Mitotic DNA synthesis (MiDAS) is a recently discovered phenomenon that is activated in early M phase for
the resolution of late replication intermediates (LRIs) as a final mechanism to support proper chromosome
segregation during mitosis. MiDAS is strongly induced after perturbed S phase with replication stress where
replication forks fail more frequently than normal conditions. Sites of MiDAS appears as punctuated sites of
DNA synthesis in prophase nuclei, which are found almost always at chromosome loci marked with FANCD2
focus formation. Such loci include common fragile sites where the completion of DNA replication is particularly
difficult under replication stress. Activation of MiDAS essentially supports the stability of such loci, but it often
causes gaps and breaks on metaphase chromosomes probably due to its “last minute” nature. Nevertheless, it
is considered that such gaps/breaks still help cells avoid chromosome mis-segregation, which may cause cell
death in the worst case scenario.
A previous model describes RAD52 as a key player, because it functions at an early step of MiDAS and
recruits POLD3, a non-catalytic subunit of Polymerase delta, which is essential for DNA synthesis in prophase.
However, our recent study revealed that this role of RAD52 is limited to human cancer cells, because its absence
has no effect on MiDAS in normal human cells. Rather, we demonstrated FANCD2 as a fundamental regulator
of MiDAS, as its deficiency non-selectively impairs this process in human cells. FANCD2 is a central protein
mutated in a human genetic disorder Fanconi anemia (FA), and our findings along with preliminary data indicate
that MiDAS is a novel cellular function that is primarily driven by a subset of FA genes.
Our revised model for MiDAS in mammalian cells is as follows. The original form of MiDAS in normal human
cells is essentially driven by the FA-driven mechanism that requires mono-ubiquitinated forms of FANCD2
(FANCD2Ub) and PCNA (PCNAUb). Probably reflecting the oncogenic replication stress that they harbor, human
cancer cells use RAD52-driven mechanisms in addition to the FA-driven mechanism. Different from primary
human cells, primary mouse cells operate MiDAS using the two mechanisms similar to human cancer cells. To
test these working hypotheses, we propose the following specific aims: 1) Determine the roles of the FA proteins
in MiDAS in normal human cells, 2) Determine the roles of PCNAUb in MiDAS in human cells, 3) Determine the
role of RAD52-driven MiDAS in human cells, and 4) Test if MiDAS in mice depends on both the FA and RAD52-
driven mechanisms. Successful completion of this proposal will unveil that MiDAS is run via different
mechanisms among mammalian cells. This information will allow us to properly assess the effect of MiDAS
deficiency in respective cell types for future studies.