The basic interest of my laboratory concerns proteins that promote the adhesion of keratinocytes to epidermal basement membrane. Our clinical interests focus on patients with autoimmune and genetic subepidermal blistering diseases. Several years ago we identified patients with a scarring and blistering disease of mucous membranes and skin who have IgG autoantibodies directed against a multiunit protein localized just beneath basal keratinocytes in epidermal basement membrane. Such patients, distinguished by their disease-specific autoantibodies, have now been identified throughout the World. Recently, we showed that the protein bound by these patients' autoantibodies is the alpha subunit (and more specifically, the G domain of this subunit) of laminins 5 and 6, adhesion molecules thought to directly promote the adhesion of keratinocytes to epidermal basement membrane. To determine if anti-laminin 5 IgG can be pathogenic in vivo, we developed a passive transfer animal model of this human blistering disease. These studies showed that anti-laminin 5 IgG induced subepidermal blisters in the skin of neonatal and adult mice as well as human skin grafted onto immunodeficient mice. Such blisters developed independent of the activation of complement or the degranulation of dermal mast cells. We also showed that human anti-laminin alpha 3 (and G domain-specific) autoantibodies also produced subepidermal blisters in the latter in vivo model. These models should prove useful in further defining the pathophysiology and treatment of this human blistering disease as well as providing insights about which domains of laminin 5 promote keratinocyte adhesion. Longitudinal studies suggested that patients with this autoimmune subepidermal blistering disease had a predisposition to cancer. To determine if such patients had an increased risk of cancer relative to the general population, a cohort of 35 patients was assembled and the cancers expected were calculated as the summed products of age- and sex-specific observatioon periods with respective incidence rates for all cancers in the NCI-SEER study. The cohort had a median age of 65 yrs; 10 patients had solitary cancers. For this cohort, 1.48 cancers were expected; the 10 cancers observed yielded a relative risk of 6.8 (95% CI: 3.3-12.5). Anti-laminin 5 IgG was not identified in the sera of normal volunteers (n=100), patients with other immunobullous diseases (n=130), or controls with solid cancers (n=50). Anti-laminin 5 autoantibodies appear to be a marker for patients with a mucosal predominant blistering disease that has an increased relative risk for cancer. In other studies, we showed that type XVII collagen, a constituent of hemidesmosomes in basal keratinocytes, is not present in the skin of patients with a distinctive inherited blistering disease (GABEB, OMIN 226650). We subsequently identified four different mutations in the COL17A1 gene of six such families. Affected individuals in three of these families were homozygous for the same 2 base pair deletion in the COL17A1 gene (4003delTC), while probands in two unrelated families carried one copy of this mutation as well as a nonsense mutation on their alternate allele. The other COL17A1 mutation was a splice site mutation. Interestingly, all of these mutations resulted in premature termination codons. Northern blot and immunoprecipitation studies of these patients' keratinocytes showed that these cells contained no detectable type XVII collagen mRNA or protein, findings supporting the idea that nonsense-mediated mRNA decay was responsible for the lack of type XVII collagen protein in these patients' skin. Using laser capture microdissection, we recently demonstrated that selected keratinocytes in one patient in the largest of these GABEB kindreds was homozygous for the 4003delTC germline deletion (like her affected siblings) yet mosaic for a second 2 base pair insertion on one COL17A1 allele. This second, post-zygotic mutation corrected the reading frame and allowed expression of partially corrected type XVII collagen in this patient's epidermal basement membrane. This was the first demonstration of mosaic partial correction of a germline deletion by a second frame-restoring mutation in man - an observation of relevance to other diseases and an example of how the study of patients with dermatologic diseases has furthered our understanding of cutaneous biology.
