TO THE EDITOR Epidermal Langerhans cells (LCs) are professional antigen-presenting dendritic cells (DCs) that reside in the epidermis and form the first immunological barrier to the external environment (Romani et al., 2010). Although considerable progress has been made in identifying the developmental requirement of LCs, the mechanisms that regulate LC development and homeostasis are still not fully elucidated. TGF-β1 is one of key regulators to control LC development and homeostasis. LCs are absent in mice that lack TGF-β1, owing to a failure in LC differentiation, survival or both (Borkowski et al., 1996; Kaplan et al., 2007; Zahner et al., 2011). This finding is further supported by the dependence of LCs development on other TGFβ1-related molecules, including inhibitor of DNA binding 2 (ID2), a TGFβ1-induced inhibitor of helix–loop–helix transcription factors (Hacker et al., 2003), and Runx3, specifically expressed by mature DCs and mediates their response to TGFβ1 (Fainaru et al., 2004). Furthermore, TGF-β1 is also required to maintain the pool of immature LCs in the epidermis (Kel et al., 2010). Although TGF-β1 is expressed by both keratinocytes and LCs, an autocrine source of TGF-β1 from LCs is required for LC development (Kaplan et al., 2007). Taken together, TGF-β1 is essential for the ontogeny, homeostasis, and function of epidermal LCs. However, the intracellular signaling pathway of TGF-β1 in epidermal LCs remains elusive. In a well-defined classical TGF-β linear signaling pathway, once activated, TGF-β1 signals through its two cell surface receptors, TGF-β receptor 1 (TβR1) and TGF-β receptor 2 (TβR2), leading to Smad-mediated transcriptional regulation. There are eight Smads: Smad1 to Smad8. Smad2 and Smad3 are activated through carboxy-terminal phosphorylation by TGFβ1. These receptor-activated Smads (R-Smads) are released from the receptor complex to form a heterotrimeric complex with a common Smad4, and translocate into the nucleus to regulate the transcription of target genes (Derynck and Zhang, 2003; Tsukazaki et al., 1998). In addition, TGF-β1 also can activate other signaling cascades, including MAPK pathways (Itoh et al., 2000; Massague, 2000; Moustakas et al., 2001). Unlike mice with the conventional disruptions of Smad2 and Smad4 that are lethal, mice with Smad3 deficiency are viable and can survive. However, Smad3 knockout (KO) mice develop progressive diseases, including leucocytosis and massive inflammation. The loss of Smad3 results in multiple cell defects, including T cells, neutriphils and macrophage (Werner et al., 2000; Yang et al., 1999). Here we used Smad3 KO mice (Datto et al., 1999) to directly test if TGFβ/Smad3 signaling pathway is involved in the ontogeny and homeostasis of epidermal LCs. During LC early development, a single wave of LC precursors recruited in the epidermis around embryonic day 18, which subsequently differentiate into LCs after birth, and LCs then undergo a massive burst proliferation during the first week of life to form a typical LC network (Chorro et al., 2009). Given that TGF-β signals are required for ontogeny and maintaining of LCs in the epidermis, we first assessed if Smad3 is required for LC homeostasis in vivo. Epidermal cell suspensions prepared from ears and dorsal skin of Smad3KO mice and WT littermates, on day 0, 1, 3, 8 after birth, as well as at 3 and 5-week old, were stained with anti-Langerin and anti-MHC-II antibodies. As shown in Figure 1 a & b, we did observe a 3–4-fold expansion of LCs from 0.78 ± 0.11% on day 0 to 2.72 ± 0.14% on day 3 in WT mice. Langerin expression was undetectable in MHCII+ LCs in the epidermis on day 0 and 1, but upregulated Langerin expression was detected in about 70%–80% LCs by day 3, which is consistent with recent report (Chorro et al., 2009; Kel et al., 2010). However, there were no significant differences on the rations of epidermal LCs between Smad3 KO and WT mice at any time points by flow cytometry (Figure 1a & b) and immunohistological staining at day 0 after born and 5 weeks old (Figure 1c & d). Thus, Smad3 is not required in the TGF-β signal pathway for oncogeny and homeostasis of epidermis LCs. Figure 1 Smad3 is not required for ontogeny and homeostasis of epidermis LCs. (a) and (b) Epidermal suspensions prepared from ears and trunk skin of Smad3 KO and WT littermates on day 0, 1, 3, 8 after birth, as well as at 3 and 5 weeks old, were stained with anti-Langerin ... Upon activation by various stimuli, immature LCs residing in epidermis collect antigen and increase their MHC II and costimulatory molecules, including CD80 and CD86, and then migrate to T cells areas of draining lymph nodes, leading to immune responses (Romani et al., 2010). However, loss of TβR1 in LCs makes LC spontaneous maturation, suggesting the TGF-β signal as an essential pathway to control the immature state of LCs (Kel et al., 2010). We next tested if TGF-β/Smad3 signaling pathway is required for maintaining the immature state of epidermal LCs. As shown in Figure 2a, the frequencies of CD80+ LCs and CD86+LCs were comparable between Smad3 KO and WT mice. Furthermore, the expression levels of MHCII, CD80 and CD86 on LCs, represented by mean fluorescence intensity (MFI), were not changed (Figure 2b). LC maturation in Smad3 KO mice were further evaluated after in vitro epidermal culture. As shown in Figure 2c, after 60h culture, both the percentages of CD86- and CD80-positive LCs and MHCIIhigh LCs as well as the relative CD80, CD86, and MHCII expression levels on LCs (Figure 2d) were significantly increased in Smad3KO and WT mice compared to unstimulation condition (Figure 2a & b), but there was no significant difference between Smad3 KO and WT mice (P > 0.05). Thus, the lack of Smad3 signaling does not affect the immature state of LCs as well as LC maturation. Figure 2 Smad3 is not an essential factor to control epidermis LC maturation and LC uptaking-antigen ability. (a) and (b) Epidermal suspensions freshly isolated from ears and trunk skin of Smad3 KO and WT littermates at 5 weeks old, were stained with anti-Langerin, ... Due to their physical location, LCs acquire and process antigens. To evaluate the role of Smad3 in antigen phagocytic function of LCs, freshly isolated epidermal cells from KO and WT mice were incubated at 37°C or 4°C (as control) with FITC-Dextran for 45 minutes and then stained with anti-mouse MHC-II and CD45.2 antibodies. As shown in Figure 2e, the phagocytic capacity of LCs in Smad3 KO mice had no significant difference (P > 0.05) compared to the WT LCs, based on the ratio of FITC-positive LCs (Figure 2e) or MFI expression levels (Figure 2f). Thus, lack of Smad3 signal pathway does not affect LC phagocytosis. In summary, lack of Smad3 surprisingly does not significantly interrupt the development and immature state of epidermal LCs, and Smad3-deficient LCs have normal maturation and phagocytosis. Our data suggest that Smad3 is not required in the TGF-β signal pathway for ontogeny, homeostasis, and function of epidermal LCs. Recent report indicated that Smad2 and Smad3 were redundantly essential for TGF-β–mediated induction of Foxp3-expressing regulatory T cells and suppression of IFN-γ production in CD4+ T cells (Takimoto et al., 2010). This raises the possibility that Smad2/Smad3 redundancy may also exist in TGF-β/Smads pathways in LCs. In addition, non-TGF-β/Smads pathways may also regulate LC ontogeny and homeostasis (Figure 2g). Further investigations are warranted to clarify the TGF-β signaling pathways through which TGF-β controls LC ontogeny and homeostasis.