Remodeling of DNA replication structures by the branch point translocase FANCM

Abstract

Fanconi anemia (FA) is a genetically heterogeneous chromosome instability syndrome associated with congenital abnormalities, bone marrow failure, and cancer predisposition. Eight FA proteins form a nuclear core complex, which promotes tolerance of DNA lesions in S phase, but the underlying mechanisms are still elusive. We reported recently that the FA core complex protein FANCM can translocate Holliday junctions. Here we show that FANCM promotes reversal of model replication forks via concerted displacement and annealing of the nascent and parental DNA strands. Fork reversal by FANCM also occurs when the lagging strand template is partially single-stranded and bound by RPA. The combined fork reversal and branch migration activities of FANCM lead to extensive regression of model replication forks. These observations provide evidence that FANCM can remodel replication fork structures and suggest a mechanism by which FANCM could promote DNA damage tolerance in S phase.

Fig. 1.

D-loop dissociation catalyzed by FANCM. (A) Experimental scheme. Asterisks denote ³²P label. Dissociation of D-loops results in the release of the labeled oligonucleotide. (B) Analysis of reaction products by agarose gel electrophoresis. FANCM (lanes 1–6) and K117R FANCM (lanes 7–12) were incubated with the D-loop substrate for the indicated time periods. (C) Experimental scheme. RPA binds to the single-stranded part of the D-loop. (D) Autoradiographs from agarose gels showing native RPA/D-loop complexes (lanes 1–3) and D-loop dissociation in the presence of RPA and ATP (lanes 4–6) or AMP-PNP (lanes 7–9) after deproteinization. (E) As in D, but D-loops were incubated with SSB instead of RPA.

Fig. 2.

FANCM can promote fork regression. (A) Experimental scheme. Fork reversal results in the transfer of the ³²P label (asterisk) from the gapped molecule to the protruding arm of the four-way junction. Extent of fork regression can be estimated by restriction analysis with AvrII (A), BamHI (B), EcoRI (E), AflIII (F), and ScaI (S). (B) Autoradiographs from polyacrylamide gels showing fork regression in the absence of proteins (lanes 1–5), in the presence of FANCM (lanes 6–10), and in the presence of K117R FANCM (lanes 11–15).

Fig. 3.

Extensive fork regression catalyzed by FANCM. (A) Experimental scheme. Complete fork regression results in the formation of linear duplex and nicked circular DNA molecules. Asterisks indicate ³²P label. Dashed lines show the two possible orientations of RuvC-mediated Holliday-junction cleavage, which result in the formation of nicked circular and labeled linear duplex DNA of 2.9 kb (cleavage in the α orientation) and labeled linear duplex DNA of 5.7 kb (cleavage in the β orientation), respectively. (B) Analysis of fork-regression products by agarose gel electrophoresis. FANCM (lanes 1–6) and K117R FANCM (lanes 7–12) were incubated with the replication-fork substrate for the indicated periods of time. The different labeled species are (top down) the four-way junction intermediate arising during the reaction (b), the original replication fork (a), the labeled gap molecule (asterisk; background in all lanes, because of incomplete annealing during fork generation), and the linear duplex molecule (c), the end product of fork regression. (C) Analysis of RuvC-mediated resolution of the Holliday junction-like intermediate formed by FANCM (b; lane 3) into linear duplex DNA of 2.9 kb (c), and linear duplex DNA of 5.7 kb (d), respectively (lane 4).

Fig. 4.

Regression of an RPA-covered replication fork. (A) Experimental scheme. RPA binds to the 114-nt gap on the lagging strand template of the replication fork. Fork regression results in the transfer of the ³²P label (asterisk) from the replication fork to the 0.5-kb linear duplex product. (B) Binding of RPA to the replication fork was verified by agarose gel-shift analysis. (C) Analysis of fork regression by agarose gel electrophoresis. FANCM and the replication-fork substrate were incubated for the indicated time periods without (lane 4) and with (lane 5) preincubation of the replication fork with RPA. As a control, the substrate was incubated without FANCM in the absence (lanes 1 and 2) or presence (lane 3) of RPA. Asterisks indicate background of labeled gapped DNA (*) and labeled linear DNA (**).

Fig. 5.

Model for the role of FANCM in the response to replication stress. (A) In this model, FANCM antagonizes the movement of replication forks advancing toward sites of damage, which might result in fork reversal once the replication fork is stalled. Remodeling of forks by FANCM would stabilize the stalled replication fork and provide time and space for the lesion site to be repaired. (B) Replication forks running into damaged sites may lead to replication-fork collapse and the formation of one-ended DSBs. These are susceptible to promiscuous repair events and are potentially the cause of chromosomal aberrations.