DNA2 and MSH2 cooperatively repair stabilized G4 and allow efficient telomere replication

Abstract

G-quadruplexes (G4s) are widely existing stable DNA secondary structures in mammalian cells. A long-standing hypothesis is that timely resolution of G4s is needed for efficient and faithful DNA replication. In vitro, G4s may be unwound by helicases or alternatively resolved via DNA2 nuclease mediated G4 cleavage. However, little is known about the biological significance and regulatory mechanism of the DNA2-mediated G4 removal pathway. Here, we report that DNA2 deficiency or its chemical inhibition leads to a significant accumulation of G4s and stalled replication forks at telomeres, which is demonstrated by a high-resolution technology: Single molecular analysis of replicating DNA (SMARD). We further identify that the DNA repair complex MutSα (MSH2-MSH6) binds G4s and stimulates G4 resolution via DNA2-mediated G4 excision. MSH2 deficiency, like DNA2 deficiency or inhibition, causes G4 accumulation and defective telomere replication. Meanwhile, G4-stabilizing environmental compounds block G4 unwinding by helicases but not G4 cleavage by DNA2. Consequently, G4 stabilizers impair telomere replication and cause telomere instabilities, especially in cells deficient in DNA2 or MSH2.

Fig. 1. DNA2 deficiency leads to G4 accumulation in mammalian cells.

a The top panel shows an AIRYSCAN joint deconvolution (jDCV) nuclear region with DNA2 in green and G4 in red. Scale bar = 1 μm. The bottom panel shows the identified spots of this region with DNA2 spots within 100 nm of a G4 spot in yellow; b Relative frequency of co-localization of DNA2 and G4. Red lines indicated mean ± SEM. P-value was calculated using a two-tailed paired t-test (n = 64 cells, p = 0.003); c Reconstituted G4 excision repair assay. Top panel: schematic diagram detailing DNA2 cleavage of G4 and gap formation. Polδ or Polβ then fills the gap using ³²P-dTTP and other dNTPs, producing ³²P-labeled products. Bottom panel: Representative denaturing PAGE image from 3 independent experiments showing G4 excision and repair; d, e G4 immunofluorescence staining in WT and DNA2^+/− MEFs. d The representative AIRYSCAN confocal microscopy images of G4s. Scale bar = 5 μm, and e The quantification of G4 by condition. The red line indicates median. P-values were calculated using one-way ANOVA (p < 0.0001, n = 45, 42, 48 cells).; f–h DNA replication at telomeres in WT and DNA2^+/− MEFs; f Top: Scheme of the IdU (red) and CldU (green) pulse labeling. Bottom: SMARD microscopy images of replicated telomeres in WT and DNA2^+/− MEF cells. SMARD fibers are arranged showing non-telomeric DNA at the left and telomeres on the right (indicated in blue). The top panels of SMARD fibers represent fully replicated telomeres, the middle partially replicated telomeres, and the bottom where replication halted immediately adjacent to telomeres. Scale = 5 μm; g Relative length of replicated telomeres (n = 118, 148 fibers). The red line indicates the median. P-value was calculated using two-sided Student’s t-test (p = 0.0008); h Percentage of incompletely replicated telomeres and stalling adjacent to telomeres. Bars indicate mean and SD (n = 119, 148 fibers). P values were calculated using a two-sided Chi-squared test (p < 0.0001). *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001. Source data are provided as a Source data file. Created in BioRender. Zhou, T. (2025) https://BioRender.com/l90i3dw.

Fig. 2. G4 binding protein complex MutSα stimulates DNA2-mediated G4 cleavage.

