Cultured Human Dermal Fibroblasts do Produce Cortisol

TO THE EDITOR

Known steroidogenic functions of the skin that are represented by the synthesis and metabolism of androgens and estrogens (Chen et al., 2002; Zouboulis and Degitz, 2004) probably include corticosteroids (Slominski et al., 2000; Slominski, 2005). Thus, the skin expresses both the genes and proteins involved in the synthesis of adrenal corticosteroids (Slominski et al., 1996, 1999, 2004; Rogoff et al., 2001). Metabolism of progesterone to deoxycorticosterone and corticosterone has been documented in melanoma cells and in rat skin (Slominski et al., 1999, 2000); corticotropin-releasing hormone- and ACTH-regulated cortisol production was demonstrated in histocultured human hair follicles (Ito et al., 2005); and corticotropin-releasing hormone and ACTH enhancement of cortisol production was demonstrated in epidermal melanocytes (Slominski et al., 2005b). In contrast, in human dermal fibroblasts cortisol production was not enhanced by corticotropin-releasing hormone and ACTH (there was only selective stimulation of corticosterone production; Slominski et al., 2005a). However, since the cortisol-like immunoreactivity was detected by ELISA, we investigated the chemical nature of the antigen.

Normal human adult dermal fibroblasts (from lightly pigmented female: lot 1C0317 (Cascade Biologics Inc., Portland, OR)) were cultured in Epilife medium with EpiLife Defined Growth Supplement free of serum (Cascade Biologics). The cells were washed twice with phosphate-buffered saline before incubation in supplement-free EpiLife medium with progesterone (10⁻⁶ m), ACTH (10⁻ m, but added again after 12 h), and/or IBMX (phosphodiesterase inhibitor) (10⁻ m). After 24 hours, media were extracted for cortisol identification and quantification as described previously (Slominski et al., 2005b). Cortisol in supernatants was measured with ELISA (Bio-Quant Inc., San Diego, CA), whereas chemical nature of the compounds was assessed by liquid chromatography mass spectrometry analysis with auto-mass spectrometry/mass spectrometry as described previously (Slominski et al., 2005b). Data were analyzed using Prism 4.00 (GraphPad Software), and is presented as mean + SEM (n = 4).

Testing dermal fibroblasts for the release of cortisol into the media showed that addition of progesterone (10⁻ m) stimulated significantly (4.3-fold) cortisol production; further enhancement (2.5 times) was observed after the addition of 10⁻ m IBMX (phosphodiestearse inhibitor) but not ACTH (11-fold increase over the basal values obtained from unstimulated cells; Figure 1). The addition of corticotropin-releasing hormone did not affect the production of cortisol (not shown). Liquid chromatography mass spectrometry analyses of extracted steroids showed the presence of an [M + H]⁺ ion at m/z 363 (real mass 362) with retention time of 11 minutes identical to the cortisol standard (Figure 2). Final confirmation of the product identity as cortisol was obtained with mass spectrometry/mass spectrometry analysis that yielded the same fragment ions as those of the cortisol standard: m/z = 345 ([M + H]⁺ –H₂O), m/z = 327 ([M + H]⁺ –2H₂O), and m/z = 309 ([M + H]⁺ –3H₂O) (Figure 2).

Figure 1. Effect of progesterone on cortisol synthesis.

Fibroblasts were cultured in 24-well plates seeded at density of 10⁵ cells/well. Differences between control and experimental treatments: *P<0.05.

Figure 2. Liquid chromatography mass spectrometry (LC/MS²) identifies cortisol in cultures of fibroblasts.

Fibroblasts were cultured in Petri dishes seeded at density of 2 × 10⁶ cells/dish. (a, c) Cortisol standard and (b, d) extracts from conditioned media.

These data document the capability of human dermal fibroblasts to produce cortisol molecules identified by rigorous analysis with liquid chromatography mass spectrometry². Moreover, and similar to normal (Slominski et al., 2005b) and malignant melanocytes (Slominski et al., 1999), fibroblasts also displayed a precursor substrate relationship in the accumulation of cortisol, for example, the addition of progesterone-stimulated production of the final corticosteroid cortisol, which was further enhanced by a phosphodiesterase inhibitor (IBMX). Nevertheless, in contrast to melanocytes ACTH and corticotropin-releasing hormone did not stimulate cortisol production in fibroblasts, related, perhaps, to defective coupling of MC-2R (if expressed) to intracellular signaling of the glucosteroidogenic pathway in cultured fibroblasts. Thus, these studies extend the ex vivo finding of cortisol production by hair follicles (Ito et al., 2005), adding an important component of heterogeneity to the pleiotropic expression of hypothalamic-pituitary-adrenal mediators in the skin (Slominski and Wortsman, 2000; Slominski, 2005). The role of these factors in the cutaneous response to stress remains to be fully characterized.