Survivin, an important antiapoptotic protein, is expressed in tumors, whereas in normal tissues the expression of this protein is extremely low, defining a role for survivin as a cancer gene. Survivin exhibits multifunctional activity in tumor cells. However, why survivin expression is sharply and invariably restricted to tumor tissue remains unclear. Here, we identified 11 putative consensus binding sites for GLI transcription factors in the survivin promoter and characterized the promoter activity. Inhibitors of the Hedgehog/GLI pathway, cyclopamine and GANT61, decreased the promoter activity in reporter assays. ΔNGLI2 (which lacks the repressor domain) was the most potent vector in activating the survivin promoter–reporter. Moreover, GANT61, a GLI1/2 inhibitor, repressed endogenous survivin protein and mRNA expression in most cells across a large panel of tumor cell lines. Chromatin immunoprecipitation showed GLI2 binding to the survivin promoter. The ectopic GLI2-evoked expression of endogenous survivin was observed in normal human fibroblasts. GANT61 decreased survivin level in nude mice tumors, mimicking the activity of GANT61 in cultured cells. The immunohistochemistry and double immunofluorescence of human tumors revealed a correlation between the tissue regions showing high GLI2 and survivin positivity. Thus, these results demonstrated that survivin is a classical transcriptional target of GLI2, a Hedgehog pathway signaling effector. This potentially reflects the high expression of survivin in human tumor cells. As the Hedgehog pathway is upregulated in virtually all types of cancer cells, these findings substantially contribute to the explanation of uniform survivin expression in tumors as a potential target for the development of a more effective treatment of cancers through the inhibition of GLI2 to restrain survivin activity.

Metastasized malignant melanoma has a poor prognosis because of its intrinsic resistance to chemotherapy and radiotherapy. The central role in the melanoma transcriptional network has the transcription factor MITF (microphthalmia-associated transcription factor). It has been shown recently that the expression of MITF and some of its target genes require the SWI/SNF chromatin remodeling complex. Here we demonstrate that survival of melanoma cells requires functional SWI/SNF complex not only by supporting expression of MITF and its targets and but also by activating expression of prosurvival proteins not directly regulated by MITF. Microarray analysis revealed that besides the MITF-driven genes, expression of proteins like osteopontin, IGF1, TGFß2 and survivin, the factors known to be generally associated with progression of tumors and the antiapoptotic properties, were reduced in acute BRG1-depleted 501mel cells. Western blots and RT-PCR confirmed the microarray findings. These proteins have been verified to be expressed independently of MITF, because MITF depletion did not impair their expression. Because these genes are not regulated by MITF, the data suggests that loss of BRG1-based SWI/SNF complexes negatively affects survival pathways beyond the MITF cascade. Immunohistochemistry showed high expression of both BRM and BRG1 in primary melanomas. Exogenous CDK2, osteopontin, or IGF1 each alone partly relieved the block of proliferation imposed by BRG1 depletion, implicating that more factors, besides the MITF target genes, are involved in melanoma cell survival. Together these results demonstrate an essential role of SWI/SNF for the expression of MITF-dependent and MITF-independent prosurvival factors in melanoma cells and suggest that SWI/SNF may be a potential and effective target in melanoma therapy.

MEK kinase inhibitors (trametinib and selumetinib) or kinase inhibitors directed against mutated BRAF(V600E) (vemurafenib and dabrafenib) have initial encouraging effects in the treatment of melanoma but acquired resistance appears almost invariably after some months. Studies revealed mutually exclusive NRAS and BRAF activating mutations driving the MAPK/ERK pathway among human melanomas. Although combination therapy exerts significantly better antitumor cell efficacy, complete remission is rarely achieved. To employ an alternative approach, we have targeted the Hedgehog/GLI pathway, which is deregulated in melanomas, through the GLI1/2 inhibitor GANT61, alone or accompanied with the treatment by the BCL2 family inhibitor obatoclax in 9 melanoma cell lines. Thus, we targeted melanoma cells irrespective of their NRAS or BRAF mutational status. After GANT61 treatment, the cell viability was drastically diminished via apoptosis, as substantial nuclear DNA fragmentation was detected. In all tested melanoma cell lines, the combined treatment was more efficient than the application of each drug alone at the end of the cell growth with inhibitors. GANT61 was efficient also alone in most cell lines without the addition of obatoclax, which had only a limited effect when used as a single drug. In most cell lines, tumor cells were eradicated after 5-9 days of combined treatment in colony outgrowth assay. To conclude, GANT61 treatment might become a hopeful and effective anti-melanoma targeted therapy, especially when combined with the BCL2 family inhibitor obatoclax.

Melanoma arises from neural crest‐derived melanocytes which reside mostly in the skin in an adult organism. Epithelial–mesenchymal transition (EMT) is a tumorigenic programme through which cells acquire mesenchymal, more pro‐oncogenic phenotype. The reversible phenotype switching is an event still not completely understood in melanoma. The EMT features and increased invasiveness are associated with lower levels of the pivotal lineage identity maintaining and melanoma‐specific transcription factor MITF (microphthalmia‐associated transcription factor), whereas increased proliferation is linked to higher MITF levels. However, the precise role of MITF in phenotype switching is still loosely characterized. To exclude the changes occurring upstream of MITF during MITF regulation in vivo, we employed a model whereby MITF expression was inducibly regulated by shRNA in melanoma cell lines. We found that the decrease in MITF caused only moderate attenuation of proliferation of the whole cell line population. Proliferation was decreased in five of 15 isolated clones, in three of them profoundly. Reduction in MITF levels alone did not generally produce EMT‐like characteristics. The stem cell marker levels also did not change appreciably, only a sharp increase in SOX2 accompanied MITF down‐regulation. Oppositely, the downstream differentiation markers and the MITF transcriptional targets melastatin and tyrosinase were profoundly decreased, as well as the downstream target livin. Surprisingly, after the MITF decline, invasiveness was not appreciably affected, independently of proliferation. The results suggest that low levels of MITF may still maintain relatively high proliferation and might reflect, rather than cause, the EMT‐like changes occurring in melanoma.

