CRAMP1-dependent histone H1 biogenesis is essential for topoisomerase II inhibitor tolerance

Abstract

Topoisomerase II (TOP2) inhibitors (TOP2i) are mainstay chemotherapeutic agents that undermine genome integrity by stabilizing TOP2-DNA complexes accompanied by DNA damage formation. Here, we reveal the uncharacterized protein CRAMP1 and H1 linker histones as key effectors of TOP2i tolerance in human cells. We demonstrate that CRAMP1 defines a dedicated histone H1 biogenesis factor stimulating transcription of both replicative and non-replicative H1 genes, driven by its concurrent targeting to histone gene loci and H1-specific promoter motifs. CRAMP1 promotes TOP2i tolerance by maintaining H1 supply, involving a novel mechanism uncoupled from TOP2i-induced DNA damage whereby reducing the H1 pool triggers unscheduled TOP2 substrate formation in low-accessibility chromatin states. This amplifies total demand for TOP2 activity, lowering the threshold for TOP2i-mediated exhaustion of TOP2. Our discoveries elucidate the mechanistic basis of histone H1 biogenesis in human cells, opening opportunities for selectively manipulating linker but not core histone supply and targeting cancer-associated H1 deficiency.

Keywords: chromatin; genome maintenance; histone H1; histone biogenesis; histone locus bodies; histone transcription; topoisomerase II; topoisomerase II inhibitors.

Graphical abstract

Figure 1

CRAMP1 is required for TOP2i resistance but not repair of TOP2i-induced DNA damage (A) Schematic outline of genome-scale CRISPR-Cas9 screens for genes whose KO sensitizes cells to etoposide or ICRF-193. LD20, 20% lethal dose; NGS, next-generation sequencing. (B) DrugZ analysis of sgRNA depletion in U2OS cells following 12 days of etoposide or ICRF-193 treatment (n = 2 technical replicates). NormZ values of <−3 and >3 were used as cutoffs for defining significant gene KOs conferring hypersensitivity and resistance, respectively, to the drugs. See Table S1 for full results. (C) Venn diagram of significant genes (NormZ < −3) from DrugZ analysis of CRISPR screens in (B). (D) Sulforhodamine B (SRB) cell growth assays using U2OS WT and derivative CRAMP1-KO cell lines treated with various doses of TOP2i (etoposide, ICRF-193, and doxorubicin; top) or other genotoxic agents (bottom) (mean ± SEM; n = 3 independent experiments). (E) SRB cell growth assays using U2OS WT, CRAMP1-KO, and CRAMP1-KO/CRAMP1-GFP cell lines treated with various doses of etoposide (mean ± SEM; n = 3 independent experiments). (F) Quantitative immunofluorescence analysis of γH2AX mean intensity in U2OS WT and CRAMP1-KO cells treated with etoposide (20 μM) for 1 h and collected at indicated times (mean ± SEM; n = 5 independent experiments; at least 5,000 cells analyzed per condition; unpaired two-tailed t test). (G) SRB cell growth assays using U2OS WT and CRAMP1-KO cells transfected with non-targeting control (CTRL) or ZNF451 siRNAs and treated with indicated doses of etoposide (mean ± SEM; n = 3 independent experiments). (H) Representative images of U2OS WT and CRAMP1-KO cells treated or not with etoposide (2 μM) for 1 h and fixed 48 h later. Scale bar, 10 μM. (I) Quantification of nuclear area in U2OS WT and CRAMP1-KO cells that were left untreated or exposed to etoposide (2 μM) for 1 h, then collected immediately or grown for an additional 48 h in the absence or presence of ATR inhibitor (ATRi; 1 μM) (mean ± SEM; n = 3 independent experiments; at least 2,000 cells analyzed per condition; unpaired two-tailed t test). The maximal nuclear area in untreated cells was used as a cutoff to determine the proportion of cells with increased nuclear size. (J) Quantification of RPA2 foci in G2 phase cells in (I) immunostained with RPA2 antibody (mean ± SEM; n = 3 independent experiments; at least 500 cells analyzed per condition; unpaired two-tailed t test). ^∗∗∗p < 0.001, ^∗∗p < 0.01, ^∗p < 0.05; ns, non-significant.

