The Role of Corneocytes in Skin Transport Revised—A Combined Computational and Experimental Approach

Abstract

Purpose

To investigate mechanisms of compound–corneocyte interactions in a combined experimental and theoretical approach.

Materials and Methods

Experimental methods are presented to investigate compound–corneocyte interactions in terms of dissolution within water of hydration and protein binding and to quantify the extent of the concurrent mechanisms. Results are presented for three compounds: caffeine, flufenamic acid, and testosterone. Two compartmental stratum corneum models M1 and M2 are formulated based on experimentally determined input parameters describing the affinity to lipid, proteins and water. M1 features a homogeneous protein compartment and considers protein interactions only via intra-corneocyte water. In M2 the protein compartment is sub-divided into a cornified envelope compartment interacting with inter-cellular lipids and a keratin compartment interacting with water.

Results

For the non-protein binding caffeine the impact of the aqueous compartment on stratum corneum partitioning is overestimated but is successfully modeled after introducing a bound water fraction that is non-accessible for compound dissolution. For lipophilic, keratin binding compounds (flufenamic acid, testosterone) only M2 correctly predicts a concentration dependence of stratum corneum partition coefficients.

Conclusions

Lipophilic and hydrophilic compounds interact with corneocytes. Interactions of lipophilic compounds are probably confined to the corneocyte surface. Interactions with intracellular keratin may be limited by their low aqueous solubility.

Appendices

Appendix A: The Volume Fractions of the SC Compartments

Conventionally water, lipid or protein content within SC is expressed as weight fractions relative to the weight of dry SC, for example

$$\omega _{{\text{SC,dry}}}^{{\text{lip}}} = \frac{{w_{{\text{lip}}} }}{{w_{{\text{SC,dry}}} }}$$

(32)

For convenience and easy comparison with other authors we also use weight fractions for representation of data. However, for calculations we need volume fractions relative to the volume of dry as well as hydrated SC. This section will explain how weight and volume fractions are related. The derivation of $\varphi _{{\text{SC,hyd}}}^i$ from $\omega _{{\text{SC,dry}}}^i$ will be shown here exemplarily for $\varphi _{{\text{SC,hyd}}}^{{\text{lip}}}$ for M1 but can be done analogously for all other compartments as well as for M2. We assume dry SC to be composed of 30% w/w lipids, i.e. $\omega _{{\text{SC,dry}}}^{{\text{lip}}} = 0.3$ and 70% w/w proteins $\omega _{{\text{SC,dry}}}^{{\text{pro}}} = \omega _{{\text{SC,dry}}}^{{\text{SC,dry}}} - \omega _{{\text{SC,dry}}}^{{\text{lip}}} = 1 - \omega _{{\text{SC,dry}}}^{{\text{lip}}} = 0.7$. These are empirical values recorded in our lab in vitro for female abdominal skin of 14 different patients in 136 samples by lipid extraction of freeze-dried SC and weighing. $\omega _{{\text{SC,dry}}}^{{\text{cpe}}}$ was previously determined to be 0.07 (30) so that $\omega _{{\text{SC,dry}}}^{\ker } = \omega _{{\text{SC,dry}}}^{{\text{pro}}} - \omega _{{\text{SC,dry}}}^{{\text{cpe}}} = 0.63$. Due to the definition in Eq. 2 we obtain

$$\omega _{{\text{SC,hyd}}}^{{\text{lip}}} = \frac{{w_{{\text{lip}}} }}{{w_{{\text{SC,dry}}} + w_{{\text{aqu}}} }}$$

(33)

where the last identity results from dividing both numerator and denominator by w _SC,dry

$$\omega _{{\text{SC,hyd}}}^{{\text{lip}}} = \frac{{\frac{{w_{{\text{lip}}} }}{{w_{{\text{SC,dry}}} }}}}{{\frac{{w_{{\text{SC,dry}}} }}{{w_{{\text{SC,dry}}} }} + \frac{{w_{{\text{aqu}}} }}{{w_{{\text{SC,dry}}} }}}} = \frac{{\omega _{{\text{SC,dry}}}^{{\text{lip}}} }}{{1 + \omega _{{\text{SC,dry}}}^{{\text{aqu}}} }}$$

(34)

Therefore $\omega _{{\text{SC,hyd}}}^i$ varies depending on the extent of SC hydration $\omega _{{\text{SC,hyd}}}^{{\text{aqu}}}$ in contrast to $\omega _{{\text{SC,dry}}}^i$ which is always constant. Relating Eq. 34 to the specific densities the volume fractions of the respective compartments are calculated:

$$\varphi _{{\text{SC,hyd}}}^{{\text{lip}}} = \omega _{{\text{SC,hyd}}}^{{\text{lip}}} \frac{{\rho _{{\text{SC,hyd}}} }}{{\rho _{{\text{lip}}} }}$$

(35)

with the density of hydrated SC defined as

$$\rho _{{\text{SC,hyd}}} = \frac{1}{{\omega _{{\text{SC,dry}}}^{{\text{lip}}} /\rho _{{\text{lip}}} + \omega _{{\text{SC,dry}}}^{{\text{pro}}} /\rho _{{\text{pro}}} + \omega _{{\text{SC,dry}}}^{{\text{aqu}}} /\rho _{{\text{aqu}}} }}$$

