The Cohesion complex maintains genome stability by preventing end joining of distant DNA ends in S phase

ABSTRACT

Genome instability is a hallmark of cancer cells. The joining of distant DNA double-strand ends (DSEs) ineluctably leads to genome rearrangements. We found that the cohesion complex maintains genome stability by repressing the joining of distant DSEs specifically in the S phase, i.e., the main phase producing one-ended DSEs.

Genetic instability is a hallmark of cancer cells. Genomes are routinely assaulted by genotoxic stresses from endogenous as well as exogenous origins. DNA double-strand breaks (DSBs) are highly toxic lesions that can lead to genetic instability, leading to carcinogenesis. In particular, the joining of distant DNA double-strand ends (DSEs) inevitably generates genome rearrangements.

There are two main general strategies for the repair of DSBs. The first is referred to as homologous recombination because it takes advantage of an intact homologous sequence, which is invaded by the broken molecule and then copied. The second strategy ligates the 2 broken DSEs without requiring sequence homologies, and is thus referred as non-homologous end joining (NHEJ). Two genetically distinct mechanisms can join DSEs in a homologous-independent manner: conservative canonical NHEJ (cNHEJ), which is KU/ligase4-dependent, and error-prone alternative end joining (A-EJ), which is KU-ligase4-independent.¹

Although it is essential to maintain genome stability, end joining can generate genetic instability at 2 levels: at the nucleotide level at the junction scar or through long-range genome rearrangements such as translocations, inversions, or long-range deletions when joining 2 distant DSEs (Fig. 1A). It is noteworthy that replication stress can generate DSBs with single DSEs², i.e., one-ended DSBs (Fig. 1B). End joining upon replication stress necessarily implies 2 DSEs that were originally distant and therefore inevitably leads to genome rearrangements and chromosome fusions (Fig. 1C). Therefore, the S phase is a particularly sensitive phase for such events. NHEJ is active throughout the cell cycle, and thus can act on replication stress-induced DSBs (see citations in ref. 1¹). Importantly, replication stress has been documented at precancerous or early steps of tumorigenesis and senescence, suggesting that it might be an important source of genetic instability at cancer initiation.^3-5 In this context, one-ended DSBs would play a primary role.

Figure 1.

DNA end joining and genetic instability. (A) Rearrangements (inversion, deletion, translocation) resulting from the end-joining of distant DSEs. (B) One-ended DSB generated by a replication fork collapse. (C) Rearrangements and chromosome fusion resulting from the end-joining of 2 one-ended DSBs. (D) Example of premature sister chromatid separation upon silencing of the cohesion protein RAD21 (left panel). Zoom: one chromosome with separated sister chromatids. Right panel: normal chromosomes exhibiting sister chromatid cohesion.

Using different substrates to monitor end joining of close (34 bp apart) versus distant (3,200 bp apart) DSEs, we found that joining of the distant ends is specifically repressed in the S/G2 phase, whereas joining of close ends is not affected.⁶ This reveals that mechanisms exist to protect against the joining of distant ends specifically during the S/G2 phase, when one-ended DSEs are generated (see above). The joining of 2 distant DSEs (a few kb apart) is not affected in the G1 phase. Because one-ended DSBs are not generated in the G1 phase (in contrast to the S phase), restriction of the joining of distant DSEs in G1 is unnecessary.

The cohesion complex maintains cohesion between the 2 sister chromatids until mitosis, ensuring even chromosome segregation. Upon replication stress, the cohesion complex favors sister chromatid exchange (SCE) to resume the progression of arrested replication forks, thus maintaining genome stability. Since the sister chromatids are identical, this leads to maintenance of genome stability. A defect in the cohesion complex decreases SCE and also results in premature sister chromatid separation (Fig. 1D), leading to mitotic catastrophe and chromosome segregation defects. Because the sister chromatids result from replication, the cohesion complex was a good candidate for repression of the end joining of distant DSEs in S phase. We found that ablation of the cohesion complex abolished the repression of end joining of distant ends, but not of close ends.⁶ We conclude that the cohesion complex represses the end joining of distant ends (even those of only a few kb apart) in S phase. On the genome scale, we found that the cohesion complex protects against long-range rearrangements (translocations, duplications, inversions, deletions). Finally, replication stress and ablation of the cohesion complex synergistically lead to the generation of chromosome fusions.⁶

These data reveal an important novel function for the cohesion complex in the maintenance of genome stability in response to replication stress. In addition to favoring SCE, the cohesion complex represses the joining of distant DNA ends that are generated in the S phase, thus protecting against genome rearrangements. Importantly, the cohesion complex does not repress the repair of close ends, allowing the repair of DSBs such as those generated by ionizing radiation or reactive oxygen species. Therefore, repression of end joining by the cohesion complex focuses on DSEs generated by replication stress.

Because of the importance of replication stress and genetic instability during the early stage of cancer, the cohesion complex should play an essential role in protection against tumor initiation. Consistent with this notion, mutations in the cohesion complex have been frequently reported in tumors.^7-9 However, the Cornelia de Lange syndrome (CdL), which results from germ-line mutations in genes encoding members of the cohesion complex leading to strong development abnormalities, is not associated with cancer predisposition. Importantly, CdL cells do not present premature sister chromatid separation.¹⁰ Given that mutations in CdL syndrome correspond to haploinsufficiency and/or hypomorphism, it is proposed that the residual activity is sufficient to ensure sister chromatid cohesion. According to our data, this should also be sufficient to support genome stability, thus protecting against cancer predisposition. Since the cohesion complex is also involved in transcription, including during the G1 phase, it is also proposed that the mutations found in this syndrome are separation of function mutations affecting transcription but not genome stability.

The cohesion complex plays multiple essential roles. In particular, the cohesion complex controls the maintenance of stability during replication through different parallel processes: it favors SCEs and restricts the end joining of distant DSEs, which are mainly generated by replication stress. Therefore, the cohesion complex should play a pivotal role in protection against tumor initiation.

Funding Statement

This work was supported by funding from the Ligue Nationale Française Contre le Cancer, ANR (Agence Nationale de la Recherche, ANR-14-CE10-0010-02), AFM-Téléthon and INCa (Institut National du Cancer, 2011-1-RT-01, 2011-1-PLBIO-09, 2013-1-PLBIO-14).

Disclosure of potential conflicts of interest

No potential conflicts of interest were disclosed.