< Back to previous page

Project

Targeting the lineage-specific long non-coding RNA LINC01212 as an effective anti-melanoma strategy.

Melanomas are aggressive malignant neoplasms of melanocytic origin, highly resistant to both chemotherapy and radiotherapy. As such, melanoma was viewed as an incurable beast for many decades and patients had virtually zero therapeutic options available. Fortunately, in the last few years major breakthroughs led to the introduction of effective therapies which significantly increased patients’ survival. Nonetheless, despite the initial enthusiasm, it rapidly became clear that melanomas could readily adapt to these therapies and form treatment-resistant clones which would ultimately lead to patients’ death. Moreover, a large fraction of the melanoma patients resulted intrinsically resistant to some of these therapies, leaving many patients without effective therapeutic options. Therefore, there is still a dire need to find novel druggable targets that could allow the design of new therapeutic strategies. Ideally, such therapies should allow to treat patients untreatable with, or unresponsive to, current treatments, but should also be combinable with current therapeutic strategies in order to lead to a complete and definitive cure or, at the very least, to postpone patients’ relapse. Furthermore, apart from a potent anti-melanoma effect, these therapies should also be specific for tumour cells, in order to minimise adverse events, one of the major drawbacks of the existing therapeutic options, thereby increasing patients’ quality of life.

 

In an effort to identify new druggable targets, I focused my attention to the non-coding portion of the genome, that is, DNA sequences that do not code for proteins. Several independent studies had shown that the vast majority of cancer-related mutations occurs in the non-coding genome, rather than in protein-coding genes, thus suggesting a potential role for these regions in the process of tumour formation and development. The non-coding DNA represents the vast majority of the DNA (nearly 99%) and contains all the sequences necessary for gene regulation and transcription but also serves as a template for the transcription of vast numbers of non-coding RNAs, that is, RNA sequences that are not translated. I decided to focus my attention on one particular subclass of non-coding RNAs, namely long non-coding RNAs (lncRNAs), RNA molecules longer than 200 nucleotides that do not code for proteins. Most lncRNAs have very restricted expression profiles, both in normal tissues and in tumour samples, making them very appealing both for diagnostic purposes and for therapeutic applications. Moreover, lncRNAs have been involved in virtually every single cellular process and many lncRNAs have been implicated in cancer development and progression. Importantly, lncRNAs can be easily targeted both in vitro and in vivo with antisense oligonucleotides, a sine qua non when envisioning therapeutic applications.

 

In order to discover lncRNAs important in the context of melanoma development and progression I decided to assess whether frequently amplified regions in melanoma harboured lncRNAs. I chose to start from recurrently amplified regions since I hypothesised that, if a lncRNA would be necessary for melanoma development, chances where that it could be amplified in melanoma samples. I thus found that the frequently-amplified 3p13-3p14 region, a region associated with poor prognosis in melanoma, where the melanoma-specific oncogene MITF resides, also encodes a lncRNA, SAMMSON, which is always co-gained with MITF. SAMMSON is a primate-specific lncRNA expressed in 95% of the melanoma cases but undetectable in normal human melanocytes or in other adult human tissues or tumours. Importantly, SAMMSON is upregulated in the transition between the radial growth phase to the vertical growth phase, during which melanoma cells acquire malignant and invasive features, suggesting that SAMMSON could possibly play a role in melanoma malignant transformation.

 

To gain some mechanistical insight on SAMMSON’s role in melanoma I pulled-down the endogenous SAMMSON transcript and analysed its protein interactome with mass spectrometry. I thus identified three proteins, namely XRN2, CARF and p32, as the strongest SAMMSON interactors. I demonstrated that all three proteins bind on the 3’-end of the SAMMSON transcript, thus raising the possibility that SAMMSON could be a hub for ribonucleoproteic complex formation. Intriguingly, one of the most prominent roles of XRN2 and p32 is to regulate ribosomal RNA (rRNA) processing, the former in the nucleoli and the latter in the mitochondria, and thus, by extension, to regulate protein synthesis either in the cytosol, in the case of XRN2, or in the mitochondria, in the case of p32. Importantly, XRN2 nuclear localisation is dynamically regulated by CARF, which titrates XRN2 excess by retaining it in the nucleoplasm to fine-tune pre-rRNA maturation. Given that the three different interactors were involved in rRNA processing and, by extension, in proteins synthesis, I speculated that SAMMSON could also control these processes.

 

I was able to conclusively demonstrate that SAMMSON can regulate the localisation of its three interactors, and that SAMMSON expression leads to an accumulation of XRN2 in the nucleoli, of CARF in the cytosol and of p32 in the mitochondria. In doing so, SAMMSON is able to affect their functions and thus promote rRNA processing. SAMMSON can achieve this goal by promoting the melanoma-specific interaction between CARF and p32 at the expense of the interaction between XRN2 and CARF. By increasing rRNA processing in both compartments, SAMMSON is able to promote protein synthesis in the cytosol and in the mitochondria and thus lead to an increase in melanoma cell growth both in vitro and in vivo.

 

Depletion of SAMMSON leads to a reduction of p32 levels in the mitochondria and to a decrease in protein synthesis. These events are followed by a sudden loss of the mitochondrial membrane potential and a diminution of the mitochondrial oxidative phosphorylation and, as a consequence, to an impairment in the ATP production. Importantly, while the mitochondria can adjust their protein synthesis rate to finely balance the income of cytosolic proteins, sudden variations of mitochondrial translation will invariably induce an adaptive response that, depending on the duration and type of stress, can result in cell cycle arrest or cell death. Accordingly, SAMMSON silencing drastically decreases the viability of melanoma cells irrespective of their mutational background or of the tumour phenotype and increases sensitivity to MAPK-targeting therapeutics both in vitro and in patient-derived xenograft models.

 

Altogether, the work herein presented shows that SAMMSON plays a pivotal role in melanoma biology and that a single lncRNA can have a major influence on melanoma growth and survival. I identified a key function for SAMMSON in nuclear-mitochondrial communication by means of its ability to rewire the RNA-binding protein network to promote a balanced increase in ribosome biogenesis and protein synthesis in both the cytosolic and mitochondrial compartments. Importantly, this role is independent of the cellular context and confers the cells a growth advantage both in vitro and in vivo. These data highlight how the aberrant expression of a single oncogenic lncRNA can solitarily hijack both the cytosolic and the mitochondrial protein synthesis machineries to enhance their activities and to promote cell growth without inducing proteotoxic stress. Furthermore, my results indicate that targeting SAMMSON affects mitochondrial homeostasis in a melanoma-specific manner and is therefore expected to deliver highly effective and tissue-restricted therapeutic responses. I speculate that SAMMSON could be used both as a diagnostic marker of early melanoma malignant transformation and as a therapeutic target, either alone or in combination with MAPK-inhibitors and possibly with immune-checkpoint inhibitors. As such SAMMSON inhibition may offer a novel rational therapeutic avenue for melanoma treatment.

 

Date:8 Dec 2014 →  31 Dec 2019
Keywords:lineage-specific, long non-coding RNA, SAMMSON, LINC01212, melanoma, mitochondria, p32 XRN2 CARF
Disciplines:Morphological sciences, Oncology
Project type:PhD project