A Hydrophobicity Scale for the Lipid Bilayer Barrier Domain from Peptide Permeabilities:  Nonadditivities in Residue Contributions

Abstract

Passive peptide transport across lipid membranes is governed by the energetics of partitioning into the ordered chain interior coupled with the rate of diffusion across this region. A hydrophobicity scale for peptide transfer into the barrier region of membranes derived from permeability coefficients would be useful to predict passive permeation of peptides across biomembranes and for determining the thermodynamics of peptide/protein insertion into the membrane interior. This study reports transport rates across large unilamellar vesicles (LUVs) composed of egg lecithin at 25 degrees C for a series of peptides having the general structure N-p-toluyl-(X)(n) (n =1-3), where X is glycine, alanine, or sarcosine. Apparent residue group contributions were calculated from permeability coefficients, P(RX), using the equation Delta(Delta G degrees )(X) = -RT ln(P(RX)/P(RH)). Multiple linear least-squares regression analysis performed for the set of 14 permeants yielded the best correlation (r(2) = 0.9993) when the following permeant descriptors were utilized: side-chain nonpolar surface area, number of -CONH- residues, number of toluyl-CON(Me)- residues, and number of other -CON(Me)- residues. The backbone -CONH- residue contribution in peptides, 4.6 kcal/mol, is significantly lower than that obtained for a single isolated -CONH- (>6 kcal/mol), suggesting a possible influence of intramolecular hydrogen bonding. Under closer scrutiny, Delta(Delta G degrees )(X) for the Ala and Gly residues decrease with increasing peptide length. The effect of N-methylation is also highly dependent on position and number of N-methyl groups on the molecule (Delta(Delta G degrees )(X) = -0.5 to -2.2 kcal/mol). These nonadditivities may be rationalized by considering the effects of peptide length and N-methylation on membrane-induced intramolecular hydrogen bonding leading to various folded conformations.

Permeability of Fluid-Phase Phospholipid Bilayers: Assessment and Useful Correlations for Permeability Screening and Other Applications

“…The most comprehensive data set for egg lecithin emerging from a single laboratory is that developed collectively in a series of papers by Xiang, Anderson, and coworkers, 16,17,26,28,29,[32][33][34][35]50 summarized in Table 1. It comprises four sets of compounds, namely: (1) p-toluic acid and a series of six related compounds with different chemical substitutions of a hydrogen atom on the "-methyl group, 29 (2) a corresponding series based on p-methylhippuric acid, 33 (3) a set of 15 other selected compounds, 26 and (4) a set of "peptidelike small organic molecules" studied most recently (including new peptide derivatives of p-toluic and pmethylhippuric acids).…”

“…It comprises four sets of compounds, namely: (1) p-toluic acid and a series of six related compounds with different chemical substitutions of a hydrogen atom on the "-methyl group, 29 (2) a corresponding series based on p-methylhippuric acid, 33 (3) a set of 15 other selected compounds, 26 and (4) a set of "peptidelike small organic molecules" studied most recently (including new peptide derivatives of p-toluic and pmethylhippuric acids). 28,35 The temperature applicable to all measurements is 25 • C. Of the 39 data points, 37 apply to pure egg lecithin, and two additional ones quantify the effects of 28.6 mol % cholesterol on permeability for p-methylhippuric acid and "-carbamoylp-methylhippuric acid. 17 In their data reduction, Xiang, Anderson, and coworkers treat carefully the existence of unionized and ionized forms of their weak acid solutes, which leads to pH dependence of apparent permeability coefficients, as well as mass transfer resistances associated with unstirred layers in diffusion cells.…”

“…26,28 The two data sets summarized in Tables 1 and 2 form the basis of our analysis. It is worth emphasizing the great value of the extensive experimental work by Xiang, Anderson, and coworkers 16,17,26,28,29,[31][32][33][34][35]39,50 embodied therein, and the strong light it sheds on P lip/w . Although significant progress has been made in understanding and correlating the trends that these data sets exhibit, 26,29,33,34,39 no single mathematical expression has yet been found that describes both of them simultaneously.…”

“…1. The bilayer permeability coefficient is given by the well-known resistances-in-series formula 16,25,26,32,34,35,41,59,62,79,80,83,84 …”

“…Application of the formula to predict P lip/w from scratch requires knowledge of an appropriate lipophilicity parameter, the 1,9-decadiene/water partition coefficient identified previously by Xiang, Anderson, and coworkers. 26,[28][29][30][31][32][33][34][35] In the absence of an experimental value, this property may be estimated in terms of K oct/w together with four of the readily available Abraham solvation parameters, 36,37 as described in a companion paper. 38 Subsequent sections draw mechanistic and practical conclusions pertinent to the prediction of fluid-phase phospholipid bilayer permeability.…”