Figure 2

CRAMP1 is a histone H1 biogenesis factor (A) Schematic outline of MS-based analysis of CRAMP1-GFP-interacting proteins. (B) MS analysis of GFP pulldowns from U2OS Flp-In T-REx CRAMP1-KO (KO) cells stably reconstituted or not with CRAMP1-GFP (CRAMP1-KO/CRAMP1-GFP). Volcano plot shows the mean difference of the protein intensity plotted against the p value (two-tailed two-sample Student’s t test). Significant differences (q value < 0.05) were calculated by permutation-based false discovery rate (FDR) control (2,500 rounds of randomization) and are indicated in blue and red (n = 4 biological replicates). See Table S2 for full results. (C) Representative images of U2OS CRAMP1-KO/CRAMP1-GFP cells transfected with indicated siRNAs and immunostained with NPAT antibody. Scale bar, 10 μm. (D) Quantification of data in (C) (mean ± SEM; n = 3 independent experiments; at least 1,000 cells analyzed per condition; unpaired two-tailed t test). (E) RNA-seq heatmap displaying relative histone H1 transcript abundance calculated from normalized transcripts per million (TPM). For simplicity, the HIST1H1A gene encoding histone H1.1 is referred to as H1.1, HIST1H1C (encoding H1.2) as H1.2, HIST1H1E (encoding H1.4) as H1.4, H1F0 (encoding H1.0) as H1.0, and H1FX (encoding H1.10) as H1.10 throughout this study. (F) RT-qPCR analysis of expression of genes encoding different histone isoforms in U2OS WT and CRAMP1-KO cells (mean ± SEM; n = 3 independent experiments). (G) Immunoblot analysis of different histone isoforms in U2OS WT and CRAMP1-KO cells. Asterisk denotes an unspecific band. (H) Scatterplot of the pooled peptide abundance of core and linker histones (x axis) from MS/MS analysis in WT/CRAMP1-KO vs. the pooled transcript abundance (TPM) of core and linker histones from RNA-seq (y axis) in WT/CRAMP1-KO. (I) RNA-seq scatterplot and Venn diagram of the log₂ fold change of genes significantly altered in WT vs. CRAMP1-KO and CRAMP1-KO vs. CRAMP1-KO/CRAMP1-GFP cell lines. Differentially expressed genes were defined as having log₂ fold change ≥ 0.58 and adjusted p > 0.05 (n = 3 technical replicates). ^∗∗∗p < 0.001, ^∗∗p < 0.01, ^∗p < 0.05; ns, non-significant.

Figure 3

Histone H1 biogenesis is mediated by a CRAMP1-GON4L-NPAT pathway (A) AlphaFold2 model (top) and domain structure (bottom) of human CRAMP1. Insets (right) show zoomed-in views of the SANT/Myb (dark purple) and GB (light purple) domains. Deletions introduced to generate the ΔSANT, ΔGB-N, and ΔGB-C mutants are indicated. (B) RNA-seq heatmap displaying relative histone H1 transcript abundance from normalized TPM in the indicated cellular backgrounds. (C) Representative images of U2OS CRAMP1-KO/CRAMP1-GFP cell lines immunostained with NPAT antibody. Scale bar, 10 μm. (D) Quantification of data in (C) (mean ± SEM; n = 3 independent experiments; at least 1,000 cells analyzed per condition; unpaired two-tailed t test). (E) AlphaFold2-Multimer model of the CRAMP1 (306–800)-GON4L (1455–2024) interaction. (F) Immunoblot analysis of GFP IPs from U2OS WT, U2OS/CRAMP1-KO, and U2OS CRAMP1-KO/CRAMP1-GFP cell lines transfected with FLAG-GON4L expression construct. (G) AlphaFold-Multimer model of the complex between CRAMP1, GON4L, and NPAT. (H) MS analysis of GFP pulldowns from U2OS CRAMP1-KO/CRAMP1-GFP vs. CRAMP1-KO/CRAMP1-GFP ΔGB-N cells. Volcano plot shows the mean difference of the protein intensity plotted against the p value (two-tailed two-sample Student’s t test). Significant differences (q value < 0.05) were calculated as in Figure 2B (n = 4 biological replicates). See Table S2 for full results. ^∗∗∗p < 0.001; ns, non-significant.

