Therapeutic targeting of Hedgehog signalling pathway in T-cell acute lymphoblastic leukemia (T-ALL)

T-cell acute lymphoblastic leukemia (T-ALL) is an aggressive hematological tumour arising from the malignant transformation of T-cell precursors. Notwithstanding intensified therapy, ~40% of adult and ~20% of pediatric patients face a dismal prognosis due to primary resistance to treatment and relapse, raising the need for more efficient and targeted therapies. Hedgehog (HH) signalling is a major developmental pathway frequently deregulated in cancer, for which a role in T-ALL is recently emerging. The therapeutic targeting of HH signalling in T-ALL by specific inhibitors in vitro and in vivo was shown to be partially effective when used as monotherapy and seems to suggest an important role for the non-canonical activation of the pathway (i.e. independent of upstream Smoothened receptor). This noncanonical activation of GLI1 transcription factor, the main HH downstream effector, underscores the necessity of dissecting the complex regulatory network upstream of GLI1 in T-ALL. In this study, we evaluated gene and protein expression of HH pathway components in T-ALL cell lines, patient-derived xenografts (PDX) and Notch1-dependent T-ALL murine models, confirming their expression in the majority of the tested samples. A significant fraction of T-ALL cell lines and PDX-derived cells were sensitive to HH pathway inhibition by GANT61 (GLI1/2 inhibitor), much less so to cyclopamine (Smoothened inhibitor). Interestingly, in a Notch1-induced T-ALL murine model, pharmacological combinations targeting commonly deregulated oncogenic pathways in T-ALL revealed a therapeutically relevant synergism between GANT61 and glucocorticoids (GCs). Combined treatment of T-ALL cell lines and PDXderived cells with GANT61 plus Dexamethasone (Dexa) showed a synergistic anti-leukemic effect, affecting both cell proliferation and survival. In order to gain mechanistic insights of the found crosstalk between the HH pathway and the GC receptor pathway, we studied the impact of synthetic GCs on GLI1 function. GCs impaired GLI1 transcriptional activity in transfected HEK293T cells stably expressing the GC receptor; however, gene expression data and Western blot analysis seemed to exclude a transcriptional effect of GCs on GLI1, but rather suggested a post-translational mechanism of action. Cell fractionation analysis did not reveal significant re-distribution of GLI1 upon Dexa treatment, while GLI1 stability was shown to be impaired, revealing a reduced protein half-life after treatment. Finally, GLI1 and the GC receptor were shown to interact, with the GC receptor recruiting the acetyltransferase PCAF and dissociating from the deacetylase HDAC1 upon Dexa treatment, thus leading to GLI1 hyperacetylation and reduced transcriptional activity. In conclusion, we demonstrated that HH pathway is active in a subset of TALLs and the differential sensitivity to HH inhibitors suggests a ligandindependent non-canonical mechanism of activation. We also collected evidence of a crosstalk between the GC receptor and HH pathway, with the GC receptor acting as a negative regulator of GLI1 transcription factor, setting the therapeutic rationale for combining GLI1 inhibitors and GCs.