Recessive epidermolysis bullosa simplex phenotype reproduced in vitro: ablation of keratin 14 is partially compensated by keratin 17

Abstract

Recessive epidermolysis bullosa simplex (REBS) is characterized by generalized cutaneous blistering in response to mechanical trauma. This results from fragility of the basal keratinocytes that lack keratin tonofilaments because of homozygote null mutation in the keratin 14 gene. REBS patients display in addition focal dyskeratotic skin lesions with histology of epidermolytic hyperkeratosis (EHK) and tonofilament clumping in the suprabasal layers of the epidermis. In this study we examined whether it is possible to mimic in vitro the bullous and dyskeratotic cellular phenotype. For this purpose, fibroblasts from nondyskeratotic (K14-/-) and dyskeratotic (K14-/-) skin of a REBS patient and fibroblasts from a healthy donor (K14+/+) were isolated and incorporated into collagen matrices. Subsequently, fresh biopsies originating from the nondyskeratotic and dyskeratotic skin of the patient and from a healthy donor were placed onto the collagen matrices and cultured at the air-liquid interface. Epidermal morphogenesis was evaluated on the basis of tissue morphology and the expression of a series of keratins. The results of the present study indicate that basal cell vacuolization in REBS can be mimicked in vitro but not the EHK. Fibroblasts seem to play an important regulatory role in establishing the REBS phenotype. These findings suggest that wild-type fibroblasts may enhance the stability of K14-/- keratinocytes in vitro.

Figure 1.

Histological appearance of a dyskeratotic (A) and nondyskeratotic (B) skin. Shown are hematoxylin-stained cross sections. The epidermis in dyskeratotic EHK plaques was hyperproliferative consisting of 4 to 11 cell layers, whereas the epidermis of nondyskeratotic nEHK skin consists only of five to six cell layers. Original magnifications, ×200.

Figure 2.

Macroscopic view of the passaging procedure. Generation of a HSE with a biopsy obtained from a REBS patient. Primary culture is established by placing a skin biopsy onto fibroblast-populated collagen matrix and culturing for 32 days at the air-liquid interface. After this culture period the collagen matrix was fully covered with keratinocytes originating from the biopsy. To establish a secondary culture, a fragment from the primary HSE is transferred onto fibroblast-populated collagen matrix.

Figure 3.

The REBS phenotype is reproduced under the in vitro conditions. Histological appearance of reconstructed epidermis established on collagen matrices populated with dyskeratotic (EHK), nondyskeratotic (nEHK), and healthy fibroblasts (C) after 32-day culture at the air-liquid interface. Cross sections of primary (A) and secondary (B) cultures are shown. In both primary and secondary cultures, a fully differentiated epidermis was formed irrespective of the skin donor used. In HSEs generated with a biopsy originating from nondyskeratotic or dyskeratotic skin on a dermal matrix containing fibroblasts derived either from nondyskeratotic or dyskeratotic skin, vacuolization of the basal cells was observed in both primary and secondary cultures. H&E-stained paraffin cross sections are shown. Original magnifications, ×800.

Figure 4.

Keratin expression in native skin. Shown are paraffin cross sections of skin stained for K14 (using antibodies RCK107, LLOO1, and LLOO2), K5, K15, and K10. Positively stained cells in the basal and first suprabasal epidermal layer were seen with RCK107 antibody, with both L001 and L002 antibodies the entire viable epidermis was stained. K5 decorated continuously and K15 locally the basal and first suprabasal epidermal layer. K10 expression was observed in all suprabasal cell layers. Original magnifications, ×200.

Figure 5.

Differential expression of K14 protein in REBS-HSE constructs. Keratin expression in HSEs generated with fibroblasts and keratinocytes originating from a healthy donor; in the skin of the REBS patient; and in HSEs generated with biopsies originating from nondyskeratotic (nEHK), dyskeratotic (EHK) skin on collagen matrices populated with fibroblasts isolated from EHK, nEHK, and healthy fibroblasts (C). Shown are paraffin cross-sections stained for K14 using different antibodies: RCK107, LLOO1, and LLOO2 (A); for K5, K15, K10 (B); and K6, K16, K17 (C). Note: RCK107 antibody showed a weak local staining in B(nEHK)/D(nEHK) and B(EHK)/D(nEHK) cultures; with LL001 and LL002 antibodies no positively stained cells were detected. A similar observation has been made in biopsies taken from the patient. Strong K17 staining was observed in basal cell layer in B(nEHK)/D(nEHK) and B(EHK)/D(EHK) cultures with K17 deposition concentrated in the vacuoles. Original magnifications, ×200.

Figure 5A.

Continued

Figure 6.

RCK107 is expressed in keratinocyte cultures established from REBS skin whereas LL001 is not. Keratinocytes originating from a healthy donor or REBS patient were cultured under submerged conditions in low-calcium medium. Shown is immunofluorescence staining with RCK107 and LLOO1 and double staining with both antibodies. Note the positive staining with RCK107 in REBS keratinocytes. Original magnifications, ×1000.