(36)

For dry SC and SC lipids the following densities are reported in literature: ρ _SC,dry = 1.3 g/cm³, and ρ _lip = 0.973 g/cm³ (72,83). The density of SC proteins can be calculated from ρ _SC,dry as shown in Eq. 37 for M1.

$$\rho _{{\text{SC,dry}}} = \frac{{w_{{\text{SC,dry}}} }}{{V_{{\text{SC,dry}}} }} = \frac{{w_{{\text{SC,dry}}} }}{{\frac{{w_{{\text{lip}}} }}{{\rho _{{\text{lip}}} }} + \frac{{w_{{\text{pro}}} }}{{\rho _{{\text{pro}}} }}}}$$

(37)

Dividing both numerator and denominator by w _SC,dry Eq. 37 may be expressed in terms of weight fractions:

$$\rho _{{\text{SC,dry}}} = \frac{{{{w_{{\text{SC,dry}}} } \mathord{\left/{\vphantom {{w_{{\text{SC,dry}}} } {w_{{\text{SC,dry}}} }}} \right.\kern-\nulldelimiterspace} {w_{{\text{SC,dry}}} }}}}{{\frac{{{{w_{{\text{lip}}} } \mathord{\left/{\vphantom {{w_{{\text{lip}}} } {w_{{\text{SC,dry}}} }}} \right.\kern-\nulldelimiterspace} {w_{{\text{SC,dry}}} }}}}{{\rho _{{\text{lip}}} }} + \frac{{{{w_{{\text{pro}}} } \mathord{\left/{\vphantom {{w_{{\text{pro}}} } {w_{{\text{SC,dry}}} }}} \right.\kern-\nulldelimiterspace} {w_{{\text{SC,dry}}} }}}}{{\rho _{{\text{pro}}} }}}} = \frac{1}{{\frac{{\omega _{{\text{SC,dry}}}^{lip} }}{{\rho _{{\text{lip}}} }} + \frac{{\omega _{{\text{SC,dry}}}^{{\text{pro}}} }}{{\rho _{{\text{pro}}} }}}}$$

(38)

$$\rho _{{\text{pro}}} = \frac{{\omega _{{\text{SC,dry}}}^{{\text{pro}}} }}{{\frac{1}{{\rho _{{\text{SC,dry}}} }} - \frac{{\omega _{{\text{SC,dry}}}^{lip} }}{{\rho _{{\text{lip}}} }}}}$$

(39)

After reorganisation for ρ _pro the protein density was calculated as 1.52 g/cm³. This value is well within the experimentally determined range (84). We assumed ρ _ker and ρ _cpe to be equal to ρ _pro.

Appendix B: Transformation of Partition Coefficients

According to Eq. 17 the partition coefficients K _SC,dry/don and K _SC,hyd/don can be written as

$$K_{{\text{SC,}}{{{\text{dry}}} \mathord{\left/{\vphantom {{{\text{dry}}} {{\text{don}}}}} \right.\kern-\nulldelimiterspace} {{\text{don}}}}} = \frac{{c_{{\text{SC,dry}}} }}{{c_{{\text{don}}} }} = \frac{{{w \mathord{\left/{\vphantom {w {V_{{\text{SC,dry}}} }}} \right.\kern-\nulldelimiterspace} {V_{{\text{SC,dry}}} }}}}{{{w \mathord{\left/{\vphantom {w {V_{{\text{don}}} }}} \right.\kern-\nulldelimiterspace} {V_{{\text{don}}} }}}}$$

(40)

$$K_{{\text{SC,}}{{{\text{hyd}}} \mathord{\left/{\vphantom {{{\text{hyd}}} {{\text{don}}}}} \right.\kern-\nulldelimiterspace} {{\text{don}}}}} = \frac{{c_{{\text{SC,hyd}}} }}{{c_{{\text{don}}} }} = \frac{{{w \mathord{\left/{\vphantom {w {V_{{\text{SC,hyd}}} }}} \right.\kern-\nulldelimiterspace} {V_{{\text{SC,hyd}}} }}}}{{{w \mathord{\left/{\vphantom {w {V_{{\text{don}}} }}} \right.\kern-\nulldelimiterspace} {V_{{\text{don}}} }}}}$$

(41)

So that K _SC,dry/don and K _SC,hyd/don are related via

$$K_{{\text{SC,}}{{{\text{dry}}} \mathord{\left/{\vphantom {{{\text{dry}}} {{\text{don}}}}} \right.\kern-\nulldelimiterspace} {{\text{don}}}}} = K_{{\text{SC,}}{{{\text{hyd}}} \mathord{\left/{\vphantom {{{\text{hyd}}} {{\text{don}}}}} \right.\kern-\nulldelimiterspace} {{\text{don}}}}} \frac{{V_{{\text{SC,hyd}}} }}{{V_{{\text{SC,dry}}} }} = K_{{\text{SC,}}{{{\text{hyd}}} \mathord{\left/{\vphantom {{{\text{hyd}}} {{\text{don}}}}} \right.\kern-\nulldelimiterspace} {{\text{don}}}}} \frac{1}{{\varphi _{{\text{SC,hyd}}}^{{\text{SC,dry}}} }}$$

(42)