STAT3 promotes melanoma metastasis by CEBP-induced repression of the MITF pigmentation pathway
Preprint posted on September 20, 2018 https://www.biorxiv.org/content/early/2018/09/20/422832
Melanoma is the most aggressive form of skin cancer but despite advances in melanoma research, the prognosis for advanced melanoma remains poor, as tumours are highly resistant to chemotherapy. One reason for this is that melanomas can switch their phenotype between highly proliferative but non-invasive, to slowly proliferating and highly invasive, thus evading the actions of drugs and continuing to grow and metastasise.
A key genetic factor in this melanoma phenotype switching is the transcription factor MITF, which controls the development, differentiation and pigmentation of melanocytes, the cells from which melanomas occur. Melanomas with high MITF levels are characteristic of having the highly proliferative but non-invasive phenotype, whereas MITF-low melanomas proliferate slowly but are highly metastatic.
Therefore, understanding the regulation of MITF and its targets is important when designing therapies to halt melanoma growth and progression. A promising candidate for such a regulator is the transcription factor STAT3. 50% of melanoma patients have enhanced activation of STAT3, and in vitro, STAT3 activation causes cells to acquire melanoma-initiating and stem cell-like properties, thus promoting melanoma growth. However, its precise role in vivo, and whether directly it regulates gene expression – particularly MITF – remains unclear.
To investigate the effect of STAT3 loss in melanoma, the authors conditionally deleted the Stat3 gene in an existing genetic mouse model of human cutaneous melanoma to obtain the line Tyr::NRASQ61K;Ink4a-/- ; Stat3flox/flox; Tyr::Cre, which they termed Stat3Δ. Control mice were from the same genetic background, but lacked Tyr::Cre and so expressed Stat3 at normal levels.
Stat3Δ mice developed tumours significantly earlier than controls, and these tumours were reminiscent of MITF-high melanomas as they were more proliferative and pigmented, but non-invasive. Conversely, control mice developed more micrometastases in their lungs, phenotypically characteristic of MITF-low melanomas (see Figure below). Xenograft studies and in vitro invasion assays using Stat3Δ-derived cells supported these findings.
From preprint Figure 1F, with permission from the authors. Stat3 deletion reduces the formation of metastatic lung colonies (arrowed). Sections of 40 week old Stat3Δ mouse (termed Stat3fl here) lungs were stained with the melanoma marker S100b.
This led the authors to hypothesise that STAT3 activity might influence melanoma phenotype switching by regulating MITF expression. To test this, the authors compared the transcriptomes of cells isolated from Stat3Δ tumours to controls through analysis of a microarray dataset. Indeed, Stat3Δ cells had strongly upregulated levels of MITF and its targets, and in general, their transcriptome resembled MITF-driven proliferation signatures found in previous melanoma cohort studies.
Next, to explore the mechanism of this MITF upregulation in Stat3Δ cells, the authors performed ATAC-seq to see whether STAT3 or its downstream targets might directly bind to the Mitf promoter region to influence gene expression. Interestingly, an accessible binding region for the STAT3 targets CEBPA and CEBPB was found within the Mitf promoter region in control cells, but not in Stat3Δ cells. Moreover, rescuing Cebpa levels in Stat3Δ cells led to Mitf downregulation. This suggests that STAT3 deletion in melanoma causes epigenetic changes to chromatin, which leads to large changes in gene expression and tumour behaviour as a result of its targets CEBPA/B repressing the MITF pathway.
Finally, the authors showed that these findings have clinical relevance in human melanoma. shRNA-mediated knockdown of STAT3 in human melanoma cell lines recapitulated changes in gene expression seen in Stat3Δ cells. The authors also searched publicly available cancer genome databases to confirm STAT3-low/MITF-high expression signatures are found in humans. Interestingly, patients with poorer prognoses were more likely to have high levels of MITF, and low STAT3, CEBPA and CEBPB levels. This suggests that the balance between STAT3 and MITF in different stages of melanoma is important, and this could be a good target for designing new therapies.
