Elucidating the expression pattern and role of the oncogenic GLI transcription factors in basal cell carcinoma

Abstract

Basal Cell Carcinoma (BCC) is the most common type of skin cancer; its development is associated with mutational inactivation of the PTCH1 tumour suppressor gene leading to deregulation of the Hedgehog pathway through increased expression of the downstream GLI (GLI1 and GLI2) transcription factors. The high expression of GLI1 and GLI2 has also been shown to be maintained through a positive feedback mechanism in human keratinocytes, but this has partly relied upon mRNA expression and has not been fully explored at the protein level. Moreover, expression analysis in human BCCs has largely relied upon in situ hybridization or RT-PCR and no studies have investigated co-expression of the GLI1 and GLI2 proteins. Whereas an active mutant of GLI2 (GLI2N) has been shown to increase endogenous GLI1 protein expression in normal human keratinocytes, preliminary evidence suggests that ectopic GLI1 suppresses endogenous GLI2 protein expression, which refutes the presence of a positive GLI feedback loop. In human BCC samples, there was a relatively high degree of heterogeneity in the expression pattern of GLI1 and GLI2 with the various antibodies employed and this was also affected by how the tissue was processed i.e. frozen versus formalin-fixed. Surprisingly, many BCCs displayed strong epidermal GLI1 expression that was decreased in the basal layer and tumour islands. In most tumours GLI1 and GLI2 showed areas of co-expression, but this was not uniform and in many areas one protein was more abundant than the other. Moreover, there was no clear difference in the expression level of GLI1 or GLI2 between BCC subtypes (aggressive versus non-aggressive). Finally, one antibody (GLI1 H-300) detected strong GLI1 expression in mesenchymal cells surrounding tumour islands (in frozen samples) but this did not correlate with epithelial Sonic Hedgehog ligand expression, suggesting a lack of paracrine Hedgehog signalling in BCC biology. In summary, this project has shown that GLI regulation and expression are more complex than previous studies have suggested and that research in this field is hampered by a lack of consistency and reliability of commercial GLI antibodies.