Nonsolvent Water in Human Erythrocytes

From the ability of a concentrated suspension of human erythrocytes to regulate the pH of unbuffered, anisotonic, external media it is possible to calculate the fractional cell volume in which chloride is dissolved. The difference between this volume and the total cell water gives the nonsolvent water (for chloride) of the cell. Nonsolvent water is less than 3 % of the isotonic cell volume. The quantity of nonsolvent water per cell may increase as the cells shrink in hypertonic solutions. A significant fraction of intracellular water appears to be in a form different from bulk water; and much of this water is bound to proteins (for reviews, see Dick, 1965; Bernal, 1966; Hechter, 1965). O f pertinence to red cells, several lines of evidence indicate that hemoglobin in solution binds about 0.3 g H~O/g protein (Adair and Adair, 1936; Schwann, 1965), which amounts to about 14% of the total cell volume or 20% of the total cell water. These observations have been used to explain the departures from "ideali ty" of erythrocyte volume changes in anisotonic media. Since the cells are about 70% water by volume, and yet only about 55% of the cell appears to respond to changes in tonicity (Ponder, 1948), the difference has been ascribed to bound water which is not available for solution of the principal solutes (LeFevre, 1964; Savitz et al., 1964). There is, however, other evidence which makes it difficult to accept this conclusion. For example, measurements of Miller (1964) showed that glucose and fructose dissolve in all of the cell water over a wide range of cell volumes. Moreover, if all of the known constituents of the cell were dissolved in only 80% of the cell water, the calculated internal osmotic pressure would be significantly higher than that of the medium; to reconcile the difference, it is necessary to make further assumptions, such as binding or sequestering of cell solutes, or reduced activity coefficients of solutes, or a high internal hydrostatic pressure, etc. There is no compelling evidence to give quanti tative support to such assumptions. This paper presents a study of nonsolvent water in human erythrocytes, with nonsolvent water being defined as that amount of water which is not