In this study, we demonstrated that Brn2 mRNA and protein levels were increased in 14 days after calcium treatment of primary keratinocytes, consistent with its exclusive expression in the granular layer of the normal epidermis. Using a recombinant adenovirus technique, we showed that Brn2 has a potential for promoting the keratinocyte differentiation. Actually, Brn2 induced the expression of differentiation markers such as involucrin, loricrin and filaggrin. These effects were believed to be a consequence of direct binding of Brn2 to the promoters of such genes, evidenced by ChIP assay. The most frequent targets of the POU domain proteins are the “octamer” motifs (ATGCAAAT), which are involved in both ubiquitous and cell-type specific regulation of various genes 
. These proteins also have the flexibility to bind heterogeneous sequences, such as the so-called TAATGARAT motif 
. The proximal promoter region of the human profilaggrin gene contains two AT-rich motifs that are homologous to the consensus recognition sequence TAATGARAT of the POU domain transcription factors 
. In our study, the Brn2 binding motifs of the involucrin, loricrin and filaggrin promoter regions are not the same as the Oct binding motifs. However, they do have AT-rich motifs and nearly all binding sites have the same ATTTT motifs. Therefore, Brn2 increases differentiation of keratinocytes through increased transcriptional activity via specific DNA binding to the promoter regions of differentiation-related genes.
Despite its potential for promoting the keratinocyte differentiation, in our study, the epidermal thickness of Brn2-injected rat skin significantly increased with the increased differentiation in the spinous and granular layers. Brn2 markedly inhibited cell proliferation when overexpressed in cultured keratinocytes, however cell proliferation in the basal layer looked like unaffected by Brn2-injecton in rat skin. The discrepancy between in vitro and in vivo data regarding cell proliferation remains to be elucidated. Interestingly, in keap1-null mice, epidermal thickness is increased along with the increased differentiation marker expression 
. These results suggest that epidermal thickness is determined by the balance of keratinocyte proliferation and differentiation, and it is likely that the impairment of the balance between proliferation and differentiation could lead to the thickening of epidermis.
Since intradermal injection of adenovirus can transduce almost all cells in the dermis including fibroblasts and possibly immune cells and endothelial cells, there can be paracrine effects by which the changes in the epidermis are happened. Actually, in our preliminary study, overexpression of Brn2 in fibroblasts resulted in increased expressions of IL-6 and IL-8 (Figure S1
). Furthermore, the conditioned medium collected from Brn2-overexpressed fibroblasts induced the growth of keratinocytes as compared with the control conditioned medium (Figure S3
). Thus, it is likely that epidermal thickening in Brn2-injected rat skin may be partly linked to the paracrine effects of neighboring cells.
The histopathology of Brn2-injected rat skin was similar to human lichen planus, showing hyperparakeratosis with thickening of the granular cell layer, degeneration of the basal cell layer, and infiltration of inflammatory cells into the subepithelial layer of connective tissue. We demonstrated the Brn2 is highly expressed in almost all epidermal cell nuclei in lichen planus using immunohistochemistry with the anti-Brn2 antibody. High levels of loricrin have been identified in the hypergranulotic and hyperorthokeratotic epidermis of lichen planus in a study exploring molecular alterations in keratinocyte differentiation in lichen planus 
. Involucrin has also been identified as a diagnostic marker in oral lichenoid lesions based on observations of involucrin reactivity in the skin and in oral lichen planus 
. There are also reports of high p53 and p21 expression levels in oral lichen planus 
. We found that overexpression of Brn2 in keratinocytes changed the expressions of involucrin, loricrin, p53 and p21, supporting the notion that Brn2 is closely linked to the pathogenesis of lichen planus.
Lichen planus is a pruritic, papular eruption characterized by a violaceous color, polygonal shape, and, sometimes, a fine scale. It is most commonly found on the flexor surfaces of the upper extremities, on the genitalia, and on mucous membranes. Lichen planus skin lesions are thought to be an immunologically mediated reaction involving T-cells because lichen planus histology is characterized by a dense, subepithelial lympho-histiocytic infiltrate with an increased number of intra-epithelial lymphocytes 
. As a consequence, degeneration of basal keratinocytes occurs forming colloid (Civatte, hyaline, cytoid) bodies that appear as homogenous eosinophilic globules 
. Brn2 adenoviral injection of rat skin led to detection of CD3 positive lymphocytes in the epidermis, but not in GFP injected or normal control tissues. However, the density of inflammatory cells in Brn2 injected skin was not as high as in lichen planus, which is characterized by a band-like pattern of inflammatory cell infiltration in subepidermal areas, denying the role of Brn2 in the cause of lichen planus. Our observations of rat skin were made after a one time injection with the Brn2 adenovirus. It is possible that Brn2 could attract more inflammatory infiltrates, like in lichen planus, in a model in which Brn2 is continuously produced. It is also possible that inflammation is triggered by other mechanisms that induce Brn2 expression in keratinocytes. Such mechanisms may exaggerate inflammation while inducing terminal differentiation via positive feed back cycles.
The lymphocytic infiltrates of lichen planus are composed almost exclusively of T-cells with a high expression level of CD3+
on the cell surface 
. While the majority of T-cells within the epithelium and adjacent to damaged basal keratinocytes is activated CD8+
lymphocytes with cytotoxic effects, most lymphocytes in the lamina propria are CD4+
helper T-cells lacking cytotoxic effects 
. We also detected CD3 positive lymphocytes in Brn2 adenovirus injected rat skin, but not in GFP injected control tissues. Most inflammatory cells were observed in the upper dermis with a few approaching the epidermis. Some basal keratinocytes exhibited vacuolar changes, which probably indicates the beginning of lichen planus. Altogether, these results suggest that Brn2 may have a role for pathogenesis of lichen planus. The precise relationship between Brn2 and lichen planus, however, remains to be elucidated.
In conclusion, we provide evidence that Brn2 has a role in mediating keratinocyte differentiation, and is possibly linked to pathogenesis of lichen planus. Further research in this area should document what kinds of phenotypes would be followed in the absence of Brn2 expression in the epidermis.