Search

Background Histone modification H4K20me3 and its methyltransferase SUV420H2 have been implicated in suppression of tumorigenesis. The underlying mechanism is unclear, although H4K20me3 abundance increases during cellular senescence, a stable proliferation arrest and tumor suppressor process, triggered by diverse molecular cues, including activated oncogenes. Here, we investigate the function of H4K20me3 in senescence and tumor suppression. Results Using immunofluorescence and ChIP-seq we determine the distribution of H4K20me3 in proliferating and senescent human cells. Altered H4K20me3 in senescence is coupled to H4K16ac and DNA methylation changes in senescence. In senescent cells, H4K20me3 is especially enriched at DNA sequences contained within specialized domains of senescence-associated heterochromatin foci (SAHF), as well as specific families of non-genic and genic repeats. Altered H4K20me3 does not correlate strongly with changes in gene expression between proliferating and senescent cells; however, in senescent cells, but not proliferating cells, H4K20me3 enrichment at gene bodies correlates inversely with gene expression, reflecting de novo accumulation of H4K20me3 at repressed genes in senescent cells, including at genes also repressed in proliferating cells. Although elevated SUV420H2 upregulates H4K20me3, this does not accelerate senescence of primary human cells. However, elevated SUV420H2/H4K20me3 reinforces oncogene-induced senescence-associated proliferation arrest and slows tumorigenesis in vivo. Conclusions These results corroborate a role for chromatin in underpinning the senescence phenotype but do not support a major role for H4K20me3 in initiation of senescence. Rather, we speculate that H4K20me3 plays a role in heterochromatinization and stabilization of the epigenome and genome of pre-malignant, oncogene-expressing senescent cells, thereby suppressing epigenetic and genetic instability and contributing to long-term senescence-mediated tumor suppression. Electronic supplementary material The online version of this article (doi:10.1186/s13059-016-1017-x) contains supplementary material, which is available to authorized users.

Descriptive statistics for H4K20me3, histone H4, and input DNA ChIP sequencing reads from proliferating (PRO) and RS, control (CON) and OIS, and CON and H2 IMR90 cells. Table S2. Descriptive statistics for RNA sequencing reads from PRO, RS, CON and OIS IMR90 cells. (DOC 75 kb)

Six additional supplementary figures and legends. (PDF 10.1 MB)

Telomeres cap the end of eukaryotic chromosomes and prevent the natural end of a chromosome\ud from being recognised as a double-stranded DNA break. Dysfunctional telomeres may trigger\ud replicative senescence, or fuse with other telomeres or with non-telomeric DNA breaks. The length\ud of a telomere plays a key role in telomere function. Relatively little is known about how telomeric\ud chromatin influences telomere length and function. A number of studies in mammalian cells have\ud identified a handful of chromatin remodelling proteins and the chromatin marks they deposit in\ud telomere length regulation. However such examples at human telomeres are scarce.\ud The primary aim of this thesis was to investigate whether the chromatin structure of a telomere is a\ud determinant of its length in human cells. Two approaches were taken to address this issue: Firstly,\ud the chromatin structure of telomeres of differing lengths were directly analysed by measuring\ud enrichment of histone modifications known to be prominent at telomeres in other model organisms.\ud Secondly, selected chromatin remodelling proteins were studied to determine whether they play a\ud role in telomeric chromatin structure and telomere length.\ud Single Telomere Length Analysis (STELA) provides a high resolution method to measure telomere\ud length distributions at individual chromosome ends. STELA assays were previously designed for the\ud 2p, 9p, 11q, 12q, 16p, 17p and 18q telomeres. An allele‐specific STELA assay has also been designed\ud for the XpYp chromosome end. In this study novel telomere and telomeric allele‐specific qPCR assays\ud were developed for the same chromosome ends. These qPCR assays, when used in conjunction with\ud ChIP provide a tool for analysing telomeric chromatin structure at individual chromosome ends.\ud Applying this ChIP‐qPCR approach alongside STELA allows any correlations between telomeric\ud chromatin structure and telomere length to be identified. This approach suggested differences in\ud telomeric chromatin structure between telomeres of different lengths in telomerase‐positive\ud HT1080 fibrosarcoma cells. Shorter HT1080 telomeres were less abundant in H3 and TRF1 and also\ud had lower levels of H4K20me3 and, to a lesser extent, H3K4me3 compared to longer telomeres.\ud Differences in chromatin structure were not observed between telomeres of different lengths in\ud telomerase negative MRC5 fibroblasts. Changes in chromatin structure were observed at individual\ud telomeres/telomere alleles were observed between actively proliferating cells and in cells\ud undergoing senescence. Telomeric enrichment of H3 and TRF1 as well as the histone methylation\ud marks H3K4me3, H3K9me3 and H4K20me3 were reduced in senescent cells. The degree of\ud chromatin structural change as the cells entered senescence differed between chromosome ends.\ud This highlights the benefits of using the telomere-specific ChIP-qPCR approach over the more\ud traditional ChIP-dot blot assays which would not be able to differentiate between the chromatin\ud structure of different chromosome ends.\ud To identify roles for chromatin remodelling proteins in telomere length maintenance siRNA\ud mediated knockdown of selected chromatin remodelers was performed in a clonal population of\ud HT1080 cells followed by STELA analysis. RNAi-depletion of the histone methyltransferase (HMTase)\ud III\ud EHMT2 resulted in an increase in very short 17p telomeres whereas loss of another HMTase, DOT1L\ud caused a divergence in the 17p telomere length distribution suggesting the presence of two\ud subpopulations of cells each with differing telomere length distributions. Subtle changes in mean\ud telomere length was observed after siRNA mediated knockdown of the HMTases MLL and EZH2, the\ud histone deacetylases (HDACs) HDAC1 and SIRT6, the ATP dependent chromatin remodelling complex\ud subunit BAF155 and the H3.3 histone chaperone DAXX. However due to certain limitations of the\ud RNAi screen the validity of these observations is questionable and more work would have to be\ud performed to confirm whether these chromatin remodelers have an effect on telomere length.\ud Finally, dramatic telomere shortening was observed in a keratinocyte holoclone population after\ud siRNA mediated knockdown of DAXX at number of chromosome ends. Prolonged depletion of DAXX\ud also caused an increase in telomere‐to‐telomere fusions. A similarly dramatic loss in telomere length\ud was seen in these cells after knockdown of EHMT2.

Life expectancy for patients living with Cystic Fibrosis (CF) is increasing year on year and there is growing interest in the ageing process in CF. Telomeres are repetitive sequences of DNA that cap the ends of eukaryotic chromosomes and shorten with ongoing cell division, thus providing a marker of replicative history and biological ageing. We aimed to investigate whether telomere length as a function of age differs between patients with CF and healthy individuals and whether telomere length is associated with severity of the patient’s CF condition. Peripheral blood samples and demographic data were collected from 47 consenting patients (age 1 to 57 years) with CF attending their routine annual review appointment at the All Wales Adult CF Centre and Noah’s Ark Children’s’ Hospital in Cardiff, UK. Telomere length profiles were assessed from peripheral blood samples, using the high resolution single telomere length analysis technique (STELA) and compared to healthy control telomere length data. Patients with CF had significantly shorter telomere lengths than healthy individuals, when adjusting for age (p