HLTF resolves G4s and promotes G4-induced replication fork slowing to maintain genome stability

Abstract

G-quadruplexes (G4s) form throughout the genome and influence important cellular processes. Their deregulation can challenge DNA replication fork progression and threaten genome stability. Here, we demonstrate an unexpected role for the double-stranded DNA (dsDNA) translocase helicase-like transcription factor (HLTF) in responding to G4s. We show that HLTF, which is enriched at G4s in the human genome, can directly unfold G4s in vitro and uses this ATP-dependent translocase function to suppress G4 accumulation throughout the cell cycle. Additionally, MSH2 (a component of MutS heterodimers that bind G4s) and HLTF act synergistically to suppress G4 accumulation, restrict alternative lengthening of telomeres, and promote resistance to G4-stabilizing drugs. In a discrete but complementary role, HLTF restrains DNA synthesis when G4s are stabilized by suppressing primase-polymerase (PrimPol)-dependent repriming. Together, the distinct roles of HLTF in the G4 response prevent DNA damage and potentially mutagenic replication to safeguard genome stability.

Keywords: DNA replication stress response; DNA translocase; G-quadruplex; HLTF; MSH2; PrimPol; RNA-DNA hybrid; alternative lengthening of telomeres; genome stability; nucleic acid secondary structure.

Figure 1.. The cell cycle independent regulation of HLTF chromatin binding by MSH2 and transcription. See also Figure S1 and Table S1&2.

A. Schematic of iPOND-SILAC-MS. “Heavy” and “light” amino acids labeling of WT or HLTF-KO cells was reversed for the two biological repeats. B. Volcano plots for proteins identified by iPOND-SILAC-MS. p-values are calculated based on two biological repeats. Protein abundance changes with at least 50% decrease (green) or increase (red) are highlighted. Significance cutoff for protein enrichment was set at p=0.05 (horizontal dotted line). C. HLTF-KO/WT SILAC ratio for proteins from B. For CMGs, RFCs, RPAs, and replicative polymerases, mean ± SEM is calculated from the normalized SILAC ratio of each subunit comprising the complex. For MSH2, 3 and 6, the normalized SILAC ratio is used to calculate the mean ± SD (n=2). D. Representative immunofluorescence (IF) images of U2OS cells expressing GFP-HLTF after sgRNA-mediated knockdown. GFP-HLTF is detected using a GFP antibody. E. Mean intensity of chromatin-bound MSH2 as shown in D. See also Figure S1E. F. Mean intensity of chromatin-bound GFP-HLTF as shown in D. See also Figure S1G. G. Representative IF images of U2OS cells expressing GFP-HLTF after mock or DRB treatment (100 μM, 4h). GFP-HLTF is detected using a GFP antibody. H. Mean intensity of chromatin-bound GFP-HLTF from G. See also Figure S1L. Mann-Whitney tests were performed for all data shown in this figure.

Figure 2.. HLTF suppresses G4 accumulation in cells and is enriched at G4 structures in the human genome. See also Figure S2.

