Genome Integrity Discussion Group June 2021

The breast cancer susceptibility genes BRCA1 and BRCA2, are potent tumor suppressors. The loss of these genes causes profound genomic instability due to their role in homologous recombination and replication fork management. While BRCA-deficient tumors initially respond well to standard-of-care therapies, such as cisplatin and PARP inhibitors (PARPi), tumors invariably acquire resistance to these agents and do recur, leaving patients with very few therapeutic options. It is therefore imperative that we identify new genetic vulnerabilities caused by BRCA1/2 mutations, as a means to identify new therapeutic targets for BRCA-deficient tumors. In my presentation, I will discuss our recent efforts in mapping genome maintenance pathways using genome-scale CRISPR/Cas9 screens in human cells, including BRCA synthetic lethality screens. I will highlight how our development of a genome stability network has identified potentially actionable synthetic lethal genetic interactions in BRCA-deficient cells. In particular I will discuss our discovery that the CIP2A protein mediates tolerance of DNA lesions during cell division that accumulate in BRCA-deficient cells.

Control of DNA replication is critical for cell cycle progression and genomic integrity. DNA replication requires a stepwise assembly of the origin of replication complex (ORC), Cdc6, Cdt1 and the loading of helicases Mcm2-7 on DNA in G1 phase. Commitment to this assembly depends on Cdc6, an indispensable substrate for cyclin-dependent kinases (CDKs). Cdc6 contains eight Cdk1 consensus sites, including a Cks1 docking site at T7 to prime Cdc6 for multiple phosphorylations and two phospho-degrons at T39-S43 and T368-S372. Thus, each phosphorylation site has a distinct role in regulating Cdc6 function. During S-phase, Clb5/Cdk1 phosphorylates Cdc6-T7 to recruit Cks1, the Cdk1 phospho-adaptor, to trigger multisite phosphorylation and create phospho-degrons to promote SCF-dependent degradation. Thereby, replication is inhibited until the next cell cycle. During mitosis, the Clb2/Cdk1/Cks1 complex tightly binds to the phosphorylated Cdc6 to prevent premature origin licensing and potentially shields phospho-degrons from SCF recognition to maintain Cdc6 protein levels during mitosis.

PP2A-Cdc55 and Cdc14 phosphatases oppose Cdk1 phosphorylation and have been shown to interact with Cdc6 during mitosis. Previous reports also show that Cdk1 inhibitor Sic1 genetically interacts with Cdc6 during late mitosis. However, a detailed mechanism on how Sic1 and these phosphatases control Cdc6 function has not been elucidated. Here, we show that PP2A-Cdc55 dephosphorylates Cdc6-T7 and –T23 sites to release Clb2 binding, exposing the phosho-degrons. Cdc14 dephosphorylates C-terminal phospho-degrons T368-S372, leading to Cdc6 stabilization. We also obtained evidence that Sic1 releases Clb2/Cdk1/Cks1 inhibitory complex from Cdc6 to load Mcm2-7 complexes on chromatin.

Our results suggest that PP2A and Cdc14 sequentially dephosphorylate distinct Cdc6 CDK sites during mitosis. Sic1 also releases the Clb2/Cdk1/Cks1 inhibitory complex from Cdc6 to trigger origin licensing upon mitotic exit. Such a mechanism ensures faithful DNA replication to maintain genome integrity.

Recent sequencing analyses have found that histone lysine-to-methionine mutations are associated with distinct types of cancers. Previous study shows that these oncogenic histone K-to-M mutations can block the methylation of their corresponding lysine residues on wild-type histones. One attractive model is that these mutations sequester histone methyltransferases, but genome-wide studies show that mutant histones and histone methyltransferases often do not colocalize. Using fission yeast histone H3K9M mutation as a model, we show that, even though the H3K9M-containing nucleosomes are broadly distributed across the genome, the histone H3K9 methyltransferase Clr4 is mainly sequestered at pericentric repeats. This selective sequestration of Clr4 depends not only on H3K9M, but also on H3K14 ubiquitylation (H3K14ub), a modification deposited by a Clr4-associated E3 ubiquitin ligase complex. In vitro, H3K14ub synergizes with H3K9M to interact with Clr4 and potentiates the inhibitory effects of H3K9M on Clr4 enzymatic activity. Moreover, binding kinetics show that H3K14ub overcomes Clr4’s aversion of H3K9M and reduces its dissociation. The selective sequestration model reconciles previous discrepancies and demonstrates the importance of protein interaction kinetics in regulating biological processes.

Fanconi Anemia (FA), a model syndrome of genome instability, is caused by a deficiency in DNA interstrand crosslink (ICL) repair resulting in chromosome breakage. The FA repair pathway comprises at least 22 FANC proteins including BRCA1 and BRCA2, which protect against carcinogenic endogenous and exogenous aldehydes. Individuals with FA are hundreds to thousands-fold more likely to develop head and neck (HNSCC), esophageal and anogenital squamous cell carcinomas (SCCs) with a median onset age of 31 years. The aggressive nature of these tumors and patient intolerance of platinum and radiation-based therapy is associated with poor survival in FA. Molecular studies on SCCs from individuals with FA (FA SCCs) have been limited, and it is unclear how they relate to sporadic HNSCCs driven by tobacco and alcohol exposure or human papillomavirus (HPV) infection.

