In Vitro Cell.Dev.Biol.--Animal 35:318--326,June 1999 © 1999Society for In Vitro Biology 1071-2690199 $05.00 + 0.00 CHARACTERIZATION OF A NEW TISSUE-ENGINEERED HUMAN SKIN EQUIVALENT WITH HAIR MARTINE MICHEL] NICOLAS UHEUREUX/ROXANE POULIOT, WEN XU, FRANOOIS A. AUGER, ANDLUCIE GERMAIN 3 Laboratoire de Recherche des Grands Bral4s/LOEX, CHAUQ Pavillon Saint-Sacrement, and D@artement de chirurgie, Universit4Laval, Sainte-Foy, Qugbec, Canada GIS 4L8 (Received 20 January 1999: accepted 2 February 1999) SUMMARY We designed a new tissue-engineered skin equivalent in which complete pilosebaceous units were integrated. This model was produced exclusively from human fibroblasts and keratinocytes and did not contain any synthetic material. Fibroblasts were cultured for 35 d with ascorbie acid and formed a thick fibrous sheet in the culture dish. The dermal equivalent was composed of stacked fibroblast sheets and exhibited some ultrastructural organization found in normal connective tissues. Keratinocytes seeded on this tissue formed a stratified and cornified epidermis and expressed typical markers of differen- tiation (keratin 10, filaggrin, and transglutaminase). After 4 wk of culture, a continuous and ultrastructurally organized basement membrane was observed and associated with the expression of laminin and collagen IV and VII. Complete pilosebaceous units were obtained by thermolysin digestion and inserted in this skin equivalent in order to assess the role of the transfollicular route in percutaneous absoqption. The presence of hair follicles abolished the lag-time observed during hydrocortisone diffusion and increased significantly its rate of penetration in comparison to the control (skin equivalent with sham hair insertion). Therefore, this new hairy, human skin equivalent mode/allowed an experimental design in which the only variable was the presence of pilosebaceous units and provided new data confirming the importance of hair follicles in pereutaneous absorption. Key u,ords: extracellu]ar nmtrix; keratinocytes: ascurbic acid; basement membrane. ] NTI/OI)I;CTION A major role of skin and especially epidermis is to protect the body against external environmental aggression. To perform this pro- tective function, normal morphogenesis and complete differentiation of keratinocyte are both required to form a full thickness epidermis. Several steps participate in this process during which specific pro- teins are synthesized (Fuehs, 1990; Steven and Steinert, 1994) such as epidermal differentiation-specific keratins K1/K10 or proteins in- volved in the latter stages of differentiation (filaggrin, loricrin, in- voluerin, small rich proline proteins, and epidermal transglutamin- ase) (Dale, 1994). Progress in cell culture has allowed researchers to produce skin equivalents by seeding cultured keratinoeytes upon various sub- strates such as: artificial membranes (Maket al., 1991), de-epider- mized dermis (R6gnier et al., 1981), collagen gels (Bell et al., 1979; Auger et al., 1995), chitosan-glycosaminoglycan sponges, or nylon meshes populated by fibroblasts (Berthod et al., 1993, 1997; Boyee and Williams, 1993; Contard et al., 1993; Slivka et al., 1993; Black et al., 1998; for a review, see Gernmin and Auger, 1995). Keratinocyte ~Present address: Department of Oral Biology, University of Washington, Box 357132, Seattle, WA 98195-7132. ePresent address: Department of Bioengineering, University of California- San Diego, La Jolla, CA 92093-0412. aTo whom correspondence should be addressed at Laboratoire de Recher- che des Grands Brfilds/LOgX. CHAUQ Pavilion Saint-Sacrement, 1050 Chemin Sainte-Foy, Sainte-Foy, Quebec, Canada G1S 4L8. seeding and proliferation on these various dermal equi,~alents lead Io the formation of an epidermal layer, whieh becomes stratified am[ cornified when skin equivalents are raised at the air-liquid interface (Pruni6ras et al., 1983: Asselineau et al., 1985; Michel et al., 1993, 1997). We have previously performed pereutaneous absm'ption stud- ies with such skin equivalents and have observed qualitatively simi- lar but quantitatively increased cutaneous permeability compared to normal mouse and human skin (Michel et al., 1995). If all of these in vitro raft cultures have their own characteristics (advantages and limits), they only partially share similarities with their native coun- terpart since they lack hair follicles, nerves, sebaceous glands, sweat glands, and blood vessels (except Black et al., 1998) and in some cases contain artificial materials. Hair follicles have been previously cultured in vitro, in an organ culture system, or in sponge-supported histoculture (Philpott et al., 1990; Li et al. 1992). Their rate of growth and hair fber production have been evaluated to be elose to the physiological values (Li and Hoffman, 1993), and their use for hair drug targeting or gene therapy has been reported (Li and Hoffman, 1993, 1995; Rolland et al., 1993). Hair follicles have also been used as a source of keratinoeytes in a skin equivalent model (Lenoir et al., 1988; Lenoir-Viale et al., 1993). However, none of these models have been designed to recreate a skin equivalent with hair that would also allow percutaneous ab- sorption studies. The process of drug migration through the skin has been described as a passive diffusion process with two distinct routes: the transe- 318