A. Representative IF images in WT and HLTF-KO U2OS cells. G4s are detected using the 1H6 antibody. B. Mean G4 intensity from A. See also Figure S2C. C. Mean G4 intensity in WT and HLTF-KO U2OS cells after sgRNA-mediated knockdown. See also Figure S2G. Mann-Whitney tests were performed in B and C. D. Representative browser tracks of G4 ChIP-seq (GSE162299) and HLTF ChIP-seq at the MYC and KRAS loci. E. Heatmaps showing ChIP-seq coverage for G4 and HLTF at G4 CUT&Tag peaks in U2OS cells (n=35,104) (GSE181373). The x-axis represents the distance from the peak in kb. Heatmaps were sorted by G4 ChIP-seq (GSE162299) signal intensity for all ChIP-seq samples. Spearman correlation coefficient between G4 and HLTF sgCON ChIP-seq signal was 0.74. F. Aggregate plot showing HLTF ChIP-seq coverage (y-axis) relative to the distance (x-axis, in kb) from G4 CUT&Tag peaks, related to E. G. Distribution of HLTF ChIP-seq peaks within each genomic compartment. TTS, transcription termination site. H. Relative enrichment of G4 CUT&Tag and HLTF ChIP-seq peaks within each genomic compartment. * indicates compartments where both G4 and HLTF showed significant enrichment. I. Frequency distribution of the number of G4 motifs within G4 CUT&Tag peaks. G4 peaks are segregated based on overlap with HLTF ChIP-seq peaks: G4 peaks that overlap with HLTF peaks are HLTF⁺ or otherwise HLTF⁻. See also Figure S2J. J. Aggregate plot showing HLTF ChIP-seq coverage (y-axis) relative to the distance (x-axis, in kb) from G4 CUT&Tag peaks. G4 peaks are segregated based on whether they are within the promoter or 5’UTR (G4 within promoter/5’UTR denoted promoter/5’UTR⁺, n=17,620; G4 out of promoter/5’UTR denoted promoter/5’UTR⁻, n=17,484). Promoters are identified as ±1kb of the annotated transcription start site. See also Figure S2K.

Figure 3.. HLTF is enriched at RNA-DNA hybrids stabilized by G4s. See also Figure S3.

A. Heatmaps showing ChIP-seq coverage for HLTF at RNA-DNA hybrid peaks identified in U2OS cells by DRIP-seq (GSE115957). The x-axis represents the distance from the DRIP peak in kb. DRIP peaks are segregated based on whether they overlap with G4 peaks identified by G4 CUT&Tag: DRIP peaks that overlap with G4 peaks are G4⁺(n=11,499) or otherwise G4⁻ (n=59,090). B. Aggregate plot showing HLTF ChIP-seq coverage (y-axis) relative to the distance in kb (x-axis) from DRIP peaks. Related to A. See also Figure S3A. C. Representative IF images of RNA-DNA hybrids in WT and HLTF-KO U2OS cells after mock or RNase H digestion. Hybrids are detected using purified, recombinant GFP-dRNH1. D. Mean nuclear hybrid intensity, as shown in C. See also Figure S3D. E. Mean nuclear hybrid intensity in WT and HLTF-KO U2OS after sgRNA transfection. See also Figure S3E. Mann-Whitney tests were performed in this figure.

Figure 4.. HLTF suppresses G4s in an ATPase-dependent manner in cells. See also Figure S4.

A. Schematic of HLTF domain structures. Arrowheads represent position of mutations. B. Western blot of HLTF-dependent Ub chain formation by UBC13/MMS2 using a ubiquitin antibody. C. ATPase rates of HLTF WT and mutant proteins measured using an NADH-coupled assay. Rates were corrected for background NADH decomposition in a no-enzyme control. Data are plotted as the mean ± SD (n=3). See also Figure S4A. D. Denaturing PAGE showing the separation of DNA substrate (fork) and product (duplex), as a measurement of the in vitro fork reversal activity of HLTF WT and mutant proteins. E. Quantification of the in vitro fork reversal activity of HLTF WT and mutant proteins. Related to D. Data are plotted as the mean ± SD (n=3). F. Mean G4 intensity in U2OS WT and HLTF-KO cells constitutively expressing HLTF WT or mutant proteins. Two clones of each cell line were analyzed separately. The median of the mean G4 intensity of the two clones is averaged to calculate the mean ± SEM (n=3). One-way ANOVA was performed followed by Dunnett’s test. G. Mean G4 intensity in U2OS WT and HLTF-KO cells inducibly expressing HLTF R890Q mutant, after dox induction (24 h, 500 ng/mL). Two clones of R890Q expressing HLTF-KO cell lines were analyzed separately. The median of the mean G4 intensity of the two clones is averaged to calculate the mean ± SEM (n=3). T-test was performed.

Figure 5.. HLTF promotes ATP-dependent G4 unfolding in dsDNA. See also Figure S5.

