Hydrophobicity

Hydrophobicity is the tendency of a chemical group to be excluded from water, driven less by attraction between non-polar surfaces than by the entropic cost water pays to order itself around them. It is quantified in several incompatible ways: octanol-water partition coefficients for whole molecules, residue scales such as Kyte-Doolittle or Eisenberg for sequences, and, most usefully for peptides, empirical retention time on a reversed-phase column.

The residue ranking is consistent across scales at the extremes, with tryptophan, isoleucine, leucine and phenylalanine at the hydrophobic end and arginine, lysine, aspartate and glutamate at the polar end. Deliberate hydrophobicity is a design tool: the C16 and C18 fatty acids on liraglutide, semaglutide and the long-acting insulins create a lipophilic patch that binds albumin reversibly. The same property creates handling problems, since hydrophobic peptides adsorb to glass and plastic surfaces.

Hydrophobicity is probably the best single predictor of the difficulties a peptide will present. It sets solubility and the pH or cosolvent needed to reach it, predicts aggregation propensity, determines chromatographic method development, and governs how much material is lost to container surfaces at low concentration. When a sequence is both hydrophobic and aggregation-prone, formulation rather than pharmacology tends to dominate development.

The inference that does not follow is that lipophilicity buys membrane crossing. For small molecules there is a rough relationship between logP and passive permeation, but for peptides size, hydrogen-bond donor count, charge and active efflux dominate, and a calculated logP for a molecule of several kilodaltons carries little predictive content. A claim that a peptide reaches the brain because it is lipophilic, without transport or tissue data, reasons from the wrong model.