a Representative silver staining SDS-PAGE from three replicates showing the pulled-down proteins by Flag-tag M2 beads in whole cell extracts from 293T cells transfected with 3x-Flag human DNA2; b DNA replication and repair proteins co-pulled down with 3x-Flag-tagged DNA2; c, d Representative co-IP and western blot analysis from three replicates of DNA2 interaction with MSH2 and MSH6; e Representative blot from three experiments of MutSα (50 nM) and MutSβ (50 nM) incubated 2 pmol substrates/µl beads. Bead-bound MutSα (MSH2 and MSH6) or MutSβ (MSH2 and MSH3) were detected by western blot; f Representative blot from three experiments with varying concentrations of MutSα (0, 25, 50, 100, 150, 200 nM) incubated with 1 µl streptavidin magnetic bead-linked G4 substrates. Bead-bound MSH2 or MSH6 was analyzed by western blot. Binding curves: the x-axis is MSH2 (g) or MSH6 (h) concentration and the y-axis is relative intensity of MSH2 (g) or MSH6 (h). Kd was determined by sigmoid curve fitting and corresponds to the concentration at which 50% of MSH2 or MSH6 bind to G4 substrates. Concentrations from each lane in f were used to determine the binding affinity of MSH2 (48 nM) and MSH6 (43 nM) to G4B oligo; i AIRYSCAN confocal microscopy of G4s; MSH2-G4 co-immunofluorescence staining in MEFs (top panel). The bottom panel shows identified spots of MSH2 and G4. Scale bar = 1 μm; j Frequency of co-localization of MSH2 and G4. Red line indicates mean ± SEM. P-value was calculated using a two-sided paired t-test (n = 79 cells, p < 0.0001); k The representative denaturing PAGE image shows DNA2 (16 nM) cleaving the ³²P-labeled G4 substrate (top) in the absence or presence of MutSα (50 nM) or XPE (100 nM); l The representative denaturing PAGE image shows cleavage of the ³²P-labeled G4 bubble substrate (top) by the DNA2 (16 nM) in the absence or presence of MutSα (50 nM) or MutSβ (50 nM). Source data are provided as a Source data file. Created in BioRender. Zhou, T. (2025) https://BioRender.com/2d65a6r.

Fig. 3. MSH2 deficiency results in G4 accumulation, defects in telomere replication.

a, b G4 immunofluorescence staining in WT, MSH2^−/−, and MSH6^−/− MEFs. a shows the representative AIRYSCAN confocal microscopy images of G4s, and b shows the quantification of G4 in WT, MSH2^-/-, and MSH6^−/− MEFs (n = 45, 51, 58 cells). The red line indicates median. All p-values were calculated using one-way ANOVA (WT vs. MSH2^−/− p = 0.0002, WT vs. MSH6^−/− p = 0.0001). Scale bar = 5 μm; c AIRYSCAN jDCV rendering of co-immunofluorescence of DNA2 and G4, and telomere FISH in WT and MSH2^−/− MEFs (top panels); spots of DNA2 and G4 and spots of DNA2 within 100 nm of a G4 spot (bottom panels); d Calculated relative co-localization frequency of DNA2 and G4 in WT and MSH2^−/− cells (n = 64, 109 cells). Red line indicates mean ± SEM. P-value was calculated using the two-sided Student’s t-test; e Relative co-localization frequency of DNA2 and G4 with telomeres in WT and MSH2^−/− cells (n = 33, 104 cells). Red line indicates mean ± SEM. P-value was calculated using the two-sided Student’s t-test (p = 0.0008); f Representative blot from three independent experiments showing increasing DNA2 binding to G4 oligos with increasing MutSα; g–i SMARD analysis of WT and MSH2^−/− MEF cells. Scale = 5 μm. g Representative microscopy images of the replicated telomeres in WT or MSH2^−/−; h Quantification of the length of replicated telomeres in different cells (n = 118, 156 fibers). The red line indicates median. P-value was calculated using a two-tailed Student’s t-test (p < 0.0001); i Percentage of partial telomere replication and replication fork stalling adjacent to telomeres in WT or MSH2^−/− cells (n = 119, 167 fibers). Bar indicates mean and SD. P-value was calculated using a two-tailed Chi-squared test (p < 0.0001); j Telomere FISH images showing telomere abnormalities in the WT, MSH2^−/−, and MSH6^−/− MEFs; k Quantification of telomere defects in WT, MSH2^−/−, and MSH6^−/− MEFs (n = 29, 15, 34 spreads). Bars indicate mean and standard deviation. All p values were calculated using the two-way ANOVA (WT vs. MSH2^−/− p < 0.0001; WT vs. MSH6^−/− p = 0.0002). *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001. Source data are provided as a Source data file.

Fig. 4. G4-stabilizing ECCs inhibit FANCJ unwinding G4, but not DNA2-mediated cleavage of G4 in vitro.

