Long chain fatty acid binding to human plasma albumin.

The binding of six physiologically important long chain fatty acids to defatted human plasma albumin was measured at 37 degrees in a calcium-free Krebs-Ringer phosphate buffer, pH 7.4. The data were analyzed in terms of multiple stepwise equilibria. With the saturated acids, the magnitude of the equilibrium (association) constants, Ki, increased as the chain length increased: laurate smaller than myristate smaller than palmitate smaller than stearate. Oleate was bound more tightly than stearate; by contrast, linoleate was bound less tightly than stearate. The equilibrium constants, K1 through K12, ranged from 2.4 times 10-6 - 3.5 times 10-3 m-1 for laurate to 2.6 times 10-8 - 3.5 times 10-5 m-1 for oleate. Successive values of Ki decrease for each of the acids, indicating that major cooperative binding effects do not occur over the physiological range of fatty acid concentrations. In no case could the Ki be segregated into distinct classes, suggesting that any grouping of albumin binding sites is somewhat arbitrary. The results were inconclusive concerning whether premicellar association of unbound fatty acid occurs. Although corrections for premicellar association produced very little change in the Ki values for myristate, they raised the Ki for palmitate and stearate by 300 to 700 per cent. A sigmoidal relationship was obtained when the logarithm of Ki was plotted against chain length for the saturated fatty acids containing 6 to 18 carbon atoms, indicating that the binding energy is not simply a statistical process dependent only on the fatty acid chain length. This selectivity that albumin contributes to the binding process may be due to varying degrees of configurational adaptability of its binding sites as the fatty acid increases in length.

Partition of fatty acids.

The partition ratios of radioactive fatty acids between n-heptane and a physiological buffer at 37 degrees C were measured. The fatty acids included the saturated acids with an even number of carbons from 10 to 18 and the unsaturated acids oleic, linoleic, and linolenic. In addition, the partition ratios of decanoate, myristate, and palmitate were determined over a wide pH range. Any single plot of partition ratio vs. aqueous concentration of an acid gave a nearly straight line, a finding consistent with very little association in the aqueous phase. In the case of the acids with 16 and 18 carbon atoms, however, comparison of the constants calculated from these plots with the assumption of no aqueous phase association revealed several inconsistencies. These inconsistencies cannot be resolved completely by assuming the existence of fatty acid association in the aqueous solution. We believe that at least some of the deviations are due to the presence of trace quantities of radioactive impurities in the labeled fatty acids. For example, purification of a sample of supposedly pure [1-(14)C]myristate by a series of solvent extractions increased the partition ratio by a factor of 1.5. Although all of the observations cannot be explained by this interpretation, we believe that our studies suggest that there is no appreciable association of fatty acids under the usual physiological conditions.

Quantitative analysis of uptake of free fatty acid by mammalian cells: lauric acid and human erythrocytes.

Quantitative aspects of the binding of free fatty acid to human erythrocytes were studied by measuring the distribution of various amounts of [1-(14)C]lauric acid between washed human erythrocytes and defatted human plasma albumin. Incubations were done at 37 degrees C in an isotonic phosphate-buffered salt solution. Laurate uptake approached a steady state value within 1 hr of incubation over the range of laurate-albumin molar ratios that were tested. Uptake was due primarily to a transfer of laurate from albumin to the cell, not to incorporation of the intact laurate-albumin complex. The fatty acid binding sites of the erythrocyte are located predominantly on or within the cell membrane. The binding model which best fitted the laurate uptake data consisted of two classes of erythrocyte binding sites. This model contains a small number of sites, 2.0 x 10(-13) moles/10(6) cells, that have an average apparent association constant of 1.8 x 10(6) m(-1) for laurate. Thus, the average strength of these sites is of the same order of magnitude as the stronger laurate binding sites of albumin. The binding model also contains a relatively large number of weaker fatty acid binding sites, 1.3 x 10(-11) moles/10(6) cells, that have an average apparent association constant of 1.3 x 10(4) m(-1) for laurate. These sites are too weak to bind appreciable amounts of laurate unless the fatty acid-albumin molar ratio is elevated.