Summary. Nonelectrolyte partition coefficients (K's) and free energies of solution (AFt's) in dimyristoyl lecithin liposomes and in bulk nonpolar solvents were compared. Individual substituent groups tend to have consistent effects on K, permitting the extraction of incremental free energies (OAF), enthalpies (OAH), and entropies (OAS) of partition and of solution. Values of the selectivity constant s and of 6riFt for the --CHzand --OH groups in lecithin suggest that partitioned solutes are mainly located in a region slightly less hydrophobic than octanol and similar to C5I-I11OH in its solvent properties. Lecithin discriminates against branched solutes more than does a bulk solvent with the same s value. Below the endothermic phase-transition temperature (i.e., when the hydrocarbon tails "freeze"), AS and dH of partition increase 10-fold, K jumps down slightly, AS and AH of solution reverse in sign from negative to positive, and the Barclay-Butler constants become more positive. Partition in lecithin and in erythrocytes is similar, except for the absence of surface charge effects in lecithin. Resistance to nonelectrolyte permeation is inhomogeneously distributed through the bilayer, and the region of maximum partition does not provide the rate-limiting barrier. An appendix derives a simple general exPression for the nonelectrolyte permeability of a membrane that may be asymmetrical, may have position-dependent partition coefficients and diffusion coefficients, and may have significant interracial resistances. This paper, the last in a series of four, analyzes nonelectrolyte average partition coefficients that were measured between water and the sucroseexcluding space of hydrated dimyristoyl lecithin liposomes and water and reported in the preceding paper (Katz & Diamond, 1974c, referred to as paper III). As discussed in the Appendix of the present paper, to attain a detailed understanding of permeation in biological membranes and thin lipid membranes will require separating the dependence of permeability coefficients on three groups of factors: those that determine, respectively, equilibrium solute concentrations in the membrane interior, solute mobilities in the membrane interior, and interracial rate processes. Partition measurements provide information about the first of these three sets of factors,

Summary. Partition coefficients (K's) between dimyristoyl lecithin liposomes and water were measured as a function of temperature for 16 nonelectrolytes. From K and its temperature dependence, the partial molar free energy (,JF), enthalpy (,JH) and entropy (LtS) of partition for each solute were extracted. By subtraction of dE', AH and AS of hydration of each solute (where known), AF, zlH and AS of solution in lecithin could be calculated for some solutes. Among different solutes AS of solution increases approximately linearly with AH of solution, as also found in bulk solvents (the so-called Barclay-Butler relations). The Barclay-Butler slope dztS/dzlH is approximately twice as steep for lecithin as for bulk nonpolar solvents. This effect is attributed in part to greater immobilization of solutes in bilayers than in bulk solvents.The previous papers in this series described methods for measuring nonelectrolyte partition coefficients between liposomes and water, and for estimating the amount of nonsolvent water in liposomes (Katz & Diamond, 1974a, b, referred to as papers I and II, respectively). The present paper reports experimental measurements of partition coefficients between dimyristoyl lecithin liposomes and water, corrected for the effects of trapped water and nonsolvent water in the liposomes. It then extracts the following thermodynamic parameters from these measurements: partial molar free energies, enthalpies, and entropies of partition; partial molar free energies, enthalpies, and entropies of solution in lecithin from the gas phase; and the Barclay-Butler constants for solution in lecithin.

Summary. In liposomes of dimyristoyl lecithin at 40 ~ a quantity of water equal to about 11.5 moles water per mole lecithin, or about one-third of the enclosed liposome water or one-fifth of the total pellet water, behaves as if it is unavailable for dissolving sucrose. This phenomenon represents permanent exclusion of sucrose, not simply a space that equilibrates slowly due to the low permeability of sucrose. The amount of nonsolvent water increases with temperature, and is similar to the amount of water bound to the phosphorylcholine groups as estimated by other methods. Nonsolvent water arises from a combination of the forces responsible for "salting-out" of nonelectrolytes from aqueous solutions by ions, and of steric effects adjacent to a surface. Measured liposome: water partition coefficients must be corrected for the effect of nonsolvent water.

Summary. This series of papers describes methods for measuring solute partition coefficients between membrane suspensions and water, and applies these methods to the system dimyristoyl lecithin:water. Knowledge of partition coefficients is relevant to analyzing permeability coefficients, probing membrane structure, and evaluating permeation models. The theoretical section of this first paper derives equations to correct measured partition coefficients for the effects of trapped water in the precipitate, tritium exchange with lipids or membranes, and nonsolvent water adjacent to membranes. The experimental section presents methods for measuring partition coefficients in liposomes of dimyristoyl lecithin, including techniques for minimizing effects of counter drift and quenching on double-label radioactive counting. These methods offer advantages in measuring membrane:water partition coefficients that are not high.A partition coefficient (K) is the ratio, at equilibrium, of a solute's concentrations in two immiscible phases between which solute can migrate. In the present series of papers we describe methods for measuring nonelectrolyte partition coefficients between membrane suspensions and water, and we apply these methods to aqueous suspensions of dimyristoyl lecithin liposomes. The methods should also be appficable to determining partition coefficients between biological membranes and water.These measurements are relevant to at least three general problems in membrane biology. First, the solute permeability coefficient (P), as commonly measured in biological membranes, is a black-box quantity which actually depends upon the solute's membrane:water K, the solute's diffusion coefficient within the membrane D, the membrane thickness Xo, and the resistances r' and r" of the two membrane:water interfaces to the * These papers are dedicated to the memory of Aharon Katchalsky, admired teacher and friend.

No abstract