THE PHYSICAL STATE OF SOLUTES AND WATER IN LIVING CELLS ACCORDING TO THE ASSOCIATION‐INDUCTION HYPOTHESIS *

Cells shrink in concentrated aqueous solutions of sodium chloride or sucrose; they swell in more dilute solutions. Similarly, a sac made of parchment paper, dead animal bladder, or copper ferrocyanide gel, filled with an aqueous solution, may gain or lose water in such environments. On the basis of the resemblance among these osmotic behaviors, Pfeffer formulated the membrane theory a century ag0.l Membranes of the above sort-with high permeability to water and much lower permeability to alcohols, sugars, and salts-were later referred to as semipermeable membranes. Pfeffer suggested that a semipermeable membrane surrounds all living cells and is responsible for the discontinuity of the cell from its aqueous environment. At the turn of the century, Overton postulated that the plasma membrane consists of a continuous lipid layer.2, In support of this idea, Overton and, later, Collander 4 9 demonstrated a linear correlation between the solubilities in oil of nonelectrolytes ranging from monohydric alcohols to glycerol and their relative rates of penetration into living cells, both parameters covering ranges of 5 logarithmic scales. Later measurement of the surface tensions of living cells showed inconsistency with this lipid membrane theory. The measured surface tension was found to be more than one-hundredfold lower than predicted if the cell surface is A remedial suggestion was then introduced; i.e., the lipid layer is sandwiched between two hydrophilic protein layers, thus diminishing the cell surface tensi0n.O The implicit assumption was that these protein layers, themselves, would not alter the cell's permeability. A serious deficiency persists in this model even after the remedial postulation of protein covering the cell surfaces; i.e., a lipid membrane is not a bona fide semipermeable membrane. Thus, ethyl alcohol is 40 times more soluble in oil than water is.1O From the relation given by Collander between the oil/ water distribution coefficient and permeability, ethyl alcohol should be 100 times more permeant to the cell membrane than water, rather than less permeant. This is hardly acceptable, because it was the much greater permeability of pig's bladder to water than to ethyl alcohol that led Abbe Nollet to the first recorded observation of osmotic phenomena in 1748." To remedy this and other defects, Collander and Barlund 1 2 then suggested that the lipid layer had small pores through which water molecules, but not the larger alcohols and sugars, could pass.