Bone marrow transplantation restores epidermal basement membrane protein expression and rescues epidermolysis bullosa model mice

Abstract

Attempts to treat congenital protein deficiencies using bone marrow-derived cells have been reported. These efforts have been based on the concepts of stem cell plasticity. However, it is considered more difficult to restore structural proteins than to restore secretory enzymes. This study aims to clarify whether bone marrow transplantation (BMT) treatment can rescue epidermolysis bullosa (EB) caused by defects in keratinocyte structural proteins. BMT treatment of adult collagen XVII (Col17) knockout mice induced donor-derived keratinocytes and Col17 expression associated with the recovery of hemidesmosomal structure and better skin manifestations, as well improving the survival rate. Both hematopoietic and mesenchymal stem cells have the potential to produce Col17 in the BMT treatment model. Furthermore, human cord blood CD34(+) cells also differentiated into keratinocytes and expressed human skin component proteins in transplanted immunocompromised (NOD/SCID/gamma(c)(null)) mice. The current conventional BMT techniques have significant potential as a systemic therapeutic approach for the treatment of human EB.

Fig. 1.

BMT-induced donor cell-derived COL17 in the epithelized skin tissue. (A) The donor-derived hCOL17 expression is observed in the epithelized skin areas of BMT-treated C57BL/6 recipients (yellow arrows). Green: mCol17 (KT4.2) or hCOL17 (D20); red: nuclei; broken lines: skin surface; arrowheads: BMZ. (Scale bars: 50 μm.) (B) Representative RT-PCR analysis for hCOL17 expression reveals positive bands in BMT-treated C57BL/6 recipients. All RNA samples were extracted from full-thickness skin biopsies, except for HaCaT from cultured cells. (C) Immunohistochemical analysis of the BMT-treated COL17^{m−/−, h+} skin tissue demonstrates donor-derived GFP⁺ CD45⁺ blood cells (yellow arrowheads) and recipient-derived GFP⁻ CD45⁺ cells (yellow arrow). Donor-derived GFP⁺ CD45⁻ (green arrowheads) cells are sporadically noted in the epidermis. (Scale bars: 20 μm.) (D) Aggregated donor-derived GFP⁺ cells in the basal cell layer are noted, some of which also express cytokeratin (white arrows). These cells are thought to be donor-derived keratinocytes. (Scale bars: 20 μm.) (E) The skin of the recipients shows sporadic, linear deposition of mCol17 (arrows). The deposition is limited to the epithelized skin area with acanthosis. (Upper) The entire consolidated image. (Lower) Higher magnification. (Scale bars: 50 μm.) (F) RT-PCR analysis shows the recovery of mCol17 mRNA in two out of three representative mice (lanes 2 and 3).

Fig. 2.

BMT treatments induce functional mCol17 in COL17^m−/− junctional EB model mice. (A) In the epithelized skin tissue of BMT-treated mice, a cluster of GFP⁺ cytokeratin⁺ basal cells is observed. Green: GFP; blue: nuclei; broken lines: skin surface. (Scale bars: 10 μm.) (B) Sporadic GFP⁺ cells (green) are shown in the epithelized skin of the recipients (arrow). Furthermore, linear staining of mCol17 is detected in the BMZ (red, KT4.2, arrowheads). (Scale bars: 20 μm.) (C) Western blotting analysis reveals the expression of mCol17 in the epithelized area of the BMT-treated COL17^m−/− mouse (lane 3), and a weak band is seen in unwounded skin of a BMT-treated COL17^m−/− mouse (lane 2). β-actin: loading control. (D) The expression of mCol17 is detected only in the epithelized skin and not in the unwounded skin area. Also, the expression is limited to the epidermis side of the epithelized skin. (E) The sorted GFP⁺ single epidermal cells of BMT-treated COL17^m−/− mice express various keratinocyte-specific mRNAs as well as mCol17. Sorted GFP⁻ cells express these mRNAs, other than that of mCol17. (F) No fused cells are apparent in the epidermis, although donor-derived XY cells are sparsely shown. Sporadic fused cells with XXXY chromosomes are observed in the deep dermis. Dashed circles indicate the border of the nucleus. e: epidermis; d: dermis. (Scale bars: 10 μm.) (G) The epithelized skin of COL17^m−/− mice has hypoplastic hemidesmosomes with thin, poorly formed inner/outer plaques (arrowheads). In BMT-treated COL17^m−/− mice, hemidesmosomes with mature plaques are seen. (Scale bars: 500 nm.) *P < 0.01.

