Using the relations between pellet structure and electric properties derived from the preceding paper, the responses of rabbit erythrocyte pellets to osmotic or colloidal-osmotic effects from exchanged supernatants and from electroporation were investigated. Changing the ionic strength of the supernatant, or replacing it with dextran or poly(ethylene glycol) solutions, caused changes of Rp according to the osmotic behavior of the pellet. Rp was high and ohmic before electroporation, but dropped abruptly in the first few microseconds once the transmembrane voltage exceeded the membrane breakdown potential. After the initial drop, Rp increased as a result of the reduction of intercellular space. Rp increased regardless of whether the pellets were formed before or immediately after the pulse, indicating that porated cells experienced a slow colloidal-osmotic swelling. The intercellular or intermembrane distances between cells in a pellet, as a function of osmotic, colloidal-osmotic, and centrifugal pressures used to compress rabbit erythrocyte pellets, were deduced from the Rp measurement. This offered a unique opportunity to measure the intermembrane repulsive force in a disordered system including living cells. Electrohemolysis of pelleted cells was reduced because of limited swelling by the compactness of the pellet. Electrofusion was observed when the applied voltage per pellet membrane exceeded the breakdown voltage. The fusion yield was independent of pulse length greater than 10 microseconds, because after the breakdown of membrane resistance, voltage drop across the pellet became insignificant. Replacing the supernatant with poly(ethylene glycol) or dextran solutions, or coating pellets with unporated cell layers reduced the colloidal-osmotic swelling and hemolysis, but also reduced the electrofusion yield. These manipulations can be explored to increase electroloading and electrofusion efficiencies.

Cell pellets formed by centrifugation provided a good system to study the osmotic behavior, electroporation, and interaction between cells. Rabbit erythrocyte pellets were used in this study because they were simpler than nucleated cells to model analytically. Structurally, cell pellets possessed properties of porous solid bodies and gels. Electrically, cell pellets were shown to behave as a parallel set of resistance, Rp, and capacitance, Cp. Information on pellet structures was obtained from electric measurements. The pellet resistance reflected the intercellular conductivity (porosity and gap conductivity), whereas the pellet capacitance depended mostly on membrane capacitance. The pellet resistance was more sensitive to experimental conditions. The intercellular gap distance can be derived from pellet porosity measurements, providing the cell volume and surface area were known. Rp increased and relaxed exponentially with time when centrifugation started and stopped; the cycles were reversible. When supernatants were exchanged with solutions containing hypotonic electrolytes or macromolecules (such as PEG) after the pellets were formed, complicated responses to different colloidal osmotic effects were observed. A transient decrease followed by a large increase of Rp was observed after the application of a porating electric pulse, as expected from a momentary membrane breakdown, followed by a limited colloidal-osmotic swelling of pelleted cells. The equilibrium values of Rp, Cp, pellet porosity, and intercellular distances were measured and calculated as functions of cell number, centrifugation force, and ionic strength of the exchanged supernatant. Thus, the structure and properties of cell pellets can be completely characterized by electrical measurements.

High-efficiency electrofusion between cells of different sizes was achieved by application of fusing electric pulses to cells in centrifuged pellets. Larger target cells (Chinese hamster ovary or L1210 cells) were stacked among smaller human erythrocytes or erythrocyte ghosts by sequential centrifugation at 700 g to form five-tier pellets in a specially designed centrifugation-electrofusion chamber. The membranes of erythrocytes and ghost were labeled with fluorescent membrane dye (1,1' dioctadecyl-3,3,3'3'-tetramethylindocarbocyanine (Dil)), and the contents of ghosts were loaded with water-soluble fluorescent dye (42-kDa fluorescein isothiocyanate dextran (FITC-dextran)), to monitor heterogeneous cell fusion. Fusion efficiency was assayed by the extent of either membrane dye mixing or contents (FITC-dextran) mixing with target cells. Four rectangular electric pulses at 300 V and 80 microseconds each were found to give the optimal fusion results of approximately 80% heterogeneous fusion by the content-mixing assay and approximately 95% by the membrane-dye-mixing assay. Cell viability remained greater than 80% after electrofusion. Because of the electric breakdown of cell membranes at the beginning of the pulse, the pellet resistance and hence the partial voltage across the pellet reduced rapidly during the remaining pulse time. This voltage redistribution favored the survival of fused cells. The limited colloidal-osmotic swelling of cells in pellets enhanced cell-cell contact and increased the pellet resistance after each pulse. As a result, the partial voltage across the pellet was restored when the next pulse was applied. This redistribution of pulse voltage in the pellet system permitted the breakdown of cell membranes at a lower applied voltage threshold than that required for electrofusion of cells in suspension or in dielectrophoretic cell chains. The cell viability and soluble dye retention within cells (FITC-dextran) remained at the same high levels for 3 h when the cells were incubated in respective culture media with serum at 37 degrees C. Viability and dye retention decreased significantly within 30 min when cells were incubated in phosphate-buffered saline without serum. The pellet technique was applied to form hybridomas by fusion of larger SP2/0 murine myelomas with smaller naive mouse lymphocytes. An optimum of 173 +/- 70 hypoxanthine aminopterin thymidine (HAT)-selected clones of the hybridomas was obtained from 40,000 SP2/0 cells and 1.5 x 10(6) lymphocytes used in each trial. This high-efficiency fusion technique may be adapted to mediate drug and gene transfer to target cells ex vivo as well as to form hybrid cells with limited cell sources.

Polyethylene glycol (PEG) and electrofusion were applied together in a simple and highly efficient cell fusion method. PEG (8000 M(r)) was used to bring human erythrocytes into contact, and a single 4.4 kV/cm, 80 microseconds duration pulse was applied to cell suspensions. The fusion yield (FY) is PEG concentration-dependent. A maximum FY (50%) was found at about 10% PEG. Higher PEG concentrations (> 10%) suppressed FY caused by colloid osmotic shrinkage. Morphological changes, such as colloidal osmotic swelling and shrinking, and the expanding and contraction of fusion lumen, when suspension media were changed from PBS to isotonic 15% dextran solutions, was examined by microscopy. FY was found to depend on both simple osmotic and colloidal-osmotic swelling. From the swelling behavior, we propose two types of electropores: the pre-fusion sites between cell pairs, and electropores on each individual cell connecting intracellular and extracellular space. The latter type is responsible for the colloidal osmotic swelling and shrinking of cell which, together with simple osmotic swelling, is responsible for expanding the pre-fusion sites into fusion lumens. Resealing of electropores resulted in reducing FY, but the FY can be restored by simple osmotic shock. Apparently, PEG plays two opposite roles in this fusion method; one is to promote pre-pulse and post-pulse cell-cell contact, protecting pre-fusion sites, and the other suppresses FY by colloid osmotic shrinkage of cells after pulsing, especially when high PEG concentration is used. 10% PEG 8000 represents the optimal combination of these properties.