Marcelle Regnier

Human epidermis reconstructed in vitro: A model to study keratinocyte differentiation and its modulation by retinoic acid

SummaryIt was possible to reconstruct epidermis in vitro by seeding dissociated keratinocytes on de-epidermized dermis and growing such recombined cultures for 1 wk, exposed to air, at the surface of the culture medium. These conditions were chosen to mimic the transdermal feeding and the exposure to the atmosphere that occur in vivo. Contrary to classical cultures performed on plastic dishes covered with culture medium, which show rudimentary differentiation and organization, the architecture of the stratified epithelium obtained in reconstructed cultures and the distribution of differentiation markers such as suprabasal keratins, involucrin, and membrane-bound transglutaminase were similar to those of the epidermis of skin biopsies; moreover, biochemical studies showed that the synthesis of the various keratins and the production of cornified envelopes was similar to what is found with skin specimens. The reconstructed epidermis model was found to be very useful to study in vitro the effect of retinoic acid on keratinocyte differentiation and epidermal morphogenesis.

Differentiation of normal and tumoral human keratinocytes cultured on dermis: Reconstruction of either normal or tumoral architecture

SummaryNormal human keratinocytes isolated from skin and squamous carcinoma cells established from a human tumor (TR146 cell line) both exhibit limited morphologic differentiation when they are grown on conventional plastic dishes. However, when they are seeded on human de-epidermized dermis and cultured at the air-liquid interface, they are able to reform an epithelium having the morphology of the tissue of origin (i.e. skin or squamous carcinoma). The distribution in such reconstructed tissues of differentiation markers such as bullous pemphigoid antigen, 67K keratin, involucrin, membrane-bound transglutaminase, and filaggrin was very similar to their distribution in normal skin and squamous carcinoma specimens, respectively. The degree of differentiation is for both cell types extremely sensitive to culture conditions such as retinoic acid concentration, emersion of the cultures, etc. These results show that subcultured normal or tumoral keratinocytes are able to recover their specific morphogenetic potential when cultured in an environment close to their in vivo situation.

In vitro basal lamina formation may require non-epidermal cell living substrate*

In an adult human epidermal cell culture system in which three different types of dermo‐epidermal junctions could be observed, the in vitro synthesis of basement membrane and type IV collagen was studied with electron microscopy and indirect immunofluorescence via an antiserum to type IV collagen. Essentially, keratinocytes juxtaposed to non‐living substrate did not produce either type IV collagen or an ultrastructural basement membrane, whereas both products were found at dermoepidermal junctions composed of living keratinocytes juxtaposed to living dermis. This suggests that the microenvironment of the keratinocyte may influence the synthesis of junctional components.

In vitro bullous pemphigoid antigen deposited on a millipore filter

In culture, epidermal cells from skin explants will migrate (epibolize) around and under the dermal component of the explant itself. With this kind of system, it has been demonstrated that when epibolized specimens are stained against bullous pemphigoid (BP) sera by indirect immunofluorescent techniques (IIF), fluorescent material is noted between the explants dermal surface and the epidermal cells that have epibolized around alld under the explant [1, 2]. The fluorescence at this interface, the so-called dermo-epibolic junction, suggests the presence and possible in vitro synthesis of BP antigen which is thought to be a component of the lamina lucida. With a slightly different technique, we have made a similar observation that confirms and extends previous work. Normal human skin was cut into thin sheets with a Castroviejo keratome set at a depth of 0.3 mm. The sheets were then cut into small square pieces 2 x 2 ram, and four such explants were placed on a nitrose cellulose disc (millipore filter). The discs and the explants were then supported on a steel grid in a plastic petri dish, and media was added just to moisten, but not submerge, the specimens (i.e., maintaining an air fluid interface). Cultures were made in Eagles minimal essential medium with 20 mM of Hepes Buffer and supplemented with 0.1 mM of sodium pyruvate, 2.0 mM of glutamine, 100 ~tg per ml of streptomycin, 100 units per ml of penicillin, and 10 % fetal bovine serum. All cultures were kept at 37 ~ C. Explants and the discs were taken from culture on days 3 and 5, quick-frozen in liquid nitrogen, and stored at minus 20 ~ C. Later, they were subjected to sectioning on a cryostat microtome set at 4 ~m and then routinely stained by IIF against high titer (over 1/50) BP sera [3]. The conjugate used in this study was commercially prepared (Hyland Laboratories) fluorescein isothiocyanate (F.I.T.C.), labelled goat gamma globulin anti-human immunoglobulin G (antibody concentration 1.7 mg/rnl ; F.I.T.C. to protein ratio 2.6 ; working dilution 1/34). Control trials were also run substituting normal human sera (Pasteur Institute) and Phosphated Buffered Saline (PBS) alone for BP sera. All slides were then mounted in glycerol, examined and photographed (Kodak Ektachrome 160 Tungston Professional) with a Leitz Orthoplan microscope with a vertical illuminator for epi-illumination and a HBO 200 mercury vapor light and filters comb. I and Ss2s Leitz.

Sorting of basal cells with FACS IV (fluorescence activated cell sorter)

To create in vitro cellular pharmacological models, it is essential to standardize techniques of epidermal cell cultures. One aspect of this problem is the heterogeneity of epidermal cells. In this study, we aimed at extracting and purifying the basal cell populations. Three main ways to achieve this goal have been proposed so far. The first is to perform sequential trypsinization (Regnier, Delescluse & Prunieras, 1973). The second is based on differences in cell densities between differentiated keratinocytes and lower cells (Sun & Green, 1976). The third takes advantage of the preferential adherence of basal cells to collagen (Skerrow & Skerrow, 1983). Here, we used another method based on the fact that after dissociation with trypsin-EDTA, basal cells retain at their surface the bullous pemphigoid antigen (BPA) (Grekin et al., 1981), a specific constituent of the basement membrane zone. Since this antigen can be revealed by specific antibodies and indirect immunofiuorescence (Beutner, Jordon & Chorzelski, 1968), a fluorescence-activated cell sorter was used to select out basal cells after reacting them with BP antisera (dilution i: 100) in fresh unfixed suspensions. After cell sorting, cell viability was evaluated by trypan blue exclusion and in vitro cultures. The results are summarized in Table i. Table i shows that the selected fluorescent population is not viable, while the counter selected population is viable. This surprising result suggests that basal cells are essentially non-viable. A possible explanation for this paradox is that basal cells were killed during their transit through the narrow channel of the cell sorter. However, when dissociated epidermal cells were separated on Percoll gradients, similar results were obtained, as shown in Table 2 [light band corresponding to fluorescent population (FACS)]. We conclude that when the dermo-epidermal split needed to prepare epidermal suspension is made by trypsin, some of the isolated basal cells are viable and others are not, and that the cell sorter is not responsible for killing basal cells. Another explanation could be that a number of cells are killed by trypsin during the process of preparing epidermal cell suspensions. To check this point epidermis was split from dermis with EDTA. By the dye exclusion test the percentage of dead cells was above 80%. However, the percentage of cells reacting positively with BP sera in suspension, without fixation was around 15-20%, as with trypsin. (Latter data not shown.) Since the percentage of fluorescent cells was only 16% in fresh unfixed suspension (see Table i) and 42% in air-dried smears (see Table 2), we wondered whether fixation could modify the number of the BP fluorescent reacting cells. To test this hypothesis, we compared the number of BPA+ cells in fresh unfixed suspension with that in the same suspension stained after smearing on glass slides and air-drying. Results are shown in Table 3.