Adaption of a dermal in vitro method to investigate the uptake of chemicals across amphibian skin

Abstract

Background: Literature data indicate that terrestrial life stages of amphibians may be more sensitive to xenobiotics than birds or mammals. It is hypothesized that dermal exposure could potentially be a significant route of exposure for amphibians, as there is evidence that their skin is more permeable than the skin of other vertebrate species. Thus, higher amounts of xenobiotics might enter systemic circulation by dermal uptake resulting in adverse effects. Heretofore, no guidelines exist to investigate dermal toxicity of chemicals to amphibians. In order to minimize vertebrate testing, this work was targeted to develop an in vitro test system as a possible model to assess the dermal uptake of chemicals across amphibian skin.

Results: The dermal absorption in vitro method (OECD guideline 428), an established toxicological (mammal) test procedure, was adapted to amphibian skin, in a first approach using the laboratory model organism Xenopus laevis and reference compounds (caffeine and testosterone). Skin permeability to both reference substances was significantly higher compared to published mammalian data. Caffeine permeated faster across the skin than testosterone, with ventral skin tending to be more permeable than dorsal skin. As usage of frozen mammalian skin is accepted, frozen skin of X. laevis was tested in parallel. To the freshly excised skin, however, freezing led to increased skin permeability, in particular to caffeine, indicating a loss of skin integrity due to freezing (without additional preservation measures).

Conclusions: This work has demonstrated that the chosen method can be applied successfully to amphibian skin, providing the basis for further investigations. In future, well-established in vitro test systems and a broad dataset for many chemicals may help assess potential amphibian risk from xenobiotics without the need for extensive vertebrate testing.

Keywords: Amphibians; Dermal absorption; In vitro method; Skin permeability; Xenopus laevis.

Fig. 1

Absorption-time profiles of caffeine and testosterone permeation across the skin of Xenopus laevis. Mean cumulative absorbed dose ± standard deviation, found in the receptor fluids plotted against time and differentiated by skin storage and body side; mean values are based on 3–5 skin samples (as specified in brackets behind indication of skin storage and side in the diagram legends), stemming from one animal [except for testosterone data of freshly excised skin due to separation of testosterone results into 4- and 8-h exposure studies; see (Additional file 2) for 4-h testosterone data and detailed information on the individual cumulative absorbed doses]; due to incorrect sampling, testosterone data after 2 h for freshly excised skin and 3 h for frozen stored skin were excluded from this diagram

Fig. 2

Comparative overview of permeability coefficients of caffeine and testosterone applied to the skin of Xenopus laevis. Permeability coefficients are arranged as boxplots, separated into the different groups fresh and frozen stored skin samples from dorsal (=d) and ventral (=v) body sides; each boxplot is based on 4–7 skin samples (as specified in brackets behind indication of skin side beneath the boxplots), stemming from two animals (for testosterone frozen stored skin samples stemmed from four frogs); dashed gray line: conferring to Marzulli et al. [26], substances may be classified into five classes for estimation of their permeation rates according to the obtained permeability coefficients (<6 × 10⁻⁶-, 6 × 10⁻⁶ to 6 × 10⁻⁵, 6 × 10⁻⁵ to 6 × 10⁻⁴, 6 × 10⁻⁴ to 6 × 10⁻³, and >6 × 10⁻³-cm/h meaning very slow, slow, moderate, fast, and very fast, respectively); ns not significant; asterisks indicate level of significance and different letters indicate significant difference calculated as described in “Methods” section. An Additional file shows individual maximum permeability coefficients in detail (see Additional file 1)

Fig. 3

Comparative overview of impedances of the skin of Xenopus laevis prior to dermal absorption experiments. Measured skin impedance data illustrated as boxplots separated into freshly excised and frozen stored skin with dorsal (=d) and ventral (=v) body sides; each boxplot is based on 9–12 skin samples (as specified in brackets behind indication of skin side beneath the boxplots), stemming from four (freshly excised skin) to six (frozen stored skin) animals each; asterisks indicate the level of significance and different letters indicate significant difference calculated as described in “Methods” section. An Additional file shows individual measured impedances in detail (see Additional file 3)