Low replication stress leads to specific replication timing advances associated to chromatin remodelling in cancer cells

Abstract

DNA replication is very well orchestrated in mammalian cells due to a tight regulation of the temporal order of replication origin activation, known as the replication timing program. The replication timing of a given replication domain is very robust and well conserved in each cell type. Upon low replication stress, the slowing of replication forks induces delayed replication of some fragile regions leading to DNA damage and genetic instability. Except for these fragile regions, the direct impact of low replication stress on the replication timing in different cellular backgrounds has not been explored in detail. Here we analysed DNA replication timing across the whole genome in a panel of human cell lines in the presence of low replication stress. We demonstrate that low replication stress induced by aphidicolin has a stronger impact on the replication timing of cancer cells than non-tumour cells. Strikingly, we unveiled an enrichment of specific replication domains undergoing a switch from late to early replication in some cancer cells. We found that advances in replication timing correlate with heterochromatin regions poorly sensitive to DNA damage signalling while being subject to an increase of chromatin accessibility in response to aphidicolin. Finally, our data indicate that, following release from replication stress conditions, replication timing advances can be inherited by the next cellular generation, suggesting a new mechanism by which some cancer cells would adapt to cellular or environmental stress.

Replication Stress, Genomic Instability, and Replication Timing: A Complex Relationship

“…In human cells entering replicative senescence, the RT program is globally preserved (less than 0.5% of the genome is slightly affected) during senescence-associated replication stress (slowing of fork rates and activation of dormant origins [125]), suggesting that RT is largely resistant to replication stress. In contrast to this study, recent works have shown that low replication stress induced by aphidicolin had significant effects on the RT of several cell lines (with 1.5-6.5% of the genome significantly affected), resulting in advances and delays in the replication of specific genomic domains [29,30]. Notably, these global values of RT variations in the genome are not very high, suggesting that the impact of replication stress on RT is not drastic.…”

“…Thus, the participation of replication timing perturbations probably exacerbates the cancer phenotype through the deregulation of other crucial elements that shape the fate of the genome. Interestingly, it was very recently observed in human cancer cells that large advances in RT induced by replication stress in mother cells were also found in the daughter cells released from replication stress [ 29 ]. These specific genomic regions were associated with chromatin remodeling and gene upregulation in the daughter cells.…”

“…The organization and structure of the genome are known to be crucial for the identity and fate of the cell. This is similarly the case for RT, as studies using abortive approaches of different factors that regulate it and treatments with different drugs have never shown more than a limited percentage of RT changes [ 10 , 11 , 12 , 28 , 29 , 30 ]. One possible explanation is that a degree of change beyond this limit is so high that it is catastrophic for the cell, as it must significantly disturb other molecular processes in the nucleus.…”

DNA Replication Stress Generates Distinctive Landscapes of DNA Copy Number Alterations and Chromosome Scale Losses

“…There may be other factor(s) involved in predisposing these specific genomic sites to fragility. For example these genes may lie in regions that are replicated early or mid S-phase in normal cells but could be delayed in timing in the presence of aphidicolin 36,39,47 which could explain their fragility. In our study we analysed replication timing from DMSO-treated cells but it would be interesting to determine whether fragility of these regions correlates with aphidicolin-specific delays in replication timing.…”