Skin diseases affect patients of all ages worldwide. By some estimates, they affect 50 percent or more of the population at any one time and comprise over 2,000 medical conditions which can range from only mild skin problems to pathologies that are serious or even fatal. The aetiology of skin diseases is varied and some of the most severe are autoimmune and genetic disorders, including cancer.
Skin tissue engineering techniques emerged in the 1980s to address the need for extensive full-thickness burns coverage in the absence of sufficient autologous skin for grafting [1–3]. In the clinical context they are also useful in treating non-healing ulcers [4, 5]. In addition to its clinical applications, three-dimensional models (3D) of engineered skin are being broadly used in skin biology research, such as reducing animal experimentation (i.e. animal testing for human skin products), investigation of cell interactions and skin barrier penetration, and the development of models of human skin diseases, such as psoriasis [6–8] and the genetic disorder epidermolysis bullosa [9, 10].
Over the last years, genetically modified (transgenic and/or null) mouse models have been developed that recapitulate to a variable extent human diseases, including skin disorders. There is a lot to learn from engineered epidermis using keratinocytes from these animal models. However, despite the early success in culturing human keratinocytes , primary murine keratinocytes have shown to be fairly reluctant to subculture and expansion in vitro, thus precluding the establishment of reproducible 3D models to be used in skin biology research. Initial attempts to culture murine keratinocytes long-term required the use of matrix coated dishes and/or complex fibroblast-conditioned media with fetal bovine serum for cell seeding and growth [12–14]. These cells retained their ability to differentiate, when the concentration of calcium in the culture media was increased, only until passage 10  to 15 . Caldelari et al.  and Reichelt & Haase  achieved serial subculture of murine epidermal keratinocytes for more than 50 and 250 passages, respectively. While Reichelt & Haase method still required the use of feeder cells in the initial phases of growth, collagen-coated dishes and serum-conditioned culture media, Caldelari et al.  subcultured cells in uncoated culture dishes and fully defined media. Also, they showed that their cell line retained the capacity to respond to the elevation of calcium concentration in the culture media with the establishment of intercellular adhesion complex and the expression of terminal differentiation markers.
There has been great advance in the development of fully defined media to culture keratinocytes and to establish 3D epidermal models; however, theses advances are being focused and tested on primary human epidermal keratinocytes. In fact, when compared to the field of human tissue-engineered skin (for a review ), current experience on 3D models using cultured murine keratinocytes is scarce, consisting of organotypic cultures that use devitalised dermis and complex non-defined culture media [13, 18].
We have established a murine keratinocyte cell line that grows on conventional cell culture plastic dishes and with defined culture media. These cells retain its potential to differentiate for more than 75 passages when the calcium concentration is increased, and form an epidermis-like structure using the abovementioned 3D protocols. This murine cell line, termed COCA, is non-tumorigenic and can also form an epidermis in vivo when grafted onto immunodeficient mice. We propose that our cell culture technique is suitable for the establishment of other murine cell lines bearing different genetic manipulations and that these cell lines can be used to establish reproducible 3D epidermal models appropriate for skin biology research.