Regulation of the oncogenic potential of signalling through the Ras/ERK pathway by dual-specificity phosphatase 5

Abnormal activation of the Ras/ERK signalling pathway plays a causative role in disease initiation and progression and approximately one third of all human cancers exhibit mutations known to cause Ras/ERK mediated cancer formation. Furthermore, many of these cancers, such as pancreatic cancer, lung cancer and malignant melanoma have few treatment options and a poor prognosis, making this pathway an intense focus of anticancer drug development. We understand a lot about how activating mutations in ERK pathway components, such as Ras or its downstream target, Braf, cause ERK signalling to be uncontrollably "switched on" in tumours, and that such mutations can determine tumour sensitivity to anticancer drugs that inhibit the Ras/ERK pathway. In contrast, we know little about how the negative regulators of Ras/ERK signalling influence oncogenesis or drug sensitivity.

We recently demonstrated that DUSP5, an inducible protein phosphatase, inhibits nuclear Ras/ERK activity and gene expression, and that deletion of DUSP5 dramatically increases mutant Ras-induced skin tumour burdens in mice, demonstrating that DUSP5 is a tumour suppressor. We therefore hypothesise that DUSP5 limits the progression of mutant Ras tumours in other tissues and in human tumours. In contrast, our more recent studies using cultured cells indicate that in the presence of Braf mutations, DUSP5 actually promotes cell growth and its deletion halts cell proliferation. We therefore hypothesise that, much like their effect on some anticancer drugs, the presence of Ras or Braf oncogenes can determine whether DUSP5 promotes or suppresses cancer cell growth. The overall objectives of this research project are two-fold:

1) We will determine the effects of DUSP5 loss in clinically relevant mouse models of mutant Kras-induced cancer, namely pancreatic and lung carcinogenesis. These models closely resemble the human disease and will enable us to assess the role of DUSP5 in modulating the initiation and progression of tumours. This will be coupled with a detailed analysis of changes in ERK-dependent gene expression and signalling endpoints associated with cell proliferation, senescence, survival, migration and cell death. The latter will allow us to identify mechanism(s) and also key targets of ERK signalling involved in promoting tumour development. Finally, we will screen normal and tumour tissue samples from pancreatic cancer patients in order to assess the levels of DUSP5 and other key ERK signalling targets and to explore correlations with clinical endpoints in human disease.

2) We will determine whether the role of DUSP5 in modulating the oncogenic effects of Ras/ERK signalling depends on the way in which the pathway is activated. This will be studied by monitoring the influence of DUSP5 on ERK signalling and cell fate in the presence of either Ras or Braf mutations. This has important implications for predicting how effective Ras/ERK pathway inhibitors may be in the treatment of cancer. We will therefore study the molecular mechanisms which underpin this functional difference and extend these studies to cancer cell lines in which Ras or Braf are mutant and determine if DUSP5 expression affects sensitivity either acute or chronic exposure to Ras/ERK inhibitors in clinical use.

Overall, this work will inform us about DUSP5 function in the regulation of Ras/ERK signalling as it relates to cancer initiation and development, will shed light on the relative importance of the Ras/ERK pathway and its targets in cancers driven by mutant Kras and allow us to explore their clinical relevance. Finally, the idea that the role of DUSP5 in regulating ERK may become more important in the presence of activating oncogenes represents an unexplored area in terms of cellular responses to pathway inhibitors and may open new approaches to modulating or predicting patient responses to drugs which target Ras/ERK signalling in cancer.

Tumours in which the Ras/ERK pathway is abnormally activated, such as pancreatic and lung cancers are relatively common, but have few treatment options and a poor prognosis. We recently discovered that DUSP5, a protein phosphatase which "switches off" Ras/ERK signals in the nucleus, plays a key role in preventing tumours in a mouse model of skin cancer, demonstrating that DUSP5 is a bona fide tumour suppressor. We therefore hypothesise that DUSP5 plays an important role in cancer development driven by Ras/ERK signalling in other tissue types in mouse models and in humans.