The sonic Hedgehog/GLI signaling pathway (HH) is critical for maintaining tissue polarity in development and contributes to tumor stemness. Transcription factors GLI1–3 are the downstream effectors of HH and activate oncogenic targets. To explore the completeness of the expression of HH components in tumor cells, we performed a screen for all HH proteins in a wide spectrum of 56 tumor cell lines of various origin using Western blot analysis. Generally, all HH proteins were expressed. Important factors GLI1 and GLI2 were always expressed, only exceptionally one of them was lowered, suggesting the functionality of HH in all tumors tested. We determined the effect of a GLI inhibitor GANT61 on proliferation in 16 chosen cell lines. More than half of tumor cells were sensitive to GANT61 to various extents. GANT61 killed the sensitive cells through apoptosis. The inhibition of reporter activity containing 12xGLI consensus sites by GANT61 and cyclopamine roughly correlated with cell proliferation influenced by GANT61. Our results recognize the sensitivity of tumor cell types to GANT61 in cell culture and support a critical role for GLI factors in tumor progression through restraining apoptosis. The use of GANT61 in combined targeted therapy of sensitive tumors, such as melanomas, seems to be immensely helpful.

ABSTRACT

In melanoma and other cancers, invasion, epithelialto-mesenchymal transition, metastasis and cancer stem cell maintenance are regulated by transcription factors including the Snail family. Slug (Snail2) protein generally supports migration and apoptosis resistance. However, its role in melanoma is not completely understood. The present study investigated the transcriptional regulation of the SLUG gene in melanoma. It demonstrated that SLUG is under the control of the Hedgehog/GLI signaling pathway and is activated predominantly by the transcription factor GLI2. The SLUG gene promoter contains a high number of GLI-binding sites. Slug expression is activated by GLI factors in reporter assays and inhibited by GANT61 (GLI inhibitor) and cyclopamine (SMO inhibitor). SLUG mRNA levels are lowered by GANT61 as assessed by reverse transcription-quantitative PCR. Chromatin immunoprecipitation revealed abundant binding of factors GLI1-3 in the four subregions of the proximal SLUG promoter. Notably, melanoma-associated transcription factor (MITF) is an imperfect activator of the SLUG promoter in reporter assays, and downregulation of MITF had no effect on endogenous Slug protein levels. Immunohistochemical analysis confirmed the above findings and showed MITF-negative regions in metastatic melanoma that were positive for GLI2 and Slug. Taken together, the results demonstrated a previously unrecognized transcriptional activation mechanism of the SLUG gene, which may represent its main regulation of expression in melanoma cells.

Mutated BRAF oncogene, which fuels unregulated MAPK pathway in most melanomas, can be targeted by a small-molecule drug vemurafenib. However, most of patients develop an acquired resistance during treatment or have an intrinsic resistance to this drug. Nonreceptor SRC tyrosine kinase is deregulated in most cancer cell types and can be efficiently inhibited by a small-molecule inhibitor dasatinib, an anticancer drug which is already used for the treatment of some cancers and is in final clinical trials for other tumor types. Although SRC does not have its own canonical pathway, it is capable of activating and augmenting different signaling routes in cancer cells. We found that melanoma tumor cell lines analysed up to date are generally resistant to dasatinib. The exception to this rule was Hbl, a very sensitive cell line whose cells were completely killed by a 10 nM of dasatinib in a few days. The difference in resistance of melanoma cell lines to dasatinib did not depend on primary versus metastatic cell origin or the presence of BRAF or NRAS mutations. Instead, the dasatinib sensitivity was dependent on its ability to inhibit completely both the MAPK and AKT/mTOR signaling pathways. This inhibition was at least as prominent as the inhibition by specific inhibitors of the two pathways. SRC has been reported previously to have links to this signaling pathways with the capacity to inhibit them in some cell types. Moreover, extensive cross-talk exists between the two pathways. Similar results were observed in several other cell lines derived mostly from lung and colorectal cancers. Our data indicate that the singular therapy of melanoma by dasatinib only would bring no benefit to patients, except for the cases (probably rare) where dasatinib strongly blocks both MAPK and AKT signaling pathways in tumor cells. More molecular knowledge about the effects of the powerful anticancer drug dasatinib is required in order to be used for the treatment of melanoma. This work was supported by grant IGA no. NT/14005-3 from the Ministry of Health (Czech Rep.) and by Institutional research project PRVOUK-P25/LF1/2. Citation Format: Jiri Vachtenheim, Jiri Réda, Lubica Ondrušová, Petra Zakova. The requirements for dasatinib sensitivity in human melanoma cell lines. [abstract]. In: Proceedings of the 105th Annual Meeting of the American Association for Cancer Research; 2014 Apr 5-9; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2014;74(19 Suppl):Abstract nr 5487. doi:10.1158/1538-7445.AM2014-5487