The FANCM-BLM-TOP3A-RMI1/2 complex suppresses telomere replication stress and Alternative Lengthening of Telomeres

In humans, telomeres consist of ′TTAGGG′ repeats, which form a nucleoprotein complex that caps the ends of chromosomes. Telomeres are constitutively bound by a six-member protein complex called shelterin. The telomere nucleoprotein structure is essential for preventing the recognition of chromosome ends as DNA double-strand breaks, and for suppressing aberrant repair processes, including end-joining and recombination. When telomeres become critically short, they can initiate a DNA damage response, thereby triggering replicative senescence. Cancer cells are able to overcome this proliferative barrier and immortalise through the reactivation of a telomere maintenance mechanism. The majority (85-90%) of cancers utilise a ribonucleoprotein reverse-transcriptase complex called telomerase to extend telomere repeats, while the remaining 10-15%, especially several paediatric cancers, utilise a recombination-based mechanism termed Alternative Lengthening of Telomeres (ALT). ALT cancers tend to have poor prognosis and ALT activity is enriched in several paediatric cancers. ALT activity can also be activated as an adaptive response to treatment therapies that inhibit telomerase. Consequently, there is a need to develop therapies that inhibit or exploit the mechanistic weaknesses of ALT. We sought to study the role of FANCM, a DNA fork translocase that remodels and protects stalled replication forks, at ALT telomeres. FANCM canonically functions as the key sensor and fork remodeller in the Fanconi Anaemia (FA) pathway of inter-strand crosslink (ICL) repair. Its canonical function is to recruit the FA core complex and facilitate ID2 complex (FANCI/FANCD2) mono-ubiquitination at ICL-stalled forks. Yet FANCM has broader roles in preventing replication defects induced by non-ICL sources. We found that FANCM depletion dramatically induces ALT phenotypes, including extrachromosomal telomeric DNA species such as C-circles, large APBs, and break-induced replication events resulting from damaged telomeres. This induction was dependent on the key break-induced replication proteins POLD3 and BLM, the telomeric G-strand-binding shelterin component POT1, and the replication fork translocase SMARCAL1. We then determined which functional domains of FANCM were required to suppress ALT activity, and whether known domains required for ICL repair were necessary. We found that the K117R mutant and, to a lesser extent, the MID and ERCC4 domain mutants were defective at suppressing several ALT phenotypes and telomere dysfunction. Overall, the MM2 domain mutant produced the most striking results, indicating that the MM2 domain was the most important domain required for the suppression of ALT activity. This suggests that the role of FANCM in suppressing ALT requires its interaction with the BLM/TOP3A/RMI1/2 (BTR) complex. We then demonstrated the therapeutic potential of inhibiting the MM2 domain interaction with the BTR complex using a small molecule inhibitor and an inducible fusion peptide construct. While the drug was moderately selective towards ALT cells, the fusion peptide displayed remarkable inhibition specifically of ALT-cells. Overall, we demonstrate that ALT telomeres are particularly dependent on FANCM to prevent excessive ALT activity, which is ultimately detrimental for viability. Therefore, we present a means to selectively kill ALT cells via disruption of the MM2 domain of FANCM