Employing multi-platform sequencing of FA patient SCCs, we demonstrate that the primary genomic signature of FA-deficiency is the presence of a high number of structural variants (SVs). This genomic instability underlies elevated copy number alteration rates of key HNSCC-associated genes, including PIK3CA, MYC, CSMD1, PTPRD, YAP1, MXD4, and EGFR. SVs found in FA tumors are enriched for small deletions and arise in the context of TP53 loss. In contrast to sporadic HNSCC, we find no evidence of HPV infection in FA HNSCC, although positive cases were identified in gynecologic tumors. SVs in FA SCC cluster between 1-100kbp in size, with breakpoints preferentially localizing to early-replicating regions, common fragile sites, tandem repeats, and SINE elements. These SVs frequently occur in chains leading to complex multi-chromosomal rearrangements and recurrent fold-back inversions that drive oncogene amplification. A murine allograft model of FA pathway-deficient SCC was enriched in SVs, exhibited dramatic tumor growth advantage, more rapid epithelial-to-mesenchymal transition (EMT), and enhanced autonomous inflammatory signaling when compared to an FA pathway-proficient model.

In light of the protective role of the FA pathway against SV formation, and recent findings of FA pathway insufficiency in the setting of increased formaldehyde load resulting in hematopoietic stem cell failure and carcinogenesis, we propose that high copy-number instability in sporadic HNSCC may result from functional overload of the FA pathway by endogenous and exogenous DNA crosslinking agents. We propose that the FA SCC represents an excellent model to understand sporadic SCC tumorigenesis through DNA damage.

Telomere repeat containing RNAs (TERRAs) are a family of long non-coding RNAs transcribed from the sub-telomeric regions of eukaryotic chromosomes. TERRA transcripts can form R-loops at chromosome ends; however the importance of these structures or the regulation of TERRA expression and retention in telomeric R-loops remain unclear. Here, we show that the RTEL1 (Regulator of Telomere Length 1) helicase influences the abundance and localization of TERRA in human cells. Depletion of RTEL1 leads to increased levels of TERRA RNA while reducing TERRA-containing R loops at telomeres. In vitro, RTEL1 shows a strong preference for binding G-quadruplex structures which form in TERRA. This binding is mediated by the C-terminal region of RTEL1, and is independent of the RTEL1 helicase domain. RTEL1 binding to TERRA appears to be essential for cell viability, underscoring the importance of this function. Degradation of TERRA containing R-loops by overexpression of RNAse H1 partially recapitulates the increased TERRA levels and telomeric instability associated with RTEL1 deficiency. Collectively, these data suggest that regulation of TERRA is a key function of the RTEL1 helicase, and that loss of that function may contribute to the disease phenotypes of patients with RTEL1 mutations.

The mechanism and control of DNA replication initiation is critical for maintaining genome integrity. In eukaryotes, DNA replication initiates from replication origins that bind the Origin Recognition Complex (ORC). Origin establishment requires well-defined DNA sequence motifs in Saccharomyces cerevisiae and some other budding yeasts, but most eukaryotes lack sequence-specific origins. However, neither the mechanism of DNA replication origin specification nor how this mechanism evolves was clear.

ORC and Cell Division Cycle 6 (Cdc6) proteins bind to replication origins. Chromatin licensing and DNA replication factor 1 (Cdt1) and replication helicase Mcm2-7 are recruited to the ORC-Cdc6 bound origins, forming an ORC-Cdc6-Cdt1-Mcm2-7 (OCCM) structure, which is an intermediate for assembling a pre-Replicative Complex (pre-RC) at all origins prior to the initiation of replication in S phase. A 3.9 Å structure of S. cerevisiae ORC-Cdc6-Cdt1-Mcm2-7 (OCCM) bound to origin DNA revealed that a loop within Orc2 inserts into a DNA minor groove and an alpha-helix within Orc4 inserts into a DNA major groove. This observation was confirmed by a higher resolution structure of ORC-DNA published later. Interestingly, the sequence of the Orc4 insertion alpha-helix and DNA interacting loop within Orc2 has rapidly evolved within budding yeasts and is absent in Candida albicans, Pichia pastoris, S. pombe, and metazoan--species in which origins are either not sequence-specific or exhibit completely different sequence motifs than those found in S. cerevisiae.

Using a massively parallel origin selection assay coupled with a custom mutual-information-based modeling approach, and a separate analysis of whole-genome replication profiling, here we show that the Orc4 alpha-helix contributes to the DNA sequence-specificity of origins in S. cerevisiae and Orc4 alpha-helix mutations change genome-wide origin firing patterns.

Importantly, the DNA sequence specificity of replication origins, mediated by the Orc4 alpha-helix, has co-evolved with the gain of ORC-Sir4-mediated gene silencing and the loss of RNA interference. We will discuss the implications of these observations.