ACTINNUCLSHAPECANCER · Role of nuclear architecture in cancer development

Role of nuclear architecture in cancer development

In this project we aim to answer several questions regarding the crosstalk between cytoskeleton, and specifically actin filaments, and nuclear functions in the pathogenesis and evolution of aggressive traits of cancer. The specific objectives we would like to address in this proposal are: 1) How do alterations in the actin regulation cause nuclear dysmorphia? 2) How these alterations perturb genome stability and/or differentiation? 3) Are changes in nuclear architecture required for actin-dependent cell migration? To address the objectives outlined above, we propose an integrated approach that includes many disciplines. Firstly, we plan to identify molecular mechanisms by which our candidates regulate nuclear shape. Secondly, we will collaborate with experts in genome stability to determine the functional relevance of nuclear shape disruption. Thirdly, we will couple RNAi screen with biomimetic matrixes synthesis to identify regulators of nuclear structure that might be involved in cell migration. The first objective of our project aims at understanding how the cytoplasmic actomyosin network affects the nucleus. Following a siRNA screen performed previously in the Sahai laboratory (Madsen et al Nature Cell Biology 2015), I decided to focus on hits involved in PP1 regulation. Depletion of our candidate genes leads to altered nuclear shape with consequent nuclear rupture in smaller organelles without induction of apoptosis. Analysis of several nuclear and cytoplasmic proteins revealed that partial loss of nuclear envelope integrity caused exchange of nucleo-cytoplasmic content. Treatment of depleted cells with blebbistatin, Y27632 and C3 rescued the phenotype, indicating that overactivation of actomyosin contractility through Rho-ROCK-MYL9 axis downstream of our candidate genes is linked to nuclear dysmorphia. Moreover, the nuclear phenotype is mimicked by a constitutive active, phosphomimetic mutant of MYL9. I next sought the mechanism by which increased actomyosin contractility could affect nuclear shape. In collaboration with the Petronzcki laboratory (Cancer Research UK London Research Institute), we first tested the hypothesis that linker proteins of the nuclear envelope could transmit forces from the actin cytoskeleton to the nucleus, but depletion of SUN1, SUN2 and Nesprin2 had no effect on nuclear shape of cells depleted of our candidate gene. Interestingly, after depletion of our candidate genes we observed formation of actin bundles in the perinuclear region and in the space between nucleus and smaller nuclear fragments. This observation is compatible with the hypothesis that increased contractile forces surrounding the nucleus can squeeze it and cause nuclear lamina breakage. In addition to the analysis of the PP1 regulators described above, I also investigated whether recurrent RhoA mutations in tumours led to similar nuclear defects. This is a pertinent question because RhoA is an upstream regulator of ROCK and MYL9, yet there is no clear explanation as to why RhoA mutations might aid the genesis of tumours. I tested three of the most frequent RhoA mutants (positions 5, 40, and 42) by ectopic expression in the same HeLa Kyoto cells used for the phosphatase analysis. This revealed that the most common cancer-linked RhoA mutants are not able to drive nuclear dysmorphia. To confirm that nuclear dysmorphia was not just an in vitro artifact due to the cortical stress exerted by attachment on plastic, I performed intravital imaging. I used a cell line with a stable labeling of nuclear envelope and DNA localization in vivo (Lap2b-GFP and Histone2B-mCherry). These cells were injected orthotopically together with Matrigel and imaged after tumor formation with in vivo two-photon microscopy. I could observe that nuclear rupture and DNA leakage are not only in vitro phenomena, but they happen in vivo in the context of a primary tumor. Moreover I managed to perturb the actomyosin-nuclear structure in vivo by inhibiting the actomyosin apparatus via topical tumor treatment with the ROCK-inhibitor Y27632. To quantify the dynamic deformation of the nucleus we have measured the variance of geometric properties of the nucleus of time. This analysis revealed that targeting ROCK could reduce dynamic nuclear deformation in vivo. The second objective aimed at understanding if perturbations of nuclear shape affect nuclear functions important in the development of cancer, as genome stability and DNA repair. Moreover, depleted cells showed increased DNA double-stand breaks foci, as visualized by γ-H2AX staining. Further, our collaborators in the Petronzcki lab analyzed the karyotype of cells depleted of our candidate genes and observed an increased frequency of abnormal chromosomes. Importantly, this phenotype was rescued by Y27632 treatment, indicating that increased actomyosin force is causing genomic damage. I am currently investigating if this translates into an altered sensitivity of the cells to chemotherapeutic drugs. In the third objective we planned to perform an RNAi screen to identify regulators of nuclear shape involved in migration through narrow pores. As preliminary experiment we depleted lamin A from cells that cannot migrate through 3μm pores (a size considered smaller than the nuclear size), despite we observed formation of multilobulated nucleus, we could not detect gain of migratory capacity. The same result was obtained cells depleted of our candidate genes, suggesting that knock-down or a single gene might not be sufficient to induce migration through limiting size pores. To circumvent this issue, we changed approach and started from a cell line that already contains clones able to migrate through the pores. We selected clones able to migrate through the pores and obtained a population with distinct morphology. However, this feature, together with the capacity to migrate, was not stable over time, suggesting that might be detrimental for efficient cell growth. To summarize, I made important progresses with objective 1 and 2 and the results have been included in a manuscript currently under submission. We are currently translating our findings in a more cancer-oriented context, such as sensitivity of cells to chemotherapeutic drugs, and we hope to generate sufficient data for another smaller paper.

Data: CORDIS, © European Union