Dystroglycan in Skin and Cutaneous Cells: β-Subunit Is Shed from the Cell Surface  Christine Herzog, Cristina Has, Claus-Werner Franzke, Frank G. Echtermeyer, Ursula Schlötzer-Schrehardt, Stephan Kröger, Erika Gustafsson, Reinhard Fässler, Leena Bruckner-Tuderman  Journal of Investigative Dermatology  Volume 122, Issue 6, Pages 1372-1380 (June 2004) DOI: 10.1111/j.0022-202X.2004.22605.x Copyright © 2004 The Society for Investigative Dermatology, Inc Terms and Conditions

Figure 1 Schematic structure of dystroglycan (DG) and the epitope regions recognized by antibodies. The primary DG polypeptide is cleaved at the position 653–654 (arrow) to yield the N-terminal, extracellular α-DG, and the C-terminal transmembrane protein β-DG. The signal peptide (SP), the transmembrane domain (TM) and the mucin-like region with extensive O-glycosylation are indicated. For characterization of DG in skin and cutaneous cells domain-specific antibodies were used. The antibody recognition sites are indicated by colored boxes. The α-DG antibody recognizes the C-terminus of α-DG (violet), β-DG-N the N-terminus of β-DG (pink), and 43DAG/8D5 the C-terminus (blue). Journal of Investigative Dermatology 2004 122, 1372-1380DOI: (10.1111/j.0022-202X.2004.22605.x) Copyright © 2004 The Society for Investigative Dermatology, Inc Terms and Conditions

Figure 2 Expression of dystroglycan (DG) in human skin and cutaneous cells. Expression of DG was analyzed by immunohistochemistry (a, b, e, f) and in situ hybridization (c, d, g, h). (a) Immunofluorescence staining with 43DAG/8D5 antibody exhibits a signal along the basement membrane (BM) (arrow). (b) As a control for BM staining, a collagen XVII antibody was used (arrow). In primary keratinocytes (e) and fibroblasts (f) a positive DG staining was observed. (c) DG mRNA was expressed by epithelial cells in human skin. (d) The control sense probe showed no staining. DG mRNA was also detected in primary keratinocytes (g) and fibroblasts (h). (i) Ultrastructural localization of DG at the dermo-epidermal junction (DEJ). Immunoelectron microscopy with antibody 43DAG/8D5 produced a constant signal associated with hemidesmosomes (Hd). The weak reaction is due to low affinity of the antibody under conditions required for immunoelectron microscopy. Note that the gold particles are localized only around the hemidesmosomes, and that other intra- and extracellular structures remain negative. This staining pattern was reproducible in a number of sections and over long stretches of the DEJ zone, rendering the probability of a non-specific reaction very unlikely. The extracellular localization of the gold particles (arrows) is explained by the distance between the epitope and the gold particle, which is generated by use of first and second antibodies. Journal of Investigative Dermatology 2004 122, 1372-1380DOI: (10.1111/j.0022-202X.2004.22605.x) Copyright © 2004 The Society for Investigative Dermatology, Inc Terms and Conditions

Figure 3 Characterization of dystroglycan (DG) in cell extracts and media. (A) Immunoblotting of WGA–sepharose-concentrated culture media of cardiac muscle cells (M), keratinocytes (K), and fibroblasts (F) revealed an α-DG band with an apparent molecular weight of 160 kDa. (B) Immunoblotting of keratinocyte (K) and fibroblast (F) extracts with the 43DAG/8D5 antibody, which recognizes the intracellular domain of β-DG, revealed a 43 kDa β-DG band and an additional 30 kDa fragment (β-DG30). The β-DG-N antibody specific for the N-terminal ectodomain of β-DG detected only the 43 kDa β-DG. (C) Proteolytic processing of β-DG. The scheme shows the domain structure of β-DG, the cleavage site of α- and β-DG (red arrow) and the region where shedding of β-DG potentially takes place (black arrow). Journal of Investigative Dermatology 2004 122, 1372-1380DOI: (10.1111/j.0022-202X.2004.22605.x) Copyright © 2004 The Society for Investigative Dermatology, Inc Terms and Conditions

