Viscoelastic properties of the human red blood cell membrane. II. Area and volume of individual red cells entering a micropipette.

Previous work demonstrated that human red cells can be drawn into cylindrical glass micropipettes of internal diameter approximately 2.0 mum without lysing. For pipettes of less than approximately 2.9 mum inside diameter, the red cell must become less spherical, that is, reduce its volume-to-area ratio. In this work measurements were made from 16-mm film records that allowed the determination of the cellular area and volume of individual erythrocytes as they were drawn into a 2.0-mum pipette with negative pressures. The results showed that the total surface area of the membrane remains constant and that the cell endures the passage into the pipette by losing volume. The volume loss was interpreted to be due to cell water and solute loss when the membrane is under stress. The loss of cell volume, rather than the stretching of the membrane, adds confirmation that although it is very deformable, the membrane is very resistant to two-dimensional strain.

Electronic measurement of red cell flow in micropipettes.

The velocity of human erythrocytes in flow-through tubes of diameter less than 5 mum is measured as a function of driving pressure. The electrical resistance in the lumen of the tube increases when a cell is present, so a pulse can be generated of length inversely proportional to velocity. In the apparatus described the pulses and driving pressure are fed to a computer which derives the correlation between pressure and velocity. Experiments confirm that the resistance to flow of erythrocytes in a narrow tube is substantially the same as that of the suspending medium. The apparatus is being used to study the effects of changes in erythrocyte deformability on flow.

Permeability of individual human erythrocytes to thiourea.

The osmotic swelling to haemolysis of individual red blood cells by isosmotic thiourea has been studied using microcine photography. 2. Crenation occurs immediately upon addition of isosmotic thiourea. The cell becomes a crenated sphere without volume decrease. 3. Subsequently, the cell volume increases linearly with time with maximum swelling occurring at about 102 sec which is 81% of the total haemolysis time. 4. At maximum swelling, the cell volume is 92% greater than the initial cell volume. This volume increase is about double that measured with other permeating substances. 5. The much larger maximum volume implies that thiourea increases the area of the cell membrane. This increase varies from 0 to 75% for individual cells, with a mean of 22%. 6. Membrane expansion varies inversely as the initial cell membrane area and cell volume (r=0-790). 7. Using the increased surface area, increased maximum volume and the swelling time, the mean permeability is calculated to be 5-52 X 10(-7) cm/sec (S.D. of mean=+/-1-19 X 10(-7) cm/sec). The distribution of permeabilities represents a normal distribution. 8. The pre-lytic potassium loss ranged from 0 to 36% with a mean value of 16-5%. This is consistent with values reported in the literature for slow haemolysis. As with other permeants the distribution is skewed towards lower values. 9. Membrane permeability of individual cells varies with the amount of membrane expansion observed. Coefficient of correlation between permeability and expansion index is 0-674. 10. There is no correlation between permeability and the reciprocal of the haemolysis time (r=-0-035). The correlation between permeability and the reciprocal of the swelling time is also poor (r=0-303), probably owing to the variability in membrane expansion by thiourea in individual cells. 11. As has been shown previously for faster permeants, the permeability coefficient cannot be calculated from the haemolysis time. Because thiourea alters the membrane area and the haemolytic volume, the coefficient cannot be calculated from the swelling time unless the changes in the membrane area are also taken into account.

The stages of osmotic haemolysis.

The haemolysis of individual human erythrocytes has been observed using an inverted microscope and cine-camera. 2. With each permeant (glycerol, propylene glycol, ethylene glycol, urea and water) haemolysis is a multistage process. The stages are swelling, popping, reduction in volume possibly accompanied by ion leakage and, finally, haemoglobin leakage. 3. The classical haemolysis time (Th) is made up of a swelling time (Tsw) and a stress time (Tst). Tst is not negligible and with the faster permeants it may occupy more than 75% of the haemolysis time. 4. The stress time can also be divided into two parts: a K+ leak time (TK) during which the cell shrinks and a time (THb) during which haemoglobin is leaving the cell. THb occupies a substantial part of Th, from 25 to 65%, and is relatively longer in fast haemolysis. 5. There is a wide spread in the permeability coefficient to glycerol in a population of erythrocytes. The distribution is compatible with a Gaussian distribution. The mean permeability is 1-79 X 10(-6) cm/sec and the S.D. is +/0 0-45 X 10(-6) cm/sec. 6. The correlation between haemolysis time and swelling time for individual erythrocytes is poor, especially for fast haemolysis. Consequently, a measure of the distribution of haemolysis time does not give a related distribution of the swelling time or of the calculated permeability for individual erythrocytes.

Geometry of the human erythrocyte. I. Effect of albumin on cell geometry.

The effects of albumin on the geometry of human erythrocytes have been studied. Individual red cells, hanging on edge from coverslips were photographed. Enlarged cell profiles were digitized using a Gradicon digitizer (Instronics Ltd., Stittsville, Ontario). Geometric parameters including diameter, area, volume, minimum cylindrical diameter, sphericity index, swelling index, maximum and minimum cell thickness, were calculated for each cell using a CDC 6400 computer. Maximum effect of human serum albumin was reached at about 1 g/liter. Studies of cell populations showed decreases in mean cell diameter of up to 6%, area 6%, and volume 15%, varying from sample to sample. The thickness of the rim was increased while that at the dimple was decreased. Studies of single cells showed that area and volume changes do not occur equally in all cells. Cells with lower sphericity indices showed larger effects. In the presence of albumin, up to 50% of the cells assumed cup-shapes (stomatocytes). These cells had smaller volumes but the same area as biconcave cells. Mechanical agitation could reversibly induce biconcave cells to assume cup shapes without area or volume changes. Experiments with de-fatted human albumins showed that the presence of bound fatty acids in varying concentrations does not alter the observed effects. Bovine serum albumin has similar effects on human erythrocytes as human serum albumin.

Viscoelastic properties of the human red blood cell membrane. I. Deformation, volume loss, and rupture of red cells in micropipettes.

Single human red blood cells suspended in buffered Ringer's solution were rapidly drawn, at recorded pressures, into glass micropipettes of diameter 0.6-3.2 mum. Cells could enter micropipettes of diameter >/= 2.9 mum with minimal pressure. In micropipettes of 0.9-2.9 mum, the pressure required increased linearly with decreasing diameter. For diameters 2.5-2.9 mum, pressures ranged up to 7 cm Hg, and the cells returned to normal biconcave shape on release. For diameters 1.9-2.5 mum, the required pressures ranged from 7 to 17 cm Hg. The released cells were crenated. In micropipettes 0.9-1.9 mum, the pressures required ranged from 17 to 34 cm Hg. The cells hemolyzed on entry. As diameter decreased from 0.9 to 0.6 mum, cells were drawn into dumbbell shapes and parts of the cells were pinched off without complete hemolysis of the cell. Using an accepted value of 138 mum(2) for the mean cell area, the mean volume of the human red cell was calculated to be 94 mum(3). Under mechanical stress, about 12% of this volume is rapidly exchangeable with the external medium. The cell volume may further decrease by 20% which is not reversible.