Human skin is the first organ for which attempts of ex-vivo reconstruction have been made over the last two decades.
A landmark was the discovery by Rheinwald & Green that irradiated 3T3 murine fibroblasts could be efficiently used to grow keratinocytes from epidermal cell suspensions (1). Epidermal sheets were obtained with incomplete differentiation and fibroblast contamination was frequent, but variants of this technique remained as a reference for human therapy especially in the burn patient, because of the selection of cells of high clonogenic potential.

A better differentiation of the reconstructed of the epidermis was noted by Prunieras et al (2) when keratinocytes were cultured on a dead deepidermized dermis and lifted at the air-liquid interface (3-4). The presence of living fibroblasts in a dermal equivalent, like in Bell's model (5), has been considered by some as an important factor to reconstruct an epidermis with better differentiation.

A further advance in epidermal culture was the development of a serum-free medium in Ham's laboratory (6). The first generation media were based on a mixture of aminoacids, vitamins and minerals completed by fetal calf serum, EGF and cholera toxin (CT). Boyce & Ham adjusted by a careful stepwise approach the basic medium composition to the specific metabolic needs of keratinocytes. In this low calcium medium feeder layers, the cAMP agonist cholera toxin and serum were no more required. The addition of bovine pituitary extract (BPE) and aminoacid supplementation allowed to grow primary cultures to confluency and to passage epidermal monolayers.

The switch to high calcium generated multilayers much similar to those obtained in Green's system (1). Since the early eighties, melanocytes had been grown in a few laboratories using Eisinger and Marko's (7) or Halaban's (8) techniques with serum, cAMP agonists (cholera toxin or IBMX) and phorbol esters (PKC agonists). Gilchrest's system using a pituitary extract(BPE), cholera toxin and a basic medium without serum showed the way towards more physiological culture condition of melanocytes, but cell growth and differentiation were poor (9).

In these techniques phorbol esters increased the mitotic activity of melanocytes and their morphology, and cAMP agonists like CT increased dendricity and melanogenesis.

Using Boyce & Ham's technique for keratinocyte culture (MCDB 153), preliminary observations were made in several laboratories about "melanocyte contamination" of cultures, fibroblasts being eliminated by the low calcium concentration and absence of serum.

Pittelkow et al introduced in the system the phorbol ester TPA and were able to grow pure melanocyte cultures (10). Modifications to this system have been made at several institution and this technique is now used with variants worldwide.

In our lab, Donatien has successfully grown well-differentiated melanocytes without CT or phorbol esters using MCBB153 supplemented with BPE, insulin, hydrocortisone, 3% serum and calcium (11). This physiological modified medium allowed to answer the controversial question of a-MSH receptors on human melanocytes (12).

The two dimensional culture systems so far described are however poor models for studying melanocyte biology ex-vivo. The in-vivo epidermal melanization unit is a well-organized three dimensional structure in which melanocytes are polarized to the basal layers and maintain close contacts with both the basal lamina and neighbouring keratinocytes. The influence of keratinocyte factors to maintain the growth and differentiation of melanocytes ex-vivo either by contact or by diffusible factors, is now well-established (13-14-15).

The reconstruction of the epidermal melanization unit ex vivo has been achieved recently.

Bertaux et al have obtained epidermal explants and stimulated melanogenesis with UVB ex vivo (16).

De Luca et al have obtained a coculture of keratinocytes and melanocytes resulting in a moderately differentiated epidermis with melanocytes polarized in basal position when epidermal keratinocytes were used but only in basal and suprabasal position when non keratinizing oral epithelial cells were used (17-18). Similarly, the polarisation of neonatal and fetal melanocytes and the ratio keratinocytes/melanocytes was studied by Haake & Scott in a epidermis reconstructed on a collagen lattice. The role of keratinocytes in stereologic and ratio regulation was confirmed (19-20).

All these experiments show that a mature epidermis is necessary to regulate the epidermal melanization unit.

Other experiments in cocultures of keratinocytes have been conducted at various institutions [Topol (21), Stainao (22), Valyi (23)].

More recently, Todd et al (24) have stimulated with UVB an epidermis reconstructed ex vivo from pure cultures of melanocytes and keratinocytes on a dead deepidermized dermis with at least 10% melanocytes. The cells came from non caucasian donors and melanosome transfer was not demonstrated.

Our experience with keratinocyte and melanocyte culture led us to set up experiments of epidermal reconstruction with melanocytes ex vivo which have been reported at the last ESPCR meeting in Vienna (25). Using pure cultures of keratinocytes and melanocytes of the same donors, we have been successful in obtaining a well differentiated epidermis with basal functional melanocytes. The system was basically a modification of the Pruniras model using 5% melanocytes in the seeding suspension and a novel culture medium.

UVB irradiation induced a stimulation of melanogenesis macroscopically, microscopically and increased melanin concentrations. These results have been reproduced with skin of various phototypes and melanosome transfer to keratinocytes was demonstrated by electron microscopy.

Despite all these exciting new developments, the complete reconstruction of the human skin remains a challenge, since other epidermal components like Langerhans cells are not yet amenable to long-term cultivation, and that dermal reconstruction has not made real breakthroughs over the last decade.

However, the European Center for the Validation of Alternative Methods (ECVAM) setup by the European Union in Ispra, Italy, has considered skin reconstruction among the most advanced projets for the development of in vitro pharmacotoxicology and research programs in this field should be stimulated.

The future of the models of epidermal reconstruction with melanocytes looks promising. First, with better standardization, assays for pharmacologic and cosmetic agents suspected to interfere with pigmentation may be developped to screen interesting molecules, with a better relevance than monolayers of melanocytes. Second, pathophysiologic and therapeutic studies in human pigmentary disorders may be approached directly ex-vivo or after grafting the reconstructed epidermis on the nude mouse.