Remodeling of Interstrand Crosslink Proximal Replisomes Is Dependent on ATR, FANCM, and FANCD2

Abstract

Eukaryotic replisomes are driven by the mini chromosome maintenance (MCM [M]) helicase complex, an offset ring locked around the template for leading strand synthesis by CDC45 (C) and GINS (G) proteins. Although the CDC45 MCM GINS (CMG) structure implies that interstrand crosslinks (ICLs) are absolute blocks to replisomes, recent studies indicate that cells can restart DNA synthesis on the side of the ICL distal to the initial encounter. Here, we report that restart requires ATR and is promoted by FANCD2 and phosphorylated FANCM. Following introduction of genomic ICLs and dependent on ATR and FANCD2 but not on the Fanconi anemia core proteins or FAAP24, FANCM binds the replisome complex, with concomitant release of the GINS proteins. In situ analysis of replisomes proximal to ICLs confirms the ATR-dependent release of GINS proteins while CDC45 is retained on the remodeled replisome. The results demonstrate the plasticity of CMG composition in response to replication stress.

Keywords: ATR; CMG; FANCD2; FANCM; GINS; ICL; interstrand crosslink; replication traverse.

Figure 1.. Phosphorylation of FANCM by ATR Is Required for Replication Traverse

(A) Fiber patterns showing replication tracts in the vicinity of Dig-TMP ICLs. (B) Interpretation of the patterns. The pie chart shows the relative frequency of the major patterns in wild-type (WT) cells. RT, replication traverse; SF, single fork; DF, double fork. (C–F) Replication patterns in cells (C) in which ATR expression was suppressed by siRNA; (D) treated with siRNA against ATRIP; (E) treated with siRNA against ATR and exposed to Dig-Ang; and (F) in WT or FANCM-deficient cells, or FANCM-deficient cells complemented with WT FANCM or a phosphorylation resistant variant (S1045). Data are presented as mean ± SD (***p < 0.001, chi-square test). See also Figure S1 and Table S1.

Figure 2.. FANCM Interacts with the MCM Complex

(A) Cells expressing FLAG FANCM were treated with UVA or TMP/UVA and incubated for 1 h with or without ATR inhibitor (ATRi) VE-821. Chromatin proteins were prepared and either analyzed directly (chromatin) or incubated with anti-FLAG antibody linked to magnetic beads. Bound proteins were analyzed by western blotting. (B) Binding of MCM proteins by FANCM requires phosphorylation. Samples were treated with calf intestine alkaline phosphatase. (C) PLA between FLAG-FANCM and MCM2. Cells expressing FLAG-FANCM were either exposed to UVA only or treated with TMP/UVA or TMP/UVA and VE-821. After 1 h the cells were fixed, and the PLA between FANCM and MCM2 determined. (D) Quantification of (C). ***p < 0.001 (rank-sum test). (E) ATR-dependent binding of the MCM complex by the FANCM translocase mutant K117R in cells exposed to TMP/UVA. (F) Traverse frequency is reduced in FANCM knockout cells expressing FANCM K117R. Data are presented as mean ± SD (not significant [NS], p > 0.05; ***p < 0.001; chi-square test). See also Figure S2 and Tables S1 and S2.

Figure 3.. FANCD2 Is Epistatic with FANCM in Support of Replication Traverse

(A) Replication patterns in FANCD2-deficient PD20 cells or PD20 cells complemented with WT FANCD2 or the ubiquitin-resistant K561R variant. (B) Association of FANCM with MCM proteins is dependent on FANCD2. PD20 cells or PD20 cells complemented by expression of WT FANCD2 were transfected with a plasmid expressing WT FLAG-FANCM, followed by treatment as indicated and coIP from a preparation of chromatin proteins. (C) PLA between FLAG-FANCM and MCM2 in PD20 cells, or PD20 cells expressing WT FANCD2. (D) Quantification of (C). ***p < 0.001 (rank-sum test). (E) coIP between FANCM and FANCD2 from chromatin proteins from cells treated with TMP/UVA. (F) FANCM and FANCD2 are epistatic in replication traverse. (A and F) Data are presented as mean ± SD (NS, p > 0.05; ***p < 0.001; chi-square test). See also Figure S3 and Tables S1 and S2.

Figure 4.. FANCM Association with MCM Proteins Is Coincident with Release of GINS Proteins

(A) coIP against MCM2 or phospho-MCM2S108 after treatment of cells with TMP/UVA. (B) coIP against FLAG-FANCM in cells with or without treatment with TMP/UVA. (C) Inverse correlation between GINS and FANCM association with CMG complex. coIP against pMCM2S108 chromatin proteins from FANCM-expressing or -deficient cells, treated with TMP/UVA. (D) PLA between WT or translocase inactive (K117R) FANCM pMCM2S108. (E) PLA between WT or translocase inactive (K117R) FANCM and PSF1 (GINS1). (F) Quantification of (D) and (E). Data are presented as mean ± SD. NS, p > 0.05 (rank-sum test). (G) FANCD2 is required for the interaction between FANCM and the MCM complex defined by phospho-MCM2. coIP between pMCM2 and chromatin proteins in FANCD2 knockout cells or knockout cells expressing WT FANCD2. See also Figure S4 and Table S2.

Figure 5.. The Interaction of CMG Proteins with ICLs Is Regulated by ATR

(A) PLA between Dig tag on the ICLs and FANCM. Cells were treated with UVA, TMP/UVA, or Dig-TMP/UVA and, after 1-h incubation at 37°C, fixed and PLA between Dig and FANCM performed. The UVA and TMP/UVA provide controls for non-specific binding in cells exposed to UVA or TMP/UVA stress. (B) Quantification of (A). (C) PLA between Dig-TMP and MCM2. In addition to Dig-TMP/UVA, cells were treated with either DMSO or VE-821 in DMSO. (D) Quantification of (C) and (E). (E) PLA between Dig-TMP and pMCM2S108. (F) Time course of PLA between Dig-TMP and pMCM2S108. Data are presented as mean ± SD. Control cells were exposed to UVA in the absence of Dig-TMP and incubated for 1 h prior to the PLA. (G) Quantification of (F). (B, D, and G) Data are presented as mean ± SD. NS, p > 0.05; ***p < 0.001 (rank-sum test). See also Figure S5 and Table S2.

Figure 6.. The Loss of GINS from Replisomes at ICLs Is Dependent on ATR and FANCM

Cells were treated with Dig-TMP/UVA, incubated for an hour, and then PLA performed between the indicated protein and Dig. (A) PLA between Dig and CDC45 in the presence or absence of VE-821 or DMSO (V). (B) Quantification of (A). (C) PLA between Dig and PSF1 in FANCM ⁺ or ⁻ cells. (D) Quantification of (E), (G), and (H). (E) PLA between Dig and PSF1, +/− VE-821. (F) Quantification of (E), (G), and (H). (G) PLA between Dig and PSF2, +/− VE-821. (H) PLA between Dig and PSF3, +/− VE-821. (B, D, and F) Data are presented as mean ± SD. ***p < 0.001 (rank-sum test). See also Figure S6 and Table S2.