Why I chose this preprint
Working in a Cancer Research institute, it is always disheartening to learn about how quickly melanoma patients develop resistance to treatment, and how aggressively their tumours can reoccur, despite huge efforts to search for chemo and immunotherapies to target and kill melanoma. From this preprint, and other studies on MITF in melanoma, it is worrying to learn that common therapies like BRAF inhibitors cause a reduction in MITF levels, and therefore could actually worsen outcomes if used in isolation. Therefore, I think this preprint highlights a potential target for regulating MITF levels in melanoma, but also shows there are huge challenges ahead in correctly balancing and timing drug treatment to prevent unwanted activation of genes which can exert large effects on tumour growth and survival.
Open questions and future directions
- The authors conclude that increased STAT3 activity drives metastasis through performing the ‘opposite’ experiments – deleting STAT3 and comparing it to wild type levels. I wonder if the authors have tried overexpressing STAT3 in a mouse mutant or cell line to show enhanced invasiveness compared to control and/or Stat3Δ tumours?
- I also wonder whether inducible deletion of STAT3, or the use of a STAT3 inhibitor in mice with established melanoma might cause a reversal of metastatic behaviour and shrink primary tumours? Or might it drive proliferation of the existing metastases and accelerate melanoma progression?
- This might be unanswerable yet, and so a bit of an unfair question, but how would one go about inhibiting the MITF and STAT3 pathways to treat melanoma? Would targeting one pathway after another be the best approach, or would it be better to target both pathways simultaneously to ‘cancel out’ their effects?
- Kemper, K et al, 2014. Phenotype Switching: Tumor Cell Plasticity as a Resistance Mechanism and Target for Therapy. Cancer Research. 74(21)5937-5941
- Ohanna, M. et al., 2013. Secretome from senescent melanoma engages the STAT3 pathway to favour reprogramming of naïve melanoma towards a tumor-initiating cell phenotype. Oncotarget, 4(12)2212-2224
Posted on: 8th October 2018Read preprint
Also in the cancer biology category:
Quantification of microenvironmental metabolites in murine cancer models reveals determinants of tumor nutrient availability
|Selected by||Maria Rafaeva|
Mechanical Stretch Kills Transformed Cancer Cells
|Selected by||Joseph Jose Thottacherry|
Inactive USP14 and inactive UCHL5 cause accumulation of distinct ubiquitinated proteins in mammalian cells
|Selected by||Mila Basic|
A conserved MFS orchestrates a subset of O-glycosylation to facilitate macrophage dissemination and tissue invasion
|Selected by||Giuliana Clemente|
MRE11-RAD50-NBS1 activates Fanconi Anemia R-loop suppression at transcription-replication conflicts
|Selected by||Katie Weiner|
Applications, Promises, and Pitfalls of Deep Learning for Fluorescence Image Reconstruction
|Selected by||Romain F. Laine|
Basal extrusion drives cell invasion and mechanical stripping of E-cadherin
|Selected by||William Hill|
Molecularly distinct models of zebrafish Myc-induced B cell leukemia
|Selected by||Hannah Brunsdon|
The landscape of antigen-specific T cells in human cancers
|Selected by||Rob Hynds|
Unjamming overcomes kinetic and proliferation arrest in terminally differentiated cells and promotes collective motility of carcinoma.
|Selected by||Tim Fessenden|
Memory sequencing reveals heritable single cell gene expression programs associated with distinct cellular behaviors
|Selected by||Leighton Daigh|
Anti-angiogenic effects of VEGF stimulation on endothelium deficient in phosphoinositide recycling
|Selected by||Coert Margadant|
Profiling the surface proteome identifies actionable biology for TSC1 mutant cells beyond mTORC1 signaling
|Selected by||Rob Hynds|
Precise tuning of gene expression output levels in mammalian cells
|Selected by||Tim Fessenden|
Moving beyond P values: Everyday data analysis with estimation plots
|Selected by||Gautam Dey|
JNK-mediated spindle reorientation in stem cells promotes dysplasia in the aging intestine
|Selected by||Maiko Kitaoka|