Figure 4 Recombinantly expressed dystroglycan (DG) can be shed from the cell surface. COS-7 cells were transiently transfected with cDNA of DG (+) or with empty vector (-). Forty-eight hours post-transfection cell extracts were prepared and immunoblotted with 43DAG/8D5. The transfected cells contain the αβ-precursor, intact β-DG, and the shed 30 kDa fragment (β-DG30). The lower panel shows a densitometric analysis of the above signals for β-DG30 and β-DG43 in control and transfected cells. Scion Image (NIH) software was used. Journal of Investigative Dermatology 2004 122, 1372-1380DOI: (10.1111/j.0022-202X.2004.22605.x) Copyright © 2004 The Society for Investigative Dermatology, Inc Terms and Conditions

Figure 5 Detection of released ectodomain of β-dystroglycan (DG) in cell culture media. Upper panel: the schematic diagram shows the β-DG construct that was assembled to allow detection of ectodomain shedding. A 6×His-tag and an Xpress epitope were added to the N-terminus of β-DG, which is a type I transmembrane protein. The 6xHis tag allowed a simple one-step purification of β-DG shed into the culture supernatant via metal affinity Talon beads. Antibody recognition sites for 43DAG/8D5 (DAG) and β-DG-N are indicated by colored boxes. Xpress, recognition site for anti-Xpress antibody; TM, transmembrane domain. Lower panel: COS-7 cells were transiently transfected with the above β-DG construct. Forty-eight hours post-transfection the medium (M) and the cell layer (E) were analyzed separately for the presence of recombinant β-DG or its shed form using the antibodies 43DAG/8D5, β-DG-N, and anti-Xpress. The shed ectodomain, was detected in the media with both β-DG-N and Xpress antibodies. It exhibited an electrophoretic mobility corresponding to a molecular weight of about 25 kDa. The full-length 43 kDa band of β-DG was detected with all antibodies in cell extracts. The weak signal of β-DG seen in the media was due to a minor cytolysis and release of cell fragments into the medium during transfection and subsequent culture. Journal of Investigative Dermatology 2004 122, 1372-1380DOI: (10.1111/j.0022-202X.2004.22605.x) Copyright © 2004 The Society for Investigative Dermatology, Inc Terms and Conditions

Figure 6 Shedding was stimulated by IL-1β and phorbol esters. (A) Keratinocytes were cultured for 6 h in the presence of IL-1β (5 ng per mL). Aliquots were drawn at regular intervals and analyzed by immunoblotting with 43DAG/8D5 antibody. The signals were analyzed densitometrically with TotalLab1D (Phoretics) software. (B) Keratinocytes were cultured for 1 h in the presence of 0.5 μM phorbol 12-myristate 13-acetate (PMA) or 0.5 μM PMA+1 μM bisindolylmaleimide I (BIM). The stimulating effect of PMA can be compensated by BIM. Journal of Investigative Dermatology 2004 122, 1372-1380DOI: (10.1111/j.0022-202X.2004.22605.x) Copyright © 2004 The Society for Investigative Dermatology, Inc Terms and Conditions

Figure 7 Lack of perlecan, a binding partner for α-dystroglycan (DG), destabilizes the DG complex. (A) Shedding of β-DG is increased in perlecan-deficient fibroblasts. Cell extracts (upper panel) and media (lower panel) from fibroblasts isolated from perlecan +/+ and −/− mice were analyzed by immunoblotting using the antibodies 43DAG/8D5 (upper panel) and α-DG (lower panel). Equal amounts of protein were loaded. Upper panel: extracts of perlecan-deficient fibroblasts contain increased amounts of β-DG30. Lower panel: culture media of perlecan −/− fibroblasts contain more α-DG than +/+ controls. (B) Matrix deposition is reduced in perlecan −/− fibroblasts. Immunofluorescence staining with antibodies to perlecan is strongly positive in +/+ fibroblasts (a) and negative in −/− fibroblasts (c). Antibodies to α-DG showed an intense staining of extracellular matrix fibrils in +/+ fibroblasts (b), but a greatly reduced signal in −/− fibroblasts (d). Journal of Investigative Dermatology 2004 122, 1372-1380DOI: (10.1111/j.0022-202X.2004.22605.x) Copyright © 2004 The Society for Investigative Dermatology, Inc Terms and Conditions