We will test this hypothesis by addressing two main objectives.
1) Firstly, we will establish if DUSP5 influences the genesis and progression of tumours in clinically relevant experimental models of Kras-induced pancreatic and lung cancer. We will analyse both tumours and cells derived from these cancers to determine the molecular mechanisms by which DUSP5 affects the oncogenic potential of Ras/ERK signalling and study the expression of DUSP5 in human tumour samples to establish if the levels of this protein are predictive of disease severity or progression.
2) Our preliminary data comparing the response to DUSP5 loss in cells containing either mutant Ras or mutant Braf suggest that the role of DUSP5 changes depending on the nature of the oncogene that drives cancer and may modify the response of cancer cells to drugs that inhibit the Ras/ERK pathway. Our second objective is therefore to determine how DUSP5 function is altered by Ras or Braf mutation and its implications for mediating drug response and drug resistance.

This project will provide new mechanistic insights into the ways that the Ras/ERK signalling pathway is "rewired" in cancer and may reveal new biomarkers which will allow the stratification of patients receiving drugs which inhibit Ras/ERK signalling and reveal potential new strategies to prevent the emergence of drug resistance.

Outside of academia, biotechnology/diagnostics companies and the pharmaceutical industry will be the major beneficiaries of our proposed research, but we would hope that cancer patients and the wider NHS will benefit in the longer term.

Patients who suffer from cancers in which tumour growth is driven by activation of the Ras/ERK pathway make up a substantial proportion of the UK cancer burden with just mutant Kras-driven non-small cell lung cancer (NSCLC) and pancreatic cancer together accounting for approximately 15% of the 331,487 new cases in 2011. In addition, the 10-year survival after diagnosis for these cancers is relatively poor ranging from 1-5%. Thus there is an enormous research effort worldwide aimed at gaining a greater understanding of the architecture, regulation and function of Ras/ERK signalling and how this knowledge might be exploited in the optimisation of existing or development of novel anti-cancer therapies.

As part of our work to explore the wider significance of DUSP5 in more clinically relevant mouse tumour models (pancreas and lung) we will also be searching for ERK-mediated endpoints or biomarkers of pathway activity that contribute to the oncogenic potential of Ras/ERK-mediated signalling. These include SepinB2, which we identified as a DUSP5 regulated protein involved in promoting skin cancer. We will also be looking at these markers in matched normal and tumour tissues from pancreatic cancer patients along with the levels of DUSP5 itself, as this gene is reported to be silenced in gastric cancer. Such biomarkers may be of clinical use in diagnosis or as prognostic indicators as well as permitting stratification of patients in terms of drug regimes as step towards "personalised" anti cancer therapy.

In the private sector most pharmaceutical companies, most notably Astra Zeneca (AZ) in the UK, have major programmes of research centred on novel biological and chemical inhibitors of Ras/ERK signalling. These include inhibitors of receptor tyrosine kinases (Iressa and AZD9291), BRaf (dabrafenib) and MAPK kinase or MEK (selumetinib). Our preliminary work indicates that DUSP5 may play quite different roles in regulating Ras/ERK signalling depending on the nature of the activated oncogene that is driving cancer cell growth. This has profound and hitherto unexplored implications for both tumour cell behaviour and the potential response of cancer cells with different levels of DUSP5 expression to drugs such as selumetinib. DUSP5 may also influence patient acquired resistance to such drugs, as this often involves reactivation of Ras/ERK signalling and would be expected to involve DUSP5 as a transcriptional target of the pathway. In early 2015 we established collaborative links with Dr. Paul Smith in the Molecular Oncology team at AZ through the "Open Innovation" scheme to expedite progression and dialogue, ensuring we have the most up to date information on lead compounds. This demonstrates our results are already of interest and benefit to scientists trying to unravel how feedback loops control ERK pathway activity and determine drug sensitivity, and will thus have immediate impact during the course of the project. It may also become clear from our work that DUSP5 is a potential drug target itself and although this class of phosphatases are not currently considered "druggable" the advent of more sophisticated screens for agents that perturb protein-protein interactions, rather than enzyme activities, may open up new avenues in this regard.