Repository logo

A novel route to oncogenic activation of cell cycle kinase Aurora A



Change log


Cacioppo, Roberta 


Aurora kinase A (AURKA) is a major positive regulator of the cell cycle, required for the onset of mitosis and the completion of cell division. Recent research has uncovered roles of AURKA that are independent of the kinase activity and that regulate multiple cellular processes, including motility, senescence, and transcription. Therefore, control of cellular abundance of AURKA protein is crucial for the correct execution of its functions. For this, multiple mechanisms are normally in place at different steps of gene expression to ensure that AURKA levels are tightly fine-tuned during each phase of the cell cycle.

A prominent association exists between high expression of AURKA and cancer, and AURKA gene is classifiable as oncogene. AURKA is indeed a highly attractive target of anti-cancer drugs, in particular small molecule kinase inhibitors. The presence of kinase-independent roles of AURKA however strongly advocates for novel targeting approaches. Anti-cancer strategies that instead aim to reduce AURKA expression levels convincingly diminish its oncogenic potential, although they are not yet close to clinical use. Moreover, the development of therapeutic small interfering RNAs against AURKA messenger RNA (mRNA) has never taken into account that this exists in multiple different isoforms, which still remain poorly investigated in their individual physiological or pathogenic role.

The activation of AURKA oncogene by means of dysregulated gene expression is known to stem from gene amplification, enhanced transcription, or increased protein stability. It has however become clear that virtually every molecular process that controls AURKA levels potentially triggers its oncogenic activation, including impaired mRNA processing, decay, and translation. Regardless, it is surprising how the past decade has neglected fundamental questions about the modulation of AURKA gene expression, particularly at the post-transcriptional level, to rather focus on the functions and regulation of AURKA protein.

The work presented here aimed to fill this knowledge gap with explorations of processes regulating AURKA mRNA, in particular mRNA alternative polyadenylation (APA), targeting by microRNA (miRNA), and translation. Preliminary unpublished data shared by Dr. Begum Akman and colleagues pointed to a switch in AURKA APA as a feature of Triple Negative Breast Cancer correlating with poor patient prognosis. These processes were therefore researched with the intent to assess whether they can offer a basis for AURKA oncogenic overexpression. Furthermore, given the strict cell cycle-dependent pattern of expression of AURKA, these processes were also investigated in light of the cell cycle.

In this Thesis, the mechanism of APA of AURKA mRNA was initially examined. Experiments using human cell culture confirmed that AURKA mRNA undergoes APA, which generates two mRNA isoforms differing in the length of the 3’ untranslated region (3’UTR). In order to investigate if and how the length of 3’UTR contributed to regulation of AURKA expression, I created a novel fluorescence-based single cell reporter of gene expression. Experiments of time-lapse imaging using this reporter in living cells revealed that the short mRNA isoform produces more protein compared to the long isoform. Subsequently, I developed a novel biochemical assay to probe translational efficiency of the different 3’UTRs. Results from this assay indicated that the increased expression of the short isoform is due to its higher translation rate.

In order to profile UTR-dependent translation rate over the cell cycle in live cells, I devised a fluorescence-based assay to simultaneously monitor translation rate and cell cycle phase in single cells. Experiments using this assay led to the discovery that translation rate of the long AURKA mRNA isoform is targeted by hsa-let-7a miRNA, a known tumour suppressor. hsa-let-7a was in fact found to regulate the cell cycle periodicity in translation of the long isoform such that translation was suppressed in the early interphase. In contrast, translation rate of the short isoform was detected high and constant throughout the cell cycle, as it lacks the sequence element for binding and regulation by hsa-let-7a.

The differential translational regulation of AURKA APA isoforms implicated the abundance ratio of the two isoforms as a key element defining AURKA expression levels. Accordingly, I used CRISPR/Cas9 editing to manipulate expression of endogenous AURKA mRNA for production of the short isoform only. This experiment revealed that impaired APA of AURKA mRNA is sufficient to cause AURKA overexpression and promote cancer-like cellular phenotypes. The mutated cell lines were in fact characterized by increased rates of cellular proliferation and migration.

Finally, a bioinformatic analysis of AURKA expression at the level of both the protein and mRNA across 18 human solid cancers, using public datasets from The Cancer Genome Atlas, provided interesting insights on cancer-specific features of AURKA expression.

In summary, this Thesis describes the discovery of a new mechanism dependent on the cooperation between APA and miRNA targeting that contributes to the control of endogenous human AURKA levels. This mechanism is likely to be a route of oncogenic activation of AURKA when dysregulated, especially when APA is impaired. In addition, the findings presented here shed light on the dependency of AURKA translation on the cell cycle, an area which had remained uncertain until now.





Lindon, Catherine


3’UTR isoforms, cell cycle, miRNA, oncogenic Aurora Kinase A, translation


Doctor of Philosophy (PhD)

Awarding Institution

University of Cambridge