Figure 4

CRAMP1 functions as a histone H1 transcription factor (A) Hilbert curves showing CUT&RUN signal intensities on chromosome 6 for U2OS CRAMP1-KO/CRAMP1-GFP WT, ΔSANT, and ΔGB-N cells. Merged spike-in normalized reads per million signals from n = 3 technical replicates. (B) CUT&RUN signal intensity (spike-in normalized reads per million [RPM]) in U2OS CRAMP1-KO cell lines stably expressing CRAMP1-GFP WT, ΔSANT, or ΔGB-N around TSSs for all genes (left), core histone genes (center), or H1 genes (right) (mean signal in n = 3 technical replicates). (C) CUT&RUN profiles of U2OS CRAMP1-KO/CRAMP1-GFP WT, ΔSANT, and ΔGB-N cells at different histone gene loci. Merged signal from 3 technical replicates, quantitated using spike-in normalized reads per million in 1 kb non-overlapping bins. Dashed lines highlight promoter regions of indicated genes. (D–F) Rank plots showing genes differentially expressed and differentially bound by CRAMP1 for indicated comparisons between U2OS WT, U2OS/CRAMP1-KO, and U2OS CRAMP1-KO/CRAMP1-GFP cell lines. Only differentially expressed genes (log₂ fold change ≥ 0.58 and adjusted p values < 0.05) were included in the analysis. Genes were ranked based on the log₁₀ rank product calculated using BETA. (G) RT-qPCR analysis of H1.4 pre-mRNA levels relative to mature H1.4 transcript levels in U2OS cells transfected with indicated siRNAs (mean ± SEM; n = 3 biological replicates). (H) Schematic of CRAMP1 targeting to histone H1 promoters via concordant recognition of H1 promoter motifs by the SANT/Myb domain and recruitment to histone promoters via the GON4L-binding GB domain.

Figure 5

CRAMP1 promotes TOP2i tolerance via histone H1 biogenesis (A) SRB cell growth assays using U2OS/CRAMP1-KO and U2OS CRAMP1-KO/CRAMP1-GFP cell lines treated with various doses of etoposide (mean ± SEM; n = 3 independent experiments). (B) Quantitative immunofluorescence analysis of γH2AX mean intensity in U2OS WT, U2OS/CRAMP1-KO, and U2OS CRAMP1-KO/CRAMP1-GFP cell lines treated or not with etoposide (20 μM) for 1 h (mean ± SEM; n = 3 independent experiments; at least 5,000 cells analyzed per condition; unpaired two-tailed t test). (C) Schematic outline of base editor tiling screen for residues in human CRAMP1 required for etoposide or ICRF-193 tolerance. (D) Base editor tiling analysis of human CRAMP1, carried out as shown in (C). x axis shows amino acid position of CRAMP1. y axis displays log₂ fold changes for etoposide (red) or IRCF-193 treatment (blue) relative to mock treatment. Each circle represents a single sgRNA. Large circles indicate sgRNAs with significant changes (p ≤ 0.01). Lower bar plot shows sequence conservation among metazoan CRAMP1 orthologs. (E) SRB cell growth assays using U2OS cells or a derivative cell line stably overexpressing indicated histone H1 isoforms (H1+) transfected with siRNAs and treated with various doses of etoposide (mean ± SEM; n = 3 independent experiments). (F) Quantitative immunofluorescence analysis of γH2AX mean intensity in cells in (E), collected after a 1-h treatment with etoposide (20 μM) (mean ± SEM; n = 3 independent experiments; at least 5,000 cells analyzed per condition; unpaired two-tailed t test). (G) Correlation analysis of NormZ values from the etoposide CRISPR-Cas9 screen (x axis) vs. RNA-seq transcript abundance (TPM) (y axis) for histone H1 isoforms in U2OS cells (mean ± SEM, n = 3 technical replicates). R² value was calculated using linear regression. (H) SRB cell growth assays using U2OS cells transfected with non-targeting control (CTRL) siRNA or an siRNA pool targeting different H1 isoforms and treated with various doses of etoposide (mean ± SEM; n = 3 independent experiments). ^∗∗∗p < 0.001, ^∗∗p < 0.01, ^∗p < 0.05; ns, non-significant.