A. Schematic of ssDNA G4 unfolding assay. B. Native gel images showing ssDNA G4 unfolding by HLTF (top) or S. cerevisiae Pif1 (bottom). C. Quantification of ssDNA G4 unfolding by HLTF or S. cerevisiae Pif1. Related to B. D. Schematic of dsDNA G4 unfolding assay. E. Representative native gel image showing the dsDNA G4 unfolding by HLTF WT or a HIRAN (N90A,N91A) mutant. DNA substrate in all lanes was digested with EcoRI. F. Quantification of dsDNA G4 unfolding by HLTF WT or a HIRAN (N90A,N91A) mutant. Lane numbers correspond to those shown in E. Data is presented as mean ± SEM (n=3).

Figure 6.. HLTF suppresses ALT activity in an ATPase-dependent manner. See also Figure S6.

A. Representative IF images of ALT-associated PML body (APB) detection in WT and HLTF-KO U2OS cells. Arrowheads mark APBs. B. Percentage of cells with at least five APBs. Related to A. Data are represented as mean ± SEM (n = 3). C. Representative IF images of in situ ssTelo-C foci detection in WT and HLTF-KO U2OS cells. D. Quantification of ssTelo-C foci/cell (n=3). Related to C. E. Quantification of ssTelo-C foci/cell (n=4) in WT and HLTF-KO U2OS cells, after transfection with the indicated sgRNA. F. Percentage of cells with at least 5 APBs in WT and HLTF-KO U2OS cells, and HLTF-KO cells expressing HLTF WT, R71E or C760S mutant proteins. Data are represented as mean ± SEM (n ≥ 3). G. Percentage of cells with at least 5 APBs after dox induction (500 ng/mL, 24 h) in WT and HLTF-KO U2OS cells, and HLTF-KO cells expressing the R890Q mutant. Data are represented as mean ± SEM (n = 3). All statistical tests in this figure are one-way ANOVA followed by Dunnett’s test.

Figure 7.. HLTF restrains replication fork progression and protects cells from DNA damage and growth defects in response to G4-stabilization. See also Figure S7.

A. Experimental setup for replication fork progression assay. B. EdU tract lengths (n=3) in mock or PDS-treated U2OS cells. C. EdU tract lengths (n=3) in U2OS cells 72 h after siRNA transfection. In B & C, line represents the median. ns, not significant, by Kruskal-Wallis test; **** p < 0.0001, by Mann-Whitney test. D. Percentage of RPE1 cells with at least 15 γH2AX foci with mock or PDS treatment (5 μM, 24 h) after sgRNA transfection. Data are represented as mean ± SEM (n=3). T-test results comparing the sgCON and sgHLTF PDS-treated samples are shown. E. PDS-response curve in RPE1 cells transfected with the indicated sgRNAs. sgRNAs against both HLTF and MSH2 are transfected in the sgDKO sample. Data are represented as mean ± SEM (n=4). F. IC50 of RPE1 cells after sgRNA transfection and PDS treatment. Related to E. Data are represented as mean ± SEM (n=4). ** p < 0.01; *** p < 0.001; ns, not significant, by one-way ANOVA followed by Dunnett’s test compared to sgCON; * p < 0.05 by t-test. G. Proposed model for the dual roles of HLTF in regulating G4s to maintain genome stability. In its first role (left), HLTF acts in all cell cycle phases and uses its ATPase-dependent dsDNA translocase activity to travel on dsDNA. Upon encounter of the G4, the translocation activity of HLTF may destabilize the structure, allowing the structure forming sequence to reanneal to its complementary strand. MutS complexes can also bind and regulate G4s in a distinct pathway that is independent of HLTF. In a second role (right), HLTF slows DNA synthesis in response to G4 stabilization. G4s may be resolved by other resolution factors as fork reversal occurs or through reannealing of the parental DNA. In HLTF’s absence, G4s at the replication fork are bypassed by PrimPol-mediated repriming.