Search

Hedgehog signaling plays pivotal roles in embryonic development, adult homeostasis and tumorigenesis. However, its engagement in the pituitary gland has been long underestimated although Hedgehog signaling and pituitary embryogenic development are closely linked. Thus, deregulation of this signaling pathway during pituitary development results in malformation of the gland. Research of the last years further implicates a regulatory role of Hedgehog signaling in the function of the adult pituitary, because its activity is also interlinked with homeostasis, hormone production, and most likely also formation of neoplasms of the gland. The fact that this pathway can be efficiently targeted by validated therapeutic strategies makes it a promising candidate for treating pituitary diseases. We here summarize the current knowledge about the importance of Hedgehog signaling during pituitary development and review recent data that highlight the impact of Hedgehog signaling in the healthy and the diseased adult pituitary gland.

Cutaneous squamous cell carcinoma (cSCC) is the second most common skin tumor in humans. Although current therapies are sufficient to clear the tumor in many cases, the overall risk of cSCC metastasis is still 5%. Alternative treatment options could help to overcome this situation. Here we focused on the role of the Hedgehog (HH) signaling pathway and its interplay with epidermal growth factor receptor (EGFR) signaling in cSCC. The analyses revealed that, despite lack of Sonic HH (SHH) expression, a subset of human cSCC can express GLI1, a marker for active HH signaling, within distinct tumor areas. In contrast, all tumors strongly express EGFR and the hair follicle stem cell marker SOX9 at the highly proliferative tumor-stroma interface, whereas central tumor regions with a more differentiated stratum spinosum cell type lack both EGFR and SOX9 expression. In vitro experiments indicate that activation of EGFR signaling in the human cSCC cell lines SCL-1, MET-1, and MET-4 leads to GLI1 inhibition via the MEK/ERK axis without affecting cellular proliferation. Of note, EGFR activation also inhibits cellular migration of SCL-1 and MET-4 cells. Because proliferation and migration of the cells is also not altered by a GLI1 knockdown, GLI1 is apparently not involved in processes of aggressiveness in established cSCC tumors. In contrast, our data rather suggest a negative correlation between Gli1 expression level and cSCC formation because skin of Ptch+/- mice with slightly elevated Gli1 expression levels is significantly less susceptible to chemically-induced cSCC formation compared to murine wildtype skin. Although not yet formally validated, these data open the possibility that GLI1 (and thus HH signaling) may antagonize cSCC initiation and is not involved in cSCC aggressiveness, at least in a subset of cSCC.

The development of skeletal muscle from immature precursors is partially driven by canonical WNT/β-catenin signaling. Rhabdomyosarcomas (RMS) are immature skeletal muscle-like, highly lethal cancers with a variably pronounced blockade of muscle differentiation. To investigate whether canonical β-catenin signaling in RMS is involved in differentiation and aggressiveness of RMS, we analyzed the effects of WNT3A and of a siRNA-mediated or pharmacologically induced β-catenin knock-down on proliferation, apoptosis and differentiation of embryonal and alveolar RMS cell lines. While the canonical WNT pathway was maintained in all cell lines as shown by WNT3A induced AXIN expression, more distal steps including transcriptional activation of its key target genes were consistently impaired. In addition, activation or inhibition of canonical WNT/β-catenin only moderately affected proliferation, apoptosis or myodifferentiation of the RMS tumor cells and a conditional knockout of β-catenin in RMS of Ptchdel/+ mice did not alter RMS incidence or multiplicity. Together our data indicates a subordinary role of the canonical WNT/β-catenin signaling for RMS proliferation, apoptosis or differentiation and thus aggressiveness of this malignant childhood tumor.

Rhabdomyosarcoma (RMS) is the most common pediatric soft tissue sarcoma with poor prognosis. RMS frequently show Hedgehog (HH) pathway activity, which is predominantly seen in the embryonal subtype (ERMS). They also show activation of Phosphatidylinositol-4,5-bisphosphate 3-kinase (PI3K) signaling. Here we compared the therapeutic effectiveness and the impact on HH target gene expression of Smoothened (SMO) antagonists with those of the PI3K inhibitor pictilisib in ERMS with and without mutations in the HH receptor Patched1 (PTCH). Our data demonstrate that growth of ERMS showing canonical Hh signaling activity due to Ptch germline mutations is efficiently reduced by SMO antagonists. This goes along with strong downregulation of the Hh target Gli1. Likewise Ptch mutant tumors are highly responsive toward the PI3K inhibitor pictilisib, which involves modulation of AKT and caspase activity. Pictilisib also modulates Hh target gene expression, which, however, is rather not correlated with its antitumoral effects. In contrast, sporadic ERMS, which usually express HH target genes without having PTCH mutation, apparently lack canonical HH signaling activity. Thus, stimulation by Sonic HE (SHH) or SAG (Smoothened agonist) or inhibition by SMO antagonists do not modulate HH target gene expression. In addition, SMO antagonists do not provoke efficient anticancer effects and rather exert off-target effects. In contrast, pictilisib and other PI3K/AKT/mTOR inhibitors potently inhibit cellular growth. They also efficiently inhibit HH target gene expression. However, of whether this is correlated with their antitumoral effects it is not clear. Together, these data suggest that PI3K inhibitors are a good and reliable therapeutic option for all ERMS, whereas SMO inhibitors might only be beneficial for ERMS driven by PTCH mutations.