FANCJ suppresses microsatellite instability and lymphomagenesis independent of the Fanconi anemia pathway

Maintenance of genome integrity during DNA replication is of vital importance to ensure that daughter cells inherit an intact copy of the genetic code. Repetitive DNA sequences are a particular challenge to genome stability due to their propensity to form secondary structures within template or nascent DNA strands that hinder replisome progression and promote template slippage. Microsatellites are repetitive sequences of 1–10 base pairs (bp) of DNA (Boyer et al. 2013). The expansion or contraction of microsatellites during DNA replication has been implicated in a wide range of genetic disorders, including neuromuscular or neurological diseases and cancer. A link between human disease and repetitive sequence instability is most clearly illustrated for trinucleotide repeats (Mirkin 2007), whose expansion is a recurrent cause of Friedreich's ataxia (GAA/TTC repeats) and Huntington's disease (CAG/CTG repeats). Microsatellite instability (MSI) is also a hallmark of Lynch syndrome-associated cancers (Aaltonen et al. 1993; Boland and Goel 2010), which are caused by mutations in DNA mismatch repair (MMR) genes. Although MMR deficiency is associated with hereditary and sporadic colorectal cancers in humans, MMR-deficient mice are primarily predisposed to lymphoma (Li 2008). Current evidence suggests that MMR corrects slipped strand mispairing resulting from additions or deletions in the newly synthesized strand, which arise during secondary structure-triggered template slippage or when the replication of the repeats is impaired. Given the strong association with human disease, understanding the mechanisms that maintain the integrity of repetitive sequences is of great clinical importance. While mechanisms exist to directly detect structural alterations in DNA, including helix-distorting lesions and base–base mismatches, other lesions may go undetected until encountered by the DNA replication machinery. When this occurs, repair must be orchestrated in the context of the replication fork, necessitating coordination of checkpoint, repair, and replication factors. In response to replication fork blockages such as interstrand cross-links (ICLs), the ATR-dependent replication stress checkpoint, Fanconi anemia (FA), and homologous recombination (HR) pathways are essential for replication fork repair and restart (Clauson et al. 2013). Repair of ICLs requires nucleolytic processing, translesion DNA synthesis, and HR (Deans and West 2011). The requirement for HR stems from the generation of DNA double-strand breaks (DSBs) that arise from nucleolytic processing events. The FA pathway is comprised of at least 16 gene products defective in FA patients (Clauson et al. 2013), which present with progressive bone marrow failure, developmental abnormalities, subfertility, and tumor predisposition (Alter and Kupfer 2002; Alter 2003; Crossan and Patel 2012). At the molecular level, the primary function of the FA pathway appears to be to induce monoubiquitylation of the heterodimeric FANCD2/FANCI complex, which coordinates ICL incision and recruitment of downstream repair factors (for review, see Kim and D'Andrea 2012). One of the most enigmatic FA proteins is FANCJ, a DEAH superfamily 2 helicase and part of the subfamily of Fe-S cluster-containing helicases, which also includes XPD, RTEL1, and CHL1 (Rudolf et al. 2006; Gari et al. 2012; Brosh and Cantor 2014). Biallelic mutations in FANCJ give rise to FA complementation group J (Levitus et al. 2005; Levran et al. 2005; Litman et al. 2005), whereas monoallelic mutations predispose to ovarian and breast cancers (Seal et al. 2006; Rafnar et al. 2011). More recently, a significant association with pancreatic and colorectal cancer was found (Rafnar et al. 2011). In line with its role in the FA pathway, defects in FANCJ give rise to exquisite ICL sensitivity in a range of different species (Bridge et al. 2005; Levitus et al. 2005; Litman et al. 2005; Youds et al. 2008). However, attempts to place FANCJ within the FA pathway have provided little insight into its precise function in ICL repair. FANCJ is dispensable for the activation of the FA core complex and hence the monoubiquitylation of FANCD2/FANCI and its recruitment to ICL lesions (Levitus et al. 2004; Litman et al. 2005). Furthermore, an interaction between FANCJ and BRCA1 is not required for classical ICL repair (Xie et al. 2010b), and a role for FANCJ in HR downstream from ICL incision remains ambiguous. Currently, the function of FANCJ in ICL repair remains poorly defined. Biochemical studies have shown that FANCJ unwinds a variety of DNA substrates, including 5′ flaps, forked duplexes, D loops, 5′ tailed triplexes, and G4-DNA structures in a 5′–3′ direction in vitro (Gupta et al. 2005; London et al. 2008; Wu et al. 2008; Sommers et al. 2009). Of these DNA structures, a clear picture has emerged linking FANCJ to the metabolism of G4-DNA secondary structures in vivo. Such a role was first suggested from the observation of increased G/C tract deletions in dog-1 (deletion of G tracts; Caenorhabditis elegans FANCJ) mutant worms, which was proposed to reflect a defect in unwinding G4-DNA structures formed within G/C tracts (Cheung et al. 2002; Youds et al. 2008). Subsequent studies in human Fancj-deficient cells found that genomic deletions tend to accumulate in the vicinity of potential G4-DNA-forming sequences (London et al. 2008). Genome-wide transcription profiling of FANCJ knockout chicken DT40 cells has also revealed that dysregulated genes are significantly associated with G4 sequences. It was proposed that FANCJ maintains epigenetic stability near G4-DNA motifs through two independent mechanisms dependent on either the Y family polymerase REV1 or WRN/BLM helicases (Sarkies et al. 2012). Recently, it was reported that FANCJ depletion from Xenopus laevis egg extract leads to persistent replication stalling at G4 sequences (Castillo Bosch et al. 2014). Despite these observations, it is currently unclear whether FANCJ functions exclusively to maintain genome stability associated with G4-DNA-forming sequences or also participates in the metabolism of other DNA secondary structures. In this study, we report that Fancj-null mice exhibit subfertility, germ cell attrition, epithelial tumor predisposition, and exquisite sensitivity to ICL-inducing agents, which phenocopy other mouse models of FA (Bakker et al. 2012). Unexpectedly, Fancj-deficient mice also present with enhanced predisposition to lymphoma, and cells derived from these mice are hypersensitive to replication inhibitors. Furthermore, Fancj−/−Fancd2−/− double-knockout mice display heightened germ cell attrition and are considerably more sensitive to ICL-inducing agents than single knockouts. Since Fancj-deficient cells are insensitive to G4-stabilizing drugs and are devoid of telomere fragility, we considered the possibility that FANCJ performs additional functions in genome stability independent of its role in the FA pathway and distinct from a role in G4-DNA metabolism. Strikingly, we show that Fancj-deficient, but not Fancd2-deficient, mice accumulate spontaneous MSI corresponding to both expansions and contractions of repeat sequences. Similarly, FA-J patient cells and human Fancj knockouts cells also present with MSI, which is exacerbated by replication inhibition. Thus, we propose that FANCJ counteracts the formation of secondary structures that arise during replication of microsatellite sequences, which minimizes the potential for strand slippage during DNA polymerization. Our findings can potentially explain the widespread involvement of FANCJ in human cancers.