AC field with 0.5 Vpp was input to the chamber for 25 min to release GUVs from ITO surface. Scheme S1. The experimental outline to prepare monodispersed GUVs by the calcium adsorption layer models. (1) Pre-step: a swollen multilayer lipid film with phosphate headgroups exposing outwards was formed on ITO surface by appling the AC field at 0.1 Vpp for 25 min. (2) Cycle steps: the uniform calcium adsorption layer was built on the lipid film prepared by the step (1). (3) Vpp-increased steps: the Vpp was increased to a suitable value for forming monodispersed GUVs. Scheme S2. Schematic representation of the experimental cycles for the electroformation of GUVs mediated by the model of ions adsorption layers. A. Keeping AC field off and exchanging a CaCl 2 solution into the chamber by a programmable syringe pump; B. exchanging a control solution into the chamber by a programmable syringe pump; C. a sine AC field was applied between the two ITO electrodes with an increased amplitude.
Step A, B and C were repeated as a cycle and three cycles were carried out one by one at 0.3, 0.5 and 0.8 Vpp sequentially. S2 Scheme S3. Schematic representation of the experimental cycles to fabricate the CaCO 3 array on the Ca 2+ -impregnanted lipd film surface. A). Keeping AC field off and exchanging a CaCl 2 solution into the chamber by a programmable syringe pump; B). exchanging a Na 2 CO 3 solution into the chamber by a programmable syringe pump; C). a sine AC field was applied between the two ITO electrode surfaces with an increased amplitude; D). turning off AC field and exchanging a control solution into chamber to rinse the excess ions.
Step A)-D) were repeated as a cycle and three cycles were carried out one by one at 0.3, 0.5 and 0.8 Vpp sequentially. Scheme S4. Schematic representation of the experimental cycles for the electroformation of GUVs mediated by the model of minerals adsorption layer. A). Befor assembling the chamber, a Ca(OH) 2 sol was spin-coated twice on the lipid film coated onto the ITO glass; B). a sine AC field was applied between the two ITO electrode surfaces; C). turning off the AC field and exchanging a NH 4 HCO 3 solution into the chamber by a programmable syringe pump.
Step B and C were repeated as a cycle and four cycles were carried out one by one at 0.1, 0.3, 0.5 and 0.8 Vpp sequentially. S3 Figure S1. The SEM image for the lipid film (A) and EDS mapping of Ca (B) on the lipid film-coated ITO electrode surface after exchanging the 1 mM CaCl 2 solution and the control solution for the construction of Ca 2+ adsorption layer. The Ca element mapping showed a large amount of Ca 2+ retained on the lipid film after the control solution flushing the chamber, which implied the Ca 2+ adsorption layer existed robustly on the film surface. Scale bar =200 μm. . The growing solution was 50 mM sucrose solution, and the GUVs were seen to sprout from the pattern after applying AC field for 1 hr. Figure S4. The DIC image on the Ca 2+ -impregnated lipid film after exchanging the Na 2 CO 3 and control solution into the chamber. In this case, instead of a gradual increase on the voltage power via multiple cycles, the amplitude was directly elevated to 0.8 Vpp without the use of cycles shown in Scheme S3. By this process, there was no CaCO 3 array observed.
S5 Figure S5. XRD results for the CaCO 3 fabrication on the lipid film-coated ITO surfaces after contacting the surfaces with the Ca 2+ solution at different conditions: purple curve represented the deposition of Ca 2+ on the lipid film-coated ITO surface without AC field applied; bule curve represented the deposition of Ca 2+ directly on bare ITO surface without the lipid film coated and AC field applied; red curve represented the deposition of Ca 2+ directly on bare ITO surface without the lipid film coated but with AC field applied during the deposition; black curve represented that the the deposition of Ca 2+ on the lipid film-coated ITO surface with AC field applied during the deposition. In four curves, only black curve presented the clear peaks assigned to brushite and ikaite, indicating the valitidy of our hypothesis that the formation of calcite mineral was mainly controlled by the lipid film and AC field.   Figure S9. The size distribution of GUVs mediated by Mg 2+ and Na + adsorption layers respectively. For clarity, the distribution curve obtained by the conventional electroformation without any ions adsorption layer was also included as represented by the blue line.
Where z i e is the charge of the i th ionic species, φ is the electric potential, k b is the Boltzmann's constant, T is the temperature, n i r is the Ca ion number density at the point r on the lipid film. When some points are far away enough from the surface, the corresponding electric potential φ is unaffected by the presence of charged lipid film surface, i.e. φ→0, and then by eq. (1), the ion number densities n i →n r (the Ca ion number density at point r far away enough from the surface).

Part II. Electrokinetic Theory [2, 3]
Electrokinetic theory is used to derive the distribution of Ca 2+ and resultant electroosmotic flow across the lipid film surface under AC electric field. Electrokinetic potential (ψ), the ion number densities (n j ), drift velocities (v j ), the fluid velocity (u), and the pressure (p) at every point (r) in the domain are necessary for a description of the lipid film/Ca 2+ system.
Whereλ i is the drag coefficient, z i e is the charge of the i th ionic species, k b is the Boltzmann's constant, T is the temperature, and m i is the apparent ionic mass.

Boundary condition:
Under AC electric field, Electrokinetic equations are not valid for the Stern layer due to the assumption concerning the discreetness of the charge. Therefore, we need the boundary condition at the Slip plane that is a far field type boundary condition as shown in equation: When applying the boundary condition (10) into eq. (5) and (9), we can derive eq. (11) and (12): Where V peak sin(ωt) is the applied electric field (sine wave) and n j b is the bulk number density of the j th ionic species under AC electric field. Regarding to our case, eq. (11) and (12) indicate that n(Ca 2+ ) as the number density of Ca 2+ ions and ψ(r, t) as the corresponding electric potential distributed across the lipid film surface are determined by V peak sin(ωt), which fits a sine wave distribution ( Figure S1). Such nonlinear phenomena are usually classified by origin of the concentration polarization, because Ca ions in Slip plane have non-zero time average velocities in AC electrokincetics at boundary condition. Figure S1. Schematic formation of Ca ions Slip plane under AC electric field because of concentration polarization (a sine wave function) and resultant electroosmotic flow (a circular integral) in areas of C(Ca 2+ ) min .

Maxwell stress:
Eq. (13) represents force of electroosmostic flow in areas of C(Ca 2+ ) min , which is an circular integral (as shown in Figure S1).
T is the second order Maxwell stress tensor 2