In vitro differentiation of mouse ESCs to keratinocyte. We differentiated the mESCs in vitro using a protocol adapted from Bilousova et. al. and Metallo et al. with modifications to increase efficiency at the initial phase of the procedure (35, 36) (Fig. 1A). Specifically, mESCs were re-suspended in EB media and the hanging drop methods were used for EB formation. Under these conditions, compact, sphere-shaped EBs formed in nearly 80% of the handing drops (Fig. 1B). After plating on ColIV-coated plates and growing in DKSFM media, the EBs had many cells growing out from the center (Fig. 1C). The outgrown cells displayed predominant ectodermal characteristics with strong expression of the surface ectodermal marker Krt 18. After passage and culture for 21–60 days, the cells started to exhibit basal keratinocyte-like morphology with increased expression of the basal keratinocyte marker Krt 14 (Figs. 1D and 1E). At proximately passage 6 (P6), the cells began to express abundant Krt 14 that was sustained subsequently in at least 20 passages. The cells can also be stored in and recovered from liquid N2 with no apparent change of basal keratinocyte morphology and Krt 14 expression.
Examination of the cells at mRNA level at different days of in vitro culture revealed a gradual decrease of the expression of stem cell markers, Oct4 and Nanog, which reached undetectable expression levels by day 9. Conversely, there was a significant increase in the expression of Krt 18 (Figs. 2A and 2B). The Krt 18 expression reached peak levels at 8–13 days of differentiation, but decreased sharply afterwards, and remained at a steady low level at the 4th passage and onward. Concurrently, the Krt 14 expression increased gradually and arrived at a steady high level after a few passages (Fig. 2C). Cells in the later passages also exhibited a slight expression of suprabasal gene, Krt 1, but had no detectable Fillagrin (Flg) that is expressed only in the uppermost layer of the epidermis at late stage of fetal development (Fig. 2D). The adherence junctions and tight-junctions, key characteristics of the basal keratinocytes detected by E-cadherin and ZO-1 staining, were evidently formed at the intercellular contacts (Fig. 2E) (37, 38). These cells maintained a high level of Krt 14 expression for > 20 passages, including after recovery from storage in liquid N2, and are henceforth termed differentiated-keratinocyte (D-KC).
The role of MAP3K1 in keratinocyte differentiation. MAP3K1 is a signal transduction enzyme playing key roles in embryonic development, specifically in epithelial morphogenesis (16). Although the Map3k1−/− mice do not have overt skin defects, they display eye developmental defects due to abnormal epithelial morphogenesis and delayed healing of skin full-thickness wounds (18). To explore the role of MAP3K1 in skin biology, we re-analyzed the global gene expression in wild type and Map3k1−/− primary keratinocytes (18), using a more stringent cut-off criteria and performed GO analyses of the differentially expressed genes. We found that skin development was a top biological process affected by MAP3K1 (Fig. 3A). Specifically, genes in epithelial terminal differentiation were significantly up-regulated in the Map3k1−/− versus wild type cells (18) (S1 Table).
To evaluate whether the roles of MAP3K1 in epithelial differentiation could be recapitulated in vitro, wild type and Map3k1−/− mESCs were used for in vitro differentiation and the expression of marker genes were examined at different time intervals. Wild type and Map3k1−/− cells had similar expression of Krt 18 at the early phase of differentiation (Fig. 3B). While these cells also had similar Krt 14 expression, they were strikingly different on Krt 1 expression at the later phase of differentiation (Fig. 3C). The expression levels of Krt 1 were 10-fold more abundant in Map3k1−/− than in wild type cells, suggesting that loss of MAP3K1 accelerated differentiation from basal to suprabasal keratinocytes. Thus, the in vitro system validated the role of MAP3K1 in hampering epithelial terminal differentiation originally insinuated from global gene expression.
Dioxin potentiate keratinocyte differentiation in vitro. The global environmental pollutant dioxin exhibits diverse developmental toxicities and is suggested to cause acanthosis and epidermal hyperkeratosis through derailing epithelial differentiation (25). As most dioxin effects are mediated by the Aryl Hydrocarbon Receptor (AHR), a ligand-activated transcription factor that regulate dioxin-responsive genes (39), we examined AHR expression during in vitro epithelial differentiation. The expression of Ahr was negligible in mESCs, but was significantly increased as soon as the cells started to differentiate in the EBs at day 2 of differentiation, consistent with previous observations (40) (Fig. 4A). The Ahr expression continuously increased and remained at high levels after the cells committed to primary ectodermal lineages and became basal keratinocytes, suggesting that AHR signaling could be activated by dioxin as soon as the cells exit stemness.
To evaluate the effects of dioxin on differentiation, we examined cells differentiated in media with or without dioxin for 13 days. The presence of dioxin in the culture media did not alter the expression of Krt 18 and Krt 14, suggesting that dioxin did not change the course of differentiation from mESC to surface ectoderm and basal keratinocytes (Fig. 4B). After multiple passages (> 8), the steady-state D-KC exhibited a robust dioxin-induced AHR activation, reflected by the induction of Cyp1a1, the prototypical AHR target gene (Fig. 4C). While the presence of dioxin did not change Krt 14 expression, it increased Krt 1 expression by 3-fold. Hence, dioxin potentiates basal to spinous keratinocyte differentiation.
Dioxin plus Map3k1 loss-of-function further promote differentiation. In vivo, dioxin and Map3k1+/− have synergistic effects on impairing eye development. When neither dioxin nor Map3k1+/− alone are detrimental, their combination causes birth defects of the eye, a defect observed also in un-treated Map3k1−/− mice (41). Notwithstanding the intriguing phenotypic observations, how the environmental and genetic factors converge to disrupt the developmental programs has remained elusive. Given the similar effects of dioxin and Map3k1 gene mutation on promoting suprabasal differentiation, we postulated that the combination of these conditions exacerbated the differentiation abnormalities. We tested the idea by treatment of D-KC derived from wild type and Map3k1+/− mESCs with 10 nM dioxin for 3 days and examination of Krt 1 expression. Compared to the wild type cells, the Map3k1+/− cells had an increase of Krt 1 expression (Fig. 5A). Dioxin treatment increased Krt 1 expression by 3-fold in wild type cells, and remarkably, it induced Krt1 expression further in Map3k1+/− cells. The Krt 1 expression in dioxin-treated Map3k1+/− cells reached to nearly 8-fold of the levels in the untreated wild type cells and close to the levels in the untreated Map3k1−/− cells (Figs. 3C and 5A).
We also tested this idea in vivo by treatment of pregnant mice, carrying wild type and Map3k1+/− embryos, with 50 ug/kg dioxin on embryonic day (E)11.5. The embryos were collected on E15.5, as described previously (41) and the embryonic skin was examined by immunohistochemistry. The E-cadhesin staining labeled multiple layers of the epithelial cells in the embryonic skin, in which the Krt 1 positive cells were detectable at the most outer layer (Fig. 5B). Quantification of the signal intensities showed that neither dioxin exposure nor Map3k1+/− altered the level of Krt 1 expression; however, their combination significantly increased Krt 1 in more than 20 samples examined (Fig. 5C). The in vivo and in vitro data together raise an intriguing possibility that the gene-environment interactions significantly potentiate basal to suprabasal differentiation as a potential mechanism underlying the eye developmental abnormalities.