Mms1 binds to G-rich regions in Saccharomyces cerevisiae and influences replication and genome stability

Abstract

The regulation of replication is essential to preserve genome integrity. Mms1 is part of the E3 ubiquitin ligase complex that is linked to replication fork progression. By identifying Mms1 binding sites genome-wide in Saccharomyces cerevisiae we connected Mms1 function to genome integrity and replication fork progression at particular G-rich motifs. This motif can form G-quadruplex (G4) structures in vitro. G4 are stable DNA structures that are known to impede replication fork progression. In the absence of Mms1, genome stability is at risk at these G-rich/G4 regions as demonstrated by gross chromosomal rearrangement assays. Mms1 binds throughout the cell cycle to these G-rich/G4 regions and supports the binding of Pif1 DNA helicase. Based on these data we propose a mechanistic model in which Mms1 binds to specific G-rich/G4 motif located on the lagging strand template for DNA replication and supports Pif1 function, DNA replication and genome integrity.

Figure 1.

Mms1 associates with G-rich regions in vivo. (A) G-rich binding motif of Mms1 identified by MEME-ChIP search. (B) Validation and characterization of ChIPseq data. Using Myc-tagged Mms1 conventional ChIP and qPCR was performed using primers directed against endogenous regions. Regions are indicated below graph. Detail on regions and primers are listed in Supplementary data 3. Three regions identified as Mms1 binding sites by ChIPseq were tested (Chr VII_BM, X_BR, XI_BM) as well as nine additional genomic regions. Sites that contain the specific G-rich binding motif identified by MEME are marked with BM, additional G-rich binding regions identified by ChIP-seq with BR and negative controls with NC. Plotted are IP/input values as mean value ± standard deviation (SD). N≥3 biological replicates. Statistical significance compared to untagged cells was determined by Student's t-test. *P < 0.05. (C) CD analysis of the three Mms1 binding regions. (D) CD analysis of the two negative controls. The CD spectrum of a folded parallel G4 structure has a specific maximum and minimum (264 nm and 243 nm, respectively) which differs from the maximum and minimum of B-DNA (245 and 290 nm, respectively) (62,63). See Supplementary data 7 for details of analyzed regions and sequences. Shown are the ellipticity (in mdeg) values. This analysis demonstrates that G-rich Mms1 binding regions, harboring a G4_tract2, can form G4 DNA structures.

Figure 2.

Mms1 protein levels are highest in G1 phase and Mms1 binds throughout cell cycle. (A) FACS analysis of cells arrested in G1, S or G2 phase. Cells were arrested in G1 by treatment with α-factor, in S phase by HU and in G2 by nocadozole. (B) Western blot analysis of Myc-tagged Mms1 protein levels in G1, S and G2 phase. Level of Mms1 was quantified using Hsp60 as a reference protein. Shown are mean Myc-tagged Mms1 levels normalized to Hsp60 ± SD. N = 3 biological replicates. See Supplementary data 8A for the gel. (C) ChIP and qPCR analysis of Mms1-Myc to seven BR in G1-, S- and G2-phase. Plotted are IP/input values as means ± SD. N≥ 3 biological replicates. In most cases Mms1 binds similar in G1, S and G2 phase. Statistical significance compared to cells arrested in G1 phase was determined by Student's t-test. *P < 0.05.

Figure 3.

Mms1 binds independently of Rtt101 and Mms22, supports Pif1 binding at G4 motifs and by this promotes DNA replication. (A) ChIP and qPCR analysis of Mms1-Myc at seven BR and two NC. Binding of Mms1 was monitored in wild type (light), rtt101 (grey) and mms22 (dark) cells. Statistical significance compared to Myc-tagged Mms1 wild type cells. For details on regions see Supplementary data 3. (B) Replication fork progression was analyzed by detected DNA Pol2 binding levels at Mms1 binding sites. ChIP and qPCR of DNA Pol2 binding in wild type and mms1 cells was performed at five BR and two NC. Statistical significance compared to Myc-tagged DNA Pol2 wild type cells. (C) Binding of Pif1 DNA helicase was analyzed at four Mms1 binding regions (BR) in wild type and in mms1 cells. As control for Pif1 binding we used two known Pif1 binding sites, the replication fork barrier at the rDNA (rDNA) (54) and telomere VI-R (tel) (53) as well as one Pif1 independent site (tRNA). As done previously (53) IP/input values are compared to IP/input values of ARO1 where no Pif1 binds. Here, fold enrichment over ARO1 was plotted as mean value ± SD. For all ChIP; N ≥ 3 biological replicates. Statistical significance was determined by Student's t-test. *P < 0.05, **P < 0.01, ***P < 0.001.

Figure 4.

Mms1 is necessary for genome stability, especially at G4 motifs. (A and B) Mms1 and DNA Pol2 binding to G4 and mutated G4 was monitored by ChIP and qPCR. VI_G4 was mutated using Cre-LoxP (VI_G4mut). Binding of Mms1 and DNA Pol2 was monitored at VI_G4, X_G4 and I_NC in the cre-loxP background as well as unaltered background. (A) As before (Figure 1) Mms1 binds to G4 motifs VI_G4, X_G4 (light). Binding of Mms1 to VI_G is abolished upon mutation VI_G4 (gray) using cre-LoxP. In the same background (cre-LoxP) Mms1 binding to X_G4 is unaffected. (B) DNA Pol2 binding was analyzed in wild type and mms1 cells. In wild type cells, DNA Pol2 binds similarly to VI_G4 and VI_G4mut. As before (Figure 3B) in mms1 cells DNA Pol2 binding is enriched at VI_G4 and X_G4 (light grey), but upon mutation of VI_G4mut binding of DNA Pol2 is reduced (black). DNA Pol2 binding in wild type and mms1 cells was not affected at X_G4 in the cre-LoxP background. Plotted are the IP/input values as mean value ± SD. N ≥ 3 biological replicates. (C) Schematic of the genome region used in the GCR assay. The GCR rate was calculated by fluctuation analysis. (D) The GCR rate was determined in wild type (wt) (white) and mms1 (gray) cells. As inserts a G4 motif (G4-LEU2), a G-rich region (GR-LEU2), a non-G-rich region (NG-LEU2) as well as LEU2 marker were used (see Supplementary data 14 for additional information). Shown are mean values ± SD as fold enrichment over wild type without insert. n = 7 biological replicates, N≥3. Statistical significance compared to no insert strain. Significance was calculated by Student's t-test. *P < 0.05, **P < 0.05, ***P < 0.001.

Figure 5.

Mechanistic model of Mms1 function at G4 structures. Mms1, perhaps in complex with an interaction partner, binds to G4 structures formed in G1 phase or in early S phase. It binds only to G4 structures form on the lagging strand template for DNA replication. During S phase, the replication fork passes the unsolved G4 structures leaving a gap behind. Mms1 enables (directly or indirectly) the binding of Pif1, which unwinds the G4 structure in late S phase. The resulting gap at the resolved G4 structure is then repaired via a so far unknown mechanism.