Autoantibodies against adhesion molecules and structures in blistering skin diseases.
Autoantibodies against adhesion molecules and structures in blistering skin diseases.
复制标题
DOI：
10.1084/jem.181.1.1
复制
发表时间：
1995-01-01
期刊：
影响因子：
--
通讯作者：
Stanley JR
中科院分区：
文献类型：
--
作者：
Stanley JR
即使仅根据临床和组织学发现，也有理由怀疑某些水疱性皮肤病的发病机制可能与粘连结构异常有关。在这方面，最好定义的大疱性疾病是天疱疮（其中有两种基本类型：寻常型和落叶型）和大疱性类天疱疮（BP）（1-3）。组织学显示天疱疮中的水疱是由于细胞间粘附的丧失，一种称为棘层松解的病理过程。在寻常型天疱疮（PV）中，粘附力丧失发生在表皮深处，刚好在基底层上方，而在落叶型天疱疮（PF）中，粘附力缺陷更浅表，在颗粒层中。在BP中，水疱是由于基底细胞与基底膜的粘附丧失所致。当免疫荧光研究表明天疱疮和BP患者的皮肤和血清中都有自身抗体时，粘附结构可能参与这些疾病的下一个线索出现了（4）。天疱疮的自身抗体被证明是针对角质细胞表面。事实上，研究人员非常确信这些抗体直接干扰粘附，他们被称为，有点异想天开，抗细胞间水泥物质抗体（5）。在BP中，抗体针对表皮基底膜区。在这两种情况下，在有缺陷的表皮细胞粘附区域中的自身抗体的发现表明，这些抗体可能针对粘附结构或分子。我们对这些疾病的认识发展的下一个阶段是开发模型来测试患者的自身抗体是否具有致病性。在两种类型的天疱疮中，答案变得明确：天疱疮IgG自身抗体可以在没有补体或炎性细胞的额外贡献的情况下引起疾病（即细胞粘附的丧失）。皮肤器官培养模型（6）和自身抗体被动转移至新生小鼠（7，8）均显示了这一点。然而，在BP中，抗体的致病性不是那么明确。补体的固定和炎性细胞的募集可能是水疱形成所必需的（9）。直到最近，还没有一个很好的动物模型来验证这一假设。然而，现在已经表明，当被动转移到新生小鼠时，针对BP抗原之一（下文定义为BPAG 2）的主要抗原表位产生的兔抗体可以诱导疾病（10）。从上述数据，我们知道天疱疮和BP是具有针对表皮区域的致病性自身抗体的疾病，在所述表皮区域中发生细胞-细胞或细胞-基质粘附的损失。因此，可以合理地假设这些自身抗体针对粘附分子或结构。这一假设已得到证实（表1）。免疫化学研究表明，PF抗原是桥粒糖蛋白1（Dsgl），桥粒的跨膜糖蛋白（11-13）。Dsgl的分子克隆表明它是钙依赖性细胞粘附分子的钙粘蛋白超基因家族的成员（14）。分子克隆还显示PV抗原是桥粒芯糖蛋白（Dsg 3）（14 a）。Dsgl和Dsg 3都已通过免疫电子显微镜定位于桥粒（15，16），并且在角质形成细胞中，都结合斑珠蛋白，一种桥粒斑块蛋白（17）。来自PV患者的自身抗体，在Dsg 3的胞外区上亲和纯化，当被动转移到新生小鼠时引起水泡，并且Dsg 3的胞外区可以完全吸收来自PV血清的所有致病性抗体（18，19）。
E ven by clinical and histologic findings alone it is reason-able to suspect that the pathogenesis of certain blistering skin diseases might involve abnormalities in adhesion structures. The best defined buUous diseases in this regard are pemphigus (of which there are two basic types: vulgaris and foliaceus) and bullous pemphigoid (BP)(1-3). Histology shows that the blisters in pemphigus result from a loss of cell-tocell adhesion, a pathologic process called acantholysis. In pemphigus vulgaris (PV) loss of adhesion occurs deep in the epidermis, just above the basal layer, whereas in pemphigus foliaceus (PF) the defective adhesion is more superficial, in the granular layer. In BP blisters result from loss of adhesion of the basal cells to the basement membrane. The next clue that adhesion structures might be involved in these diseases came when immunofluorescence studies demonstrated that both pemphigus and BP patients have autoantibodies in their skin as well as in their serum (4). In pemphigus the autoantibodies were shown to be directed against the keratinocyte cell surface. In fact, investigators were so sure that these antibodies interfered directly with adhesion, they were termed, somewhat fancifully, anti-intercellular cement substance antibodies (5). In BP the antibodies are directed against the epidermal basement membrane zone. In both instances, the findings of autoantibodies in areas of defective epidermal cell adhesion suggested that these antibodies might be directed against adhesion structures or molecules. The next stage in the evolution of our knowledge about these diseases came when models were developed to test whether patients' autoantibodies are pathogenic. In pemphigus of both types the answer became dear: pemphigus IgG autoantibodies can cause disease (ie, loss of cell adhesion) without the additional contribution of complement or inflammatory cells. This was shown by both a skin organ culture model (6) and passive transfer of autoantibodies to neonatal mice (7, 8). However, in BP the pathogenicity of the antibodies is not so clear-cut. Probably fixation of complement and recruitment of inflammatory cells is necessary for blister formation (9). Until recently, there has not been a good animal model to test this hypothesis. Now, however, it has been shown that rabbit antibodies raised to a major antigenic epitope of one of the BP antigens (defined below as BPAG2) can induce disease when passively transferred to neonatal mice (10). From the above data, we know that pemphigus and BP are diseases with pathogenic autoantibodies directed against areas of the epidermis in which loss of cell-cell or cell-substrate adhesion occurs. Therefore, it was reasonable to hypothesize that these autoantibodies are directed against adhesion molecules or structures. This hypothesis has been confirmed (Table 1). Immunochemical studies demonstrated that PF antigen is desmoglein 1 (Dsgl), a transmembrane glycoprotein of desmosomes (11-13). Molecular cloning of Dsgl indicated that it is a member of the cadherin supergene family of calciumdependent cell adhesion molecules (14). Molecular cloning also revealed that PV antigen is a desmoglein (Dsg3)(14a). Both Dsgl and Dsg3 have been localized by immunoelectron microscopy to desmosomes (15, 16) and, in keratinocytes, both bind plakoglobin, a desmosomal plaque protein (17). Autoantibodies from PV patients, affinity purified on the extracellular region of Dsg3, cause blisters when passively transferred to neonatal mice, and the extracellular region of Dsg3 can completely absorb all pathogenic antibodies from PV sera (18, 19).Immunochemical studies indicated that BP …