GIANT VESICLES AS BIOCHEMICAL COMPARTMENTS - THE USE OF MICROINJECTION TECHNIQUES

Detailed experimental procedures are described for the preparation of thin-walled giant phosphatidylcholine vesicles, which are useful for m icroinjections. In these microinjection experiments, a target vesicle (typically about 50 to 100 mu m in diameter) was punctured by a micron eedle and an aqueous solution was injected into the internal volume of the vesicle. The method, which was used for giant vesicle preparation , is a modification of the so-called electroformation method, original ly described by Angelova and Dimitrov (Faraday Discuss. Chem. Sec. 198 8, 81, 303-311 and 345-349). With this method, the vesicles grow in an investigation chamber at a platinum wire in an aqueous medium with th e help of an alternating electric field, and we have investigated how the experimental parameters (in particular applied voltage and frequen cy and ionic strength of the aqueous medium) influence the vesicle for mation process. Using a specially constructed investigation chamber an d 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine (POPC) as lipid, th e applied voltage was varied between 0.6 and 10 V, holding the frequen cy constant at 10 Hz. At voltages <5 V, the giant vesicles formed ofte n appeared under the microscope as nonspherical (''cut spheres'') and open, ''mushroom-like'' structures. Often, however, nonsphericity was only an optical artifact, and closed vesicles could be distinguished f rom open structures by microinjecting fluorescent dye molecules, which in the case of an open structure immediately leaked out. At voltages >5 V, closed structures were observed. At constant voltage (1.3 V), '' cut spheres'' and ''mushroom'' type structures appeared mainly in the frequency range 10-100 Hz. Between 0.2 and 2 Hz, mainly closed structu res were formed. Typical conditions for vesicle formation useful for m icroinjections were 2 V and 10 Hz. Occasionally, giant vesicles with d iameters of up to 300 mu m formed. The presence of high salt concentra tions prevented the formation of giant vesicles; in the case of LiCl, NaCl, or KCl, the limiting concentration was 10 mM, while the maximal concentrations for MgCl2, CoCl2, and CaCl2 were 1.7, 1.0, and 0.2 mM, respectively. The giant vesicles formed were osmotically sensitive. Ad dition of glucose led to a vesicle shrinkage, the beginning of visible shrinkage being dependent on the glucose concentration, ranging from 6 to 7 min (with 10 nM glucose) to 30 s (with 200 mM glucose). Giant v esicles could also be formed with mixtures of POPC and phospholiponucl eosides, such as 5'-(1,2-dioleoyl-sn-glycero-3-phospho)uridine (DOP-ur idine) or 5'-(n-hexadecylphospho)uridine (HDP-uridine); or mixtures of POPC with yl-2-oleoyl-sn-glycero-3[phospho-rac-(1-glycerol)] (POPG), 1,2-dipalmitoyl-rac-glycero-3-phosphoethanolamine (DPPE), or didodecyl dimethylammonium bromide (DDAB). No stable thin-walled vesicles formed with pure POPG, bovine brain phosphatidylserine, or pure phospholipon ucleosides. Although apparently limited to a certain class of phosphol ipids (phosphatidylcholines) or to lipid mixtures containing phosphati dylcholines, and limited to low ionic strength solutions, the electrof ormation method has proven to be the method of choice for successful m icroinjections, finally allowing to use the aqueous interior of indivi dual vesicles as microreactors with volumes of about 50-100 pt. Succes sful microinjections into giant POPC vesicles were demonstrated by usi ng calcein as fluorescent probe or DNA that was stained with YO-PRO-1. In some cases, even multiple puncturing of the same vesicle was possi ble.