View metadata, citation and similar papers at core.ac.uk brought to you by CORE provided by Elsevier - Publisher Connector Hemidesmosomes Show Abnormal Association with the Keratin Filament Network in Junctional Forms of Epidermolysis Bullosa James R. McMillan, John A. McGrath, Michael J. Tidman,* and Robin A. J. Eady Department of Cell Pathology, St John’s Institute of Dermatology, UMDS, St Thomas’s Hospital, London, U.K.; *Department of Dermatology, Royal Infirmary of Edinburgh, Edinburgh, U.K. Junctional epidermolysis bullosa is a group of hereditary respectively. In junctional epidermolysis bullosa with bullous disorders resulting from defects in several hemi- pyloric atresia (α6β4 abnormalities, n J 3) the values desmosome-anchoring filament components. Because were also reduced [41.8% K 7.0 (p < 0.001) and 44.5% hemidesmosomes are involved not only in keratinocyte- K 5.7 (p < 0.001), respectively]. In the non-Herlitz group extracellular matrix adherence, but also in normal (laminin 5 mutations, n J 3) the counts were 66.7% K anchorage of keratin intermediate filaments to the basal 7.1 (p > 0.05) and 70.5% K 8.5 (p < 0.05), and in keratinocyte membrane, we questioned whether this skin from patients with bullous pemphigoid antigen 2 intracellular function of hemidesmosomes was also per- mutations (n J 3) the counts were 54.3% K 13.8 (p < 0.01) turbed in junctional epidermolysis bullosa. We used and 57.1% K 13.9 (p < 0.01). In epidermolysis bullosa quantitative electron microscopic methods to assess cer- simplex associated with plectin mutations the values were tain morphologic features of hemidesmosome–keratin 31.9% K 8.9 (p < 0.001) for keratin intermediate filaments intermediate filaments interactions in skin from normal association and 39.9% K 7.1 (p < 0.001) for inner plaques. subjects (n J 11) and from patients with different forms Findings in recessive dystrophic epidermolysis bullosa patients’ skin were indistinguishable from normal control of junctional epidermolysis bullosa (n J 13). In addition, skin with inner plaques (90.5% K 2.5) and keratin inter- skin from patients with autosomal recessive epidermolysis mediate filaments attachment (86.3% K 2.1). These bullosa simplex with plectin defects (n J 3) or with findings suggest that the molecular abnormalities under- autosomal recessive dystrophic epidermolysis bullosa lying different forms of junctional epidermolysis bullosa (n J 4) were included as controls. Values were expressed appear to affect certain critical intracellular functions as a percentage of the total number of hemidesmosomes of hemidesmosomes, such as the normal connections counted. In normal skin 83.3% K 3.3 (SEM) hemidesmo- with keratin intermediate filaments. This may have somes were associated with keratin intermediate filaments important implications for the maintenance of basal and 90.1% K 1.9 had inner plaques. In Herlitz junctional keratinocyte integrity and resilience in junctional epidermolysis bullosa (laminin 5 abnormalities, n J 4) epidermolysis bullosa. Key words: basement membrane/ these values were reduced to 45.3% K 11.5 (p < 0.001; blistering skin disease/electron microscopy. J Invest Dermatol analysis of variance) and 50.3% K 12.8 (p < 0.001), 110:132–137, 1998 emidesmosomes are junctional complexes distributed The molecular pathology of JEB is complex; this autosomal recessive along the inner or stromal-facing aspect of basal cells disease may result from mutations in any one of six distinct genes of the epidermis and other stratified epithelia (Tidman encoding different molecular components of the hemidesmosome- and Eady, 1984; Legan et al, 1992; Borradori and anchoring filament complex (Christiano and Uitto, 1996). Sonnenberg, 1996). A major role of hemidesmosomes Initially, the ‘‘lethal’’ or Herlitz form of JEB was shown to be caused His in epidermal–dermal adhesion and defective hemidesmosome func- by mutations in each of the three genes of the different polypeptide tion may result in epidermal detachment or blistering at the level of chains comprising the trimeric anchoring filament protein, laminin 5 the lamina lucida of the basement membrane, as seen in the group of (Aberdam et al, 1994; Baudoin et al, 1994; Pulkkinen et al, 1994; hereditary mechanobullous disorders, junctional epidermolysis bullosa Kivirikko et al, 1995). It was then found that mutations in either a (JEB) (Tidman and Eady, 1986; Borradori and Sonnenberg, 1996). laminin 5 gene or the gene for another hemidesmosome-anchoring filament component, the 180 kDa bullous pemphigoid antigen (BP180/ BPAG2/COL17 A1), could result in a nonlethal form of JEB known as the ‘‘generalized atrophic benign’’ form of epidermolysis bullosa Manuscript received August 13, 1997; revised October 7, 1997; accepted for (EB) (Jonkman et al, 1995; McGrath et al, 1995a; McGrath et al, publication October 14, 1997. 