Hydrogen bonding intestinal permeability

Winiwarter, S., Ax, F., Lennemas, H., Hallberg, A., Pettersson, C., Karlen, A. Hydrogen bonding descriptors in the prediction of human in vivo intestinal permeability. J. Mol. Graph. Model. 2003, 21, 273-287... 

Instead of using the oral bioavailability of a drug, one can attempt to correlate PM values with permeability coefficients generated from in situ perfused intestinal preparations. Here, one eliminates the complexities of liver metabolism, clearance, and formulation variables. Recently, this type of in vitro-in situ correlation has been conducted using the model peptides (described previously in Section V.B.2). The permeabilities of these model peptides were determined using a perfused rat intestinal preparation which involved cannulation of the mesenteric vein (Kim et al., 1993). With this preparation, it was possible to measure both the disappearance of the peptides from the intestinal perfusate and the appearance of the peptides in the mesenteric vein. Thus, clearance values (CLapp) could be calculated for each peptide. Knowing the effective surface area of the perfused rat ileum, the CLapp values could be converted to permeability coefficients (P). When the permeability coefficients of the model peptides were plotted as a function of the lipophilicity of the peptides, as measured by partition coefficients in octanol-water, a poor correlation (r2 = 0.02) was observed. A better correlation was observed between the permeabilities of these peptides and the number of potential hydrogen bonds the peptide can make with water (r2 = 0.56,... 

Winiwarter S, Ax F, Lennernas H, Hallberg A, Pettersson C and Karlen A (2003) Hydrogen Bonding Descriptors in the Prediction of Human In Vivo Intestinal Permeability. J Mol Graph Model 21 pp 273-287. 

In addition, the value of the polar surface areas (PSA) of a given molecule (Clark and Grootenhuis 2003), defined as the molecule s surface area able to undergo van-der-Waals interaction (hydrogen bond donors and acceptor groups), is used to estimate passive diffusion. A good correlation between PSA values and actual human intestinal absorption data was found for a number of compounds (Clark 1999), with good permeability found for compounds of PSA values of <61 A2 and poor absorption at values of >140 A2. 

Based on the study of Sugano et al. (2000) and our predictive VolSurf model for this series, it can be concluded that factors like size and shape previously reported to affect paracellular permeability are indeed important to explain the local structure-permeability relationship of this chemotype. Usually, permeability via paracellular aqueous pore diffusion depends on the size of the solute and its diffusion coefficient in water. Another important factor is lipophilicity. Between intestinal absorption and both volume and lipophilicity, a negative correlation was reported for this series of thrombin inhibitors. In addition, hydrogen bonding properties and dipolarity are factors that determine... 

This interpretation, although derived for a local series of MMP inhibitors, is in qualitative agreement to models for permeability and intestinal absorption. The effects of polarity versus hydrophobicity and the importance of hydrogen bonding interactions are consistently highlighted in comparison to general absorption model but with different importance of individual descriptors which have been tailored by PLS to the problem under study. 

Novartis uses the In Silico Profiling web tool. Available properties include the octanol water partition coefficient log P, molar refractivity, flexibility index, hydrogen bonding characteristics and molecular polar surface area. Various drug properties, such as intestinal absorption, BBB permeability or Plasma Protein Binding (PPB) are calculated based on in-house models. 

---