Role of CIZ1 in maintenance of epigenetic landscape in primary differentiated cells

Cells in a developing embryo have the capability to become all of the different types of cell in the adult body. As they specialize they gradually turn off genes that are not required, until they only produce products that are essential to their function. One of the ways that their choice is restricted is by modification of specific proteins that are closely associated with the genes to be shut down. Once established, these modifications, or marks, must be copied each time a cell divides so that daughter cells have the same marks as their parents, and therefore access the same repertoire of genes as their parents. We are studying the mechanism that ensures marks are faithfully copied; the 'quality control' process that stops a cell from drifting away from its chosen path.

Specifically, we are interested in how the chosen genes and the proteins that add the marks (factors) find each other among the mass of DNA packed into the cell nucleus. Do the factors diffuse around the nucleus until they happen to come across the correct genes, or is there a delivery system that ensures 'letters' reach the right 'address' in good time? Our experiments will look at whether, in normal cells, the factors are confined to factory sites, restricting their ability to diffuse around. This would mean that genes must visit factories to be marked, which could afford tight control over the process. So far, there has been very little investigation of this because (our data suggest) restriction to factories appears to be absent or overridden in most of the cell types that are commonly used by researchers (stem cells, cancer cells, cultured cells). In order to understand whether restriction to factories is the norm for non-diseased cells in the body, it is necessary to work with newly isolated cells that have not been adapted to long-term growth in a laboratory. Typically, adapted cells have undergone changes that are similar to changes that take place in cancer cells, and appear to have lost the ability to restrict key factors to factories. Only by studying newly isolated cells will we understand how normal cells maintain their choice of genes, and then be able to understand how this is eroded during the early stages of cancer.

Recently we showed that a protein called CIZ1 is required, in normal cells, for a test chromosome to i) acquire key marks and ii) briefly relocate from one position in the nucleus to another position. We think that these two observations are linked and that, by studying the function of CIZ1 we can establish whether targeted delivery of gene to spatially restricted factor normally operates inside the nucleus of normal cells in the body. We aim to use the results to understand how important CIZ1 is to maintain appropriate marks, and so guard against loss of control and emergence of disease. We are specifically concerned with the effect of alterations in CIZ1 (common in early stage cancers of several different types), as experimental removal of CIZ1 promotes the apparent tendency of cultured cells to loosen spatial separation of gene and factor.

Inaccurate polycomb repressive complex 2 (PRC2) target gene selection and changes in PRC subunit expression have been linked with many types of cancer, particularly hematological malignancies where the data argue for both oncogenic and tumour suppressor function. Changes in CIZ1 are similarly emerging as widespread early events in cellular transformation, again implicating both oncogenic and tumour suppressor activity. Our analysis of CIZ1 null murine somatic cells at early passage (before immortalization) revealed widespread deregulation of genes that are normally under the control of PRC2, including those located on the inactive X chromosome (Xi). However PRC2 subunit levels are not themselves changed. This project will investigate how CIZ1 determines PRC target gene selection, and build on our recent evidence which suggests that PRC2 is normally attached to the core nuclear matrix, and that CIZ1 supports transfer of repressed chromatin toward the interior of the nucleus (likely visiting PRC2 bodies) specifically during S phase. We will use early passage WT cells because our data show that immortalized cells and tumour-derived cells have already sustained changes in PRC2 (release from the nuclear matrix) that facilitate bypass of the process we wish to study. We will use the Xi as a model to study template transfer (TT) because we can follow its movement, during DNA replication, from its typical position at the nuclear periphery into the interior and back again within a 30 minute window. Thus, this project will address i) the contribution of spatial immobilization of PRC bodies and associated TT to fidelity of PRC target gene maintenance in primary somatic cells and ii) whether the process is bypassed by reversion to a less spatially restricted, embryonic-like, process in transformed cells, and iii) the extent to which our common cultured cell models fail to report on what appears to be an unstudied mechanism that helps to maintain epigenetic fidelity in the body.

One of the key impacts that could arise from this work is a greater realisation among cell biologists that disease-linked normal processes are often corrupted in cultured cells, so that what we think of as normal has already begun to diverge. For most questions this may not be a problem but in the area of nuclear structure and organisation, it is of particular concern because relaxation of strict organisation seems to be one of the things that accompanies immortalization. This is still under appreciated because the emergent descriptions of nuclear organisation in cells in long-term culture very often align with descriptions of stem cells. Consequently, we know very little about how the epigenetic landscape is maintained in dividing normal differentiated somatic cells before they reach the Hayflick limit.

Through the accepted routes for dissemination of scientific data and ideas (conferences, papers and reviews), we will use the data generated here to communicate the message that to understand cellular degeneration we need a more robust description of the processes that keep a normal, dividing, somatic cell from escaping its programming. We will address whether as-yet unreported fidelity mechanisms exist, that are non-essential for viability and which may therefore underlie viable, yet disease-prone cells in the body. This concept should not be so hard to convey because of the parallels with DNA repair deficient cells, which are viable but hypermutable. Loss of fidelity mechanisms, as a driver for accelerated epigenetic drift, will be the main concept addressed by this work and has the potential to yield quality control strategies that could be applied to improve cell-based regenerative medicine.

In addition to academic beneficiaries, the wider public could benefit if we uncover information that can be exploited in early stage cancer detection strategies. We have already validated one nuclear matrix protein as a blood biomarker for detection of early stage lung cancer (1), and are making progress translating this into a high-throughput immunoassay for use in hospital settings (2), initially aimed at second line evaluation of CT-positive patients. It is possible that the work proposed here will reveal further markers of early nuclear structure changes, and so contribute to a molecular description of what has been the main method of cancer diagnosis and classification for more than a century (3). Furthermore, a molecular understanding of how nuclear structure becomes compromised may reveal common vulnerabilities that could be exploited therapeutically, or at least help us understand why gross changes to nuclear structure are so common in tumour cells.

1. Higgins, G. et al. Variant Ciz1 is a circulating biomarker for early-stage lung cancer. Proc Natl Acad Sci U S A 109, E3128-3135, doi:10.1073/pnas.1210107109 (2012).
2. Coverley, D. et al. A quantitative immunoassay for lung cancer biomarker CIZ1b in patient plasma. Clinical biochemistry, doi:10.1016/j.clinbiochem.2016.11.015 (2016).
3. Zink, D., Fischer, A. H. & Nickerson, J. A. Nuclear structure in cancer cells. Nature Reviews Cancer 4, 677-687 (2004).