a, b Representative primer extension from three replicates by Polδ (20 nM) on the non-G4 or G4 template in the presence of increasing concentrations of G4-stabilizing ECCs PIPER (0.2, 0.5, 1 µM) (a) and capreomycin (5, 10, 50, 100 µM). b The top shows the diagram of the Polδ (20 nM)-catalyzed primer extension on a DNA template without or with a G4-forming sequence. The primer was labeled with FAM on the 5’ end. The bottom shows the representative denaturing-PAGE image of the primer extension assay; c, d Representative primer extension from three replicates on the G4 template in the absence or presence of 5 nM FANCJ with or without 0.2, 0.5, and 1 μM PIPER (c) or 10, 50, and 100 μM capreomycin (d); e, f Representative blot from three replicates showing the cleavage of a FAM³²P-labeled G4 substrate (50 nM) with DNA2 (16 nM) and MutSα (50 nM) in the absence or presence of increasing concentrations of PIPER (0.5, 1, 2, 5, 10 µM) (e) or capreomycin (1, 10, 50, 100 µM) (f); g, h G4 immunofluorescence staining in WT, DNA2^+/−, and MSH2^−/− MEFs. g The representative AIRYSCAN confocal microscopy images of G4s, and (h) shows the quantification of G4 in WT, DNA2^+/−, or MSH2^−/− MEFs without or with treatment with G4 stabilizers PIPER or capreomycin (n = 45, 47, 51, 44, 50, 50, 41, 50, 48 cells). Results are from at least 3 biological replicates. Red lines indicate median. All p-values were calculated using a two-tailed Student’s t-test (WT NT vs. WT Capreo p < 0.0001; WT NT vs. WT PIPER p < 0.0001; WT Capreo vs. DNA2+/− Capreo p = 0.021; WT PIPER vs. DNA2+/− PIPER p = 0.005; WT PIPER vs. MSH2^−/− PIPER p = 0.005). Scale bar = 5 μm. Source data are provided as a Source data file. Created in BioRender. Zhou, T. (2025) https://BioRender.com/d58fiu1.

Fig. 5. G4-stabilizing ECCs impair DNA replication at telomeres in WT DNA2^+/− and MSH2^−/− MEF cells.

a SMARD microscopy images of replicated telomeres in WT, DNA2^+/−, and MSH2^−/− MEF cells with or without exposure to PIPER or capreomycin. Scale = 5 μm; b Relative length of replicated telomeres (n = 118, 123, 149, 52, 63, 61, 87, 82, 87 fibers). Red lines indicate median. P values were calculated using a two-tailed Mann–Whitney test (WT vs. MSH2^−/− p < 0.0001; WT vs. DNA2+/− p < 0.0001; WT DMSO vs. WT PIPER p = 0.0109; WT DMSO vs. WT Capreo p = 0.0001; WT PIPER vs. DNA2^+/− PIPER p = 0.046); c Percentage of incompletely replicated telomeres (n = 119, 167, 148, 131, 96, 108, 93, 81, 119 fibers). Bars indicate the mean and error bars standard deviation. P values were calculated using a two-tailed Chi-squared test (WT vs. WT Piper p = 0.0004; WT vs. WT Capreo p = 0.0005; WT Capreo vs. DNA2^+/− Capreo p = 0.017). *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001. Source data are provided as a Source data file.

Fig. 6. G4-stabilizing ECCs enhance telomere abnormalities in WT DNA2^+/− and MSH2^−/− MEF cells.

a Telomere FISH images showing telomere abnormalities in the WT and DNA2^+/− MEFs treated with PIPER and capreomycin. Scale bar = 5 μm; b Quantification of fragile telomeres and shortened telomeres in different cells (n = 57, 44, 56, 24, 21, 22 spreads). All p values were calculated using the two-way ANOVA (WT DMSO vs. WT PIPER p < 0.0001; WT DMSO vs. WT Capreo p < 0.0001; WT DMSO vs. DNA2^+/− DMSO p < 0.0001; WT PIPER vs. DNA2^+/− PIPER p < 0.0001; WT Capreo vs. DNA2^+/− Capreo p < 0.0001; DNA2^+/− DMSO vs. DNA2^+/− PIPER p < 0.0001; DNA2^+/− DMSO vs. DNA2^+/− Capreo p = 0.0005); c Telomere FISH images showing telomere abnormalities in the WT and MSH2^−/− MEFs treated with PIPER and capreomycin. Scale bar = 5 μm; d Quantification of fragile telomeres and shortened telomeres in different cells (n = 29, 29, 32, 15, 20, 17 spreads). Bars indicate mean and error bars standard deviation. All p values were calculated using the two-way ANOVA (WT DMSO vs. WT PIPER p = 0.003; WT DMSO vs. WT Capreo p < 0.0001; WT DMSO vs. MSH2^−/− DMSO p < 0.0001; WT PIPER vs. MSH2^−/− PIPER p < 0.0001; WT Capreo vs. MSH2^−/− Capreo p = 0.012; MSH2^−/− DMSO vs. MSH2^−/− PIPER p = 0.0152);. *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001. Source data are provided as a Source data file.