Yap1 has been described as a key regulator of normal skin differentiation, balancing proliferation and differentiation programs.1,2 Upon activation of the Hippo pathway and other upstream pathways, Yap1 becomes phosphorylated at serine 127 and is retained in the cytoplasm through binding to the scaffold protein stratifin.3 This process results in inhibition of proliferation and activation of cellular differentiation. Conversely, in skin stem cells, Yap1 remains unphosphorylated in the nucleus, thereby promoting cell proliferation.
In skin tumors, several studies have demonstrated that Yap1 accumulates in the nucleus in basal cell carcinoma (BCC) and cutaneous squamous cell carcinoma (cSCC), and this feature appears to play an important role in the development of these keratinocyte carcinomas.4–6
In a previous study, we observed that cSCC expresses higher levels of the cytokine S100A9 (calgranulin B), whereas BCC does not exhibit S100A9 expression.7 S100A9 expression has been associated with aggressiveness and metastatic potential.8 Consistent with this observation, cSCC is generally more aggressive and metastatic than BCC. Interestingly, S100A9 expression is regulated by the Hippo–Yap1 pathway in cSCC.9 Inactivation of the Hippo pathway and the accumulation of nuclear Yap1 repress S100A9 expression. Therefore, Yap1 may be predominantly cytoplasmic and retained by stratifin in cSCC to allow S100A9 expression, which appears to contradict the nuclear role of Yap1 in promoting cSCC development.
To better understand the status of Yap1 in the most frequent nonmelanoma skin cancers, we analyzed its expression by immunohistochemistry (IHC) in samples of cSCC, BCC, Merkel cell carcinoma (MCC), atypical fibroxanthoma (AFX), and dermatofibrosarcoma protuberans (DFSP). In these samples, we also evaluated the expression of S100A9 and stratifin.
We analyzed Yap1 localization by calculating the cytoplasm-to-nucleus ratio. For this purpose, Yap1 signal intensity was measured in both the cytoplasm and nucleus. Samples of cSCC (n=19), BCC (n=27), and AFX (n=20) were highly heterogeneous and exhibited cells with cytoplasmic and nuclear staining in similar proportions (Fig. 1). DFSP samples (n=14) showed predominantly nuclear staining. Finally, Yap1 expression was not detected in MCC samples (n=9).
Yap1 expression in nonmelanoma skin cancers. BCC (A), cSCC (B), AFX (C), DFSP (D), and MCC (E) (40×). Histoscore was calculated based on staining intensity using a scale from 0 to 3 (0=no staining, 1=weak staining, 2=moderate staining, 3=strong staining). Statistical significance was evaluated using the Kruskal–Wallis test followed by Dunn multiple-comparison test (*P<.05; **P<.01).
We also performed immunohistochemistry for phosphorylated Yap1 at serine 127 (Yap1-pS127) to evaluate Hippo pathway activity. Yap1-pS127 intensity was assessed using a histoscore ranging from 0 to 300. The results showed that BCC (mean, 252.78; n=23) and cSCC (mean, 246.36; n=23) exhibited high expression levels of Yap1-pS127 (Fig. 2). AFX samples (mean, 194; n=22) showed lower levels of Yap1-pS127, whereas DFSP samples (mean, 116.92; n=18) demonstrated a marked reduction in Yap1-pS127 expression (Fig. 2).
Expression of phosphorylated Yap1 in nonmelanoma skin cancers. Immunostaining for Yap1-pS127 in BCC (A), cSCC (B), AFX (C), and DFSP (D) (40×). Quantification was performed using a histoscore (Hscore) that considers staining intensity and percentage of positive cells, according to the formula: Hscore=1×(% weak staining)+2×(% moderate staining)+3×(% strong staining). The Hscore ranges from 0 to 300. Statistical significance was evaluated using the Kruskal–Wallis test followed by Dunn multiple-comparison test (***P<.001; ns, not significant).
Finally, we analyzed the expression of the Yap1-regulated protein S100A9 and the Yap1 scaffold protein stratifin in AFX (n=20), DFSP (n=14), and MCC samples (n=8) (Fig. 3). Neither S100A9 nor stratifin expression was detected in these tumors. These proteins are also not expressed in BCC samples but are present at high levels in cSCC samples.7
We conclude that, except in MCC samples, the remaining nonmelanoma skin cancers studied displayed a characteristic mixture of tumor cells with either nuclear or cytoplasmic accumulation of Yap1. Among these tumors, DFSP samples showed a higher proportion of cells with nuclear Yap1 and, consistently, reduced Yap1-pS127 expression. This correlation supports the robustness of our findings.
Nuclear accumulation of Yap1 has been associated with increased proliferative activity and tumor aggressiveness.4–6 However, given the heterogeneous localization of Yap1 within tumor samples, several possible scenarios may be considered. For example, in cSCC samples, proliferation may be enhanced through different mechanisms depending on whether Yap1 is localized in the nucleus or cytoplasm. Cells with cytoplasmic Yap1, which may be retained in the cytoplasm by stratifin, could express S100A9 and thereby promote proliferation.8
BCC, DFSP, and AFX samples do not appear to express stratifin7 (this work). This finding is consistent with observations in many epithelial tumors, in which stratifin is considered a tumor suppressor.9 Nevertheless, we also observed a substantial proportion of cytoplasmic Yap1 in BCC and AFX tumors. It is therefore possible that other 14-3-3 proteins retain phosphorylated Yap1 in these tumors. In agreement with this hypothesis, previous studies have shown that 14-3-3ɛ, 14-3-3γ, and 14-3-3ζ proteins are expressed in skin carcinomas but not in normal keratinocytes.10
In conclusion, our findings indicate that Yap1 status is heterogeneous among the nonmelanoma skin tumors studied. Within the same tumor, aggressiveness may be regulated through different mechanisms depending on whether Yap1 is localized in the nucleus or cytoplasm of specific tumor cell populations.
Conflict of interestThe authors declare no conflict of interest.