Fig. 3.

BMT treatments in COL17^m−/− junctional EB model mice change vulnerability to friction in the skin and induce better clinical conditions. (A) epithelized areas after erosion formation are investigated by rubber stress test. In untreated COL17^m−/− mice, mild mechanical stimulus induces large erosions. Conversely, BMT-treated mice show less severe erosions. *P < 0.05. (B) The erosion area expressed as a percent of the rubbed area is measured for each group. Resistance of the skin to mechanical stimuli is significantly improved in the BMT-treated COL17^m−/− mice. (C) Survival curves of BMT-treated and -untreated COL17^m−/− mice from d 35 after birth (the day of BMT). COL17^m−/− mice treated with BMT from COL17^m−/− mice are shown as the BMT control mice; 73.7% of BMT-treated COL17^m−/− mice could be expected to live over 200 d after BMT treatment vs. only 27.5% of untreated COL17^m−/− mice and 16.7% of BMT control mice. (P = 0.015 for BMT-treated vs. untreated mice, P = 0.021 for BMT-treated vs. BMT control, and P = 0.964 for BMT control vs. untreated.). (D) Clinical manifestations at 90 d after BMT treatment (125 d after birth). Untreated COL17^m−/− mice show moderate perioral erosions with crusts and anal erosions occurring spontaneously. In contrast, BMT-treated COL17^m−/− mice show mild erosions in these areas.

Fig. 4.

HSCs and MSCs each have the potential to produce mCol17 in transplanted COL17^m−/− mice. (A) HSCs and MSCs from GFP⁺ Tg mice were sorted or cultured. One or the other type of these stem cells with supporting COL17^m−/− whole BM cells were injected into preirradiated COL17^m−/− mice. (B) Sparse GFP⁺ cytokeratin⁺ cells, shown by white arrows, are detected in the epithelized skin of HSC-BMT model mouse (Upper). Also in the MSC-BMT model, GFP⁺ cytokeratin⁺ cells are observed (Lower). (Scale bars: 10 μm.) (C) Punctate staining of mCol17 is noted, shown as yellow arrows, in the epithelized skin tissue of both HSC- and MSC-BMT model mice. (Scale bars: 20 μm.) (D) RT-PCR analysis of the epithelized skin area after full-thickness wounding. Both HSCs (lane 1) and MSCs (lane 2) in the BMT treatment model express positive mCol17. Also, single i.v. injection of GFP⁺ MSCs (lane 3) induces weak mCol17 mRNA expression. (E) At 90 d after treatment, the HSC-BMT model mice (Center) demonstrate better clinical manifestations than the untreated COL17^m−/− mice (Left), whereas mice of the MSC-BMT models (Right) tend to show more severe perianal erosions and hair loss.

Fig. 5.

Human hematopoietic stem cell transplantation induces human epidermal keratinocytes that produce BMZ proteins. (A) The epithelized skin samples of treated NOG mice include sporadic cytokeratin⁺ (red), HLA-ABC⁺ (blue) cells in the basal cell layer, which indicate human cord blood-derived keratinocytes. (Scale bars: 10 μm.) (B) 3D analyses of the immunohistochemical sections prove costaining of keratin (red) and HLA-ABC (green), indicating that these cells are human cord blood-derived keratinocytes and not two distinct overlaid cells. Blue lines: cross-section edges. (Scale bar: 10 μm.) (C) Sparse, linear deposition of hCOL17 is noted in the epithelized skin of CBSCT-treated NOG mice (yellow arrows). Green: hCOL17 (D20); red: cytokeratin; (Scale bars: 20 μm.) (D) RT-PCR analysis for two transplanted NOG mice, for both unwounded and epithelized skin (CBSCT 1 and CBSCT 2). The expression of several BMZ proteins, hCOL17, BPAG1, plectin, α6 integrin, laminin β3, and laminin γ2 mRNA is demonstrated. Unwounded skin shows faint expression of hCOL17 and α6 integrin.