Cellular senescence and tumour suppressors
Cellular senescence is a state of stable cell cycle arrest with active metabolism. Similar to apoptosis, senescence can be a failsafe program against a variety of cellular insults. In contrast to apoptosis, in which cytotoxic signals converge to a common mechanism, senescence is typically a delayed stress response involving multiple effector mechanisms.
These effector mechanisms include epigenetic regulation, the DNA damage response, the senescence-associated secretion phenotype (SASP) and autophagy. The relative contribution of these effectors varies depending on the trigger and cell type, and it is possible that the combination and balance of these effectors determines the quality of the senescence phenotype. Thus, to understand the senescence program, it is important to identify new effector mechanisms and examine how they associate with each other, and also to identify which effector mechanisms could be potential targets for cancer therapy.
Genome-wide analysis of heterochromatin components in SAHF
Certain types of cells undergo distinct alterations in chromatin structure during senescence, called senescence-associated heterochromatic foci (SAHF). SAHF have been widely used as a marker of senescence, and more importantly, several new components of senescence machinery have been successfully identified using SAHF as a readout. Thus, it is important to understand SAHF structure in more detail and how SAHF are actually formed. During oncogenic Ras-induced senescence, SAHF formation is dynamically regulated in HDFs, thus providing a unique tool to study not only senescence, but also chromatin biology. To characterise SAHF in detail, we have investigated a dynamic redistribution of the specific histone modifications using confocal microscopy as well as ChIP-seq. In collaboration with the Tavaré group (CRUK CI), we have shown that SAHF formation results in a concentric chromatin architecture, not only segregating the chromatin of individual chromosomes into heterochromatin and euchromatin, but also concentrating histones H3K9me3 and H3K27me3 (markers of constitutive and facultative heterochromatin, respectively) in non-overlapping layers within SAHFs. Surprisingly, despite the dramatic chromatin structure alteration, the linear ‘global’ epigenomic landscapes of these repressive marks are highly static, although local profiles of those histone marks can be dynamic particularly at some genic regions. Our data indicate that the high-order chromatin structure change during SAHF formation is achieved mainly through the spatial rearrangement of pre-existing heterochromatin, rather than spreading of heterochromatin (Chandra et al., Mol Cell 2012; 47: 203).
To further identify factors that facilitate SAHF formation, we have focused on dynamic alteration of the nuclear lamina structure, which is associated with H3K9me3 positive heterochromatin in HDFs. Emerging evidence suggests that Lamin B1 (LMNB1), a key component of the nuclear lamina, is specifically down-regulated during senescence, and we found that LMNB1 reduction is highly correlated with the loss of peri-nuclear H3K9me3 as well as SAHF formation. The van Steensel group (Netherlands Cancer Institute) has shown that the nuclear lamina contacts with hundreds of large genomic regions, lamina associated domains (LADs), defined through genome-wide mapping of LMNB1. LADs are associated with facultative heterochromatin markers, H3K9me2 and H3K27me3. To determine dynamic alteration of LADs during senescence, we have mapped LMNB1 binding in both growing and senescent cells. In addition to H3K9me2 and H3K27me3, H3K9me3 is also associated with LADs, where H3K9me3 primarily occupies the central regions of LADs. Despite the global down-regulation in LMNB1, LMNB1 binding is reduced mainly in H3K9me3-enriched regions, and this reduction is correlated with the spatial repositioning of H3K9me3-enriched chromatin and SAHF formation but not with gene expression changes during senescence. Furthermore, we also found de novo gains in LMNB1 binding in small sections of the genome, which includes a number of cell cycle genes, and these de novo LMNB1 binding regions in senescent cells are correlated with increased H3K27me3 and gene repression. These results suggest that LMNB1 may contribute to senescence in at least two ways due to its uneven genome-wide redistribution: first, through the spatial reorganization of chromatin and, second, through gene repression (Sadaie et al., Genes Dev. 2013; 27: 1800) (Figure 1).
Figure 1. Uneven alterations in LMNB1 genomic profile during Ras-induced senescence (RIS) in HDFs. A. Correlation between LMNB1 down-regulation and SAHF formation. Confocal images for indicating antibodies are shown. Perinuclear H3K9me3 foci are also decreased in RIS cells. B. Central regions of LADs are enriched for H3K9me3. LMNB1 is preferentially depleted from the H3K9me3-positive regions (not shown). C. Pie chart describes genomic regions based on differential binding events of LMNB1 between growing and RIS cells. Numbers represent percentage of the genome in each class.
TOR-autophagy spatial coupling compartment, TASCC
We have shown that autophagy activity is increased during senescence and that mTOR and autophagy cooperatively facilitate SASP through forming a cellular compartment, the TOR-autophagy spatial coupling compartment (TASCC), which provides a local environment enriched for amino acids and mRNA translation machinery (Young et al., Genes Dev 2009; 23: 798, Narita et al., Science 2011; 332: 966). We have also shown that a TASCC-like structure is not limited to Ras-induced senescence, and we are currently investigating an implication of this structure more generally in the cancer context.
Identification of senescence-associated p53 function (in collaboration with the Tavaré group, CRUK CI)
A tumour suppressive transcription factor, p53, plays a critical role in many stress responsive phenotypes, including DNA damage checkpoints, apoptosis, and senescence (Figure 2). Although ample data have supported a role for p53 in senescence, the precise mechanism or ‘senescence-specific p53-targets’ are not known. To address this issue, we are currently using HDFs, where we can induce different phenotypes, in which p53 plays a crucial role, depending on environmental stimuli or other conditions. These phenotypes include, senescence, apoptosis, and acute DNA damage response. Using expression microarrays in conjunction with stable RNAi technology as well as p53 ChIP-seq, this system would allow us to understand both general and phenotype-specific p53 functions.
Figure 2. Model system for understanding a comprehensive picture of p53 functions.