1995b). More recently, mutations in the α6 and β4 integrin genes Reprint requests to: Professor R. A. J. Eady, Department of Cell Pathology, St John’s Institute of Dermatology, St Thomas’s Hospital, Lambeth Palace Road, have been shown to give rise to a form of JEB associated with pyloric London SE1 7EH, U.K.. atresia (Vidal et al, 1995; Shimizu et al, 1996; Pulkkinen et al, 1997). Abbreviations: BPAG1, bullous pemphigoid antigen 1; BPAG2, bullous The α6β4 integrin is a hemidesmosome component and is a receptor pemphigoid antigen 2; BP180, 180 kDa bullous pemphigoid antigen; BP230, that binds laminin 5 (Niessen et al, 1994). The β4 integrin subunit has 230 kDa bullous pemphigoid antigen. been implicated in the second major function of hemidesmosomes, 0022-202X/98/$10.50 · Copyright © 1998 by The Society for Investigative Dermatology, Inc. 132 VOL. 110, NO. 2 FEBRUARY 1998 HEMIDESMOSOME–KERATIN FILAMENT INTERACTIONS 133 making connections with the cytoskeleton (Spinardi et al, 1993; type VII collagen monoclonal antibody (LH7:2) from I. Leigh (Royal London Mainiero et al, 1995; Spinardi et al, 1995; van der Neut et al, 1996; Hospital, U.K.) (Leigh et al, 1987). Indirect immunofluorescence microscopy Niessen et al, 1997). was performed on skin biopsies. The skin specimens were embedded in OCT The linkage of the keratin intermediate filament network to the compound (Miles Diagnostic, Elkhart, IN) and snap frozen in isopentane cooled by liquid nitrogen. Cryostat sections (5 µm) were collected on gelatin and hemidesmosome and the basal keratinocyte plasma membrane may chrome alum-coated slides and air dried. The sections were used immediately involve several components (Katz et al, 1991). Current evidence or stored until needed in air-tight boxes at –70°C. suggests that at least two distinct proteins are involved in this role. Indirect immunofluorescence was carried out with modifications to the The first is the 230 kDa bullous pemphigoid antigen (BP230) or method described previously (Kennedy et al, 1985). Sections were fixed in cold bullous pemphigoid antigen 1 (BPAG1) (Guo et al, 1995), and the acetone (–20°C) for 10 min and incubated with normal rabbit sera for 5 min second is plectin (Smith et al, 1996). Experimental mice genetically at 37°C. Sections were incubated with primary antibodies, secondary antibodies deficient in BPAG1 have small hemidesmosomes with defective connec- rabbit anti-mouse fluorescein isothiocyanate, diluted in 3% bovine serum tions with keratin filaments (Guo et al, 1995). The basal keratinocytes albumin in 0.1 M Dulbecco’s phosphate-buffered saline for 30 min at 37°C in are intrinsically fragile and can be induced by trauma to rupture a darkened, humidified chamber. The sections were then mounted in 0.2% p-phenylenediamine in phosphate-buffered saline and glycerol and examined through the cytoplasm in a zone just above the hemidesmosomes (Guo with a Nikon Optiphot 2 microscope (Nikon UK, Kingston, U.K.) equipped et al, 1995). Although no human genetic disease has yet been shown for epifluorescence. Controls included normal skin as a positive control and to arise from a mutant BPAG1 gene, a similar epidermal lesion has substituting the primary antibody with phosphate-buffered saline, myeloma been described in an autosomal recessive form of EB simplex caused supernatant, or an irrelevant immunoglobulin isotype, as negative controls. by mutations in plectin, another hemidesmosome plaque component that shares DNA sequence similarities with BPAG1 (Gache et al, 1996; Electron microscopy The skin specimens were cut into small pieces and McLean et al, 1996; Smith et al, 1996). Plectin therefore would placed in half-strength Karnovsky fixative containing 2% formaldehyde and 2.5% glutaraldehyde in 0.04 M sodium cacodylate buffer with 0.05% calcium also seem to be involved in the linkage of keratin filaments to 1 hemidesmosomes. A further hemidesmosome component, HD1, has chloride and 5% sucrose (Eady, 1985). Samples were fixed while on a rotator at room temperature for 4–5 h and then washed in three changes molecular mass, tissue distribution, and probable functional similarities (15 min each) of 0.067 M cacodylate buffer with 0.05% calcium chloride to plectin (Hieda et al, 1992; Smith et al, 1996). and 5% sucrose. Postfixation was in 1.3% osmium (2% stock diluted one Currently, therefore, BPAG1, plectin, HD1, and the β4 subunit of part water to two parts osmium) for 2 h at room temperature. Specimens the integrin α6β4 have all been implicated in the linkage of keratin were dehydrated in a graded ethanol series (15 min each), stained en bloc filaments to hemidesmosomes; however, we are unaware of any in 2.5 or 5% uranyl acetate dissolved in 50% ethanol (1 h), and embedded experimental data specifically examining the role of BP180, not only on in TAAB 812 resin with hard hardener (TAAB, Aldermaston, U.K.) via hemidesmosome assembly but also on the capacity of hemidesmosome propylene oxide (two washes). After two changes of resin the samples were plaques to associate or connect with keratin filaments. Recent data placed in silastic moulds, labeled, and placed in an oven at 60°C for 48 h.