Figure 6

Histone H1 deficiency drives TOP2i hypersensitivity by elevating TOP2 substrate formation and demand (A) Representative images of U2OS cells transfected sequentially with CRAMP1 siRNA and indicated GFP expression constructs and co-immunostained with TOP2A and fibrillarin antibodies. Scale bar, 10 μM. (B) Quantification of mean intensity of TOP2A (nucleolar/extranucleolar ratio) for the experiment in (A) (mean ± SEM; n = 3 biological replicates; at least 50 cells analyzed per condition). (C) Average of calibrated TOP2 CC-seq signals around protein-coding TSSs for U2OS WT, CRAMP1-KO, and CRAMP1-KO/CRAMP1-GFP cells (mean of n = 4 biological replicates). (D) Bar plot comparing fold change in TOP2 CC-seq signal densities in each ChromHMM-defined chromatin state, between U2OS WT, CRAMP1-KO, and CRAMP1-KO/CRAMP1-GFP cell lines. The relative accessibility of the chromatin states is indicated (mean ± SEM of n = 4 biological replicates). (E) TOP2 CC-seq coverage averaged per chromosome in regions with different chromatin accessibility defined using ATAC-seq signal in WT and CRAMP1-KO cells. Different ATAC-seq accessibility regions were defined using ChromHMM chromatin states split into three categories as in (D) (mean TOP2cc coverage from n = 2 biological replicates). (F) Venn diagram showing proportion of transcripts upregulated in CRAMP1-KO cells and corrected by stable complementation with CRAMP1-GFP that map to heterochromatic regions defined by ChromHMM (p = 6.04 × 10⁻²² when comparing to transcripts not regulated by CRAMP1 KO; Fisher’s exact test). (G) Immunoblot analysis of TOP2A expression in HCT116-TOP2A-mAID cells after treatment with indicated doses of auxin for 72 h. (H) SRB cell growth assays using HCT116-TOP2A-mAID transfected with control (CTRL) or CRAMP1 siRNAs and grown in presence of indicated auxin doses (mean ± SEM; n = 3 independent experiments). (I) As in (H), except cells were transfected with a pool of siRNAs targeting different H1 isoforms (mean ± SEM; n = 3 independent experiments). (J) Quantification of nuclear area (analyzed as in Figure 1I) in U2OS WT and CRAMP1-KO cells that were treated or not with etoposide (2 μM) for 1 h, transfected with GFP-TOP2A plasmid or empty vector 24 h later, and collected after an additional 48 h (mean ± SEM; n = 3 independent experiments; at least 2,000 cells analyzed per condition; unpaired two-tailed t test). (K) Incucyte analysis of cell area (μm²) of HCT116-TOP2A-mAID cells transfected with indicated siRNAs and grown in the presence of auxin (1.5 μM) (mean ± SEM; n = 3 independent experiments). ^∗∗∗p < 0.001, ^∗∗p < 0.01, ^∗p < 0.05; ns, non-significant.

Figure 7

CRAMP1-dependent histone H1 biogenesis is critical for the TOP2i tolerance: model (A) In conjunction with GON4L and NPAT, CRAMP1 promotes transcription of both replicative H1 genes within HLBs and non-replicative H1 genes. (B) Defective H1 supply increases TOP2 engagement in low-accessibility chromatin states undergoing decompaction, lowering the threshold for TOP2i cytotoxicity.