Plasminogen activator inhibitor-2 is expressed in different types of congenital ichthyosis:: in vivo evidence for its cross-linking into the cornified cell envelope by transglutaminase-1

Background Plasminogen activator inhibitor-2 (PAI-2), a regulatory serpin of the plasminogen activator (PA) system, has been described as a potential component of the cornified cell envelope (CE). Protease inhibitors are essential for skin homeostasis and in particular for the regulation of the desquamation process. Therefore, an aberrant expression of PAI-2 could be involved in the pathogenesis of certain cornification disorders. Objectives Evaluation of the expression of PAI-2 in different types of congenital ichthyosis, especially in lamellar ichthyosis/nonbullous congenital ichthyosiform erythroderma (LI/NCIE) and in Netherton syndrome (NTS). Demonstration of the functional relationship between PAI-2 and transglutaminase (TGase)-1. Patients and methods Using immunohistochemistry we evaluated cryosections from individuals suffering from LI/NCIE (n = 67), NTS (n = 6), ichthyosis-follicularis-atrichia-photophobia syndrome (n = 2) and Harlequin ichthyosis (n = 1) in comparison with psoriasis vulgaris and healthy skin. Moreover, we assessed the respective TGase-1 activity and the presence of TGase-1 protein. A functional assay was developed to elucidate whether PAI-2 is a substrate for TGase-1. Results PAI-2 is expressed in different types of congenital ichthyosis and there is a strong correlation between TGase-1 activity and PAI-2 protein signal. Double staining revealed a strong colocalization of TGase-1 activity and PAI-2 protein. The epidermal incorporation of the specific PAI-2 peptide containing a TGase binding site revealed a strong pericellular staining in the stratum granulosum in healthy skin. In contrast, TGase-1-deficient skin showed only a lamellar staining in the stratum corneum. Conclusions We provide in vivo evidence that PAI-2 is a substrate of TGase-1. The normal expression of PAI-2 in a large group of TGase-1-proficient LI/NCIE patients makes it rather unlikely that PAI-2 alone is a primary molecular cause of LI/NCIE.