Chemical-induced phase transition and global conformational reorganization of chromatin - PubMed (original) (raw)
doi: 10.1038/s41467-023-41340-4.
Shuxiang Shi # 1 2, Yuanyuan Shi # 3 4, Peipei Jiang # 3 4, Ganlu Hu # 5, Qinying Ye 1, Zhan Shi 6, Kexin Yu 1 7, Chenguang Wang 3 4 8, Guoping Fan 5, Suwen Zhao 1 7, Hanhui Ma 1, Alex C Y Chang 3 4 8, Zhi Li 6, Qian Bian 9 10, Chao-Po Lin 11
Affiliations
- PMID: 37689690
- PMCID: PMC10492836
- DOI: 10.1038/s41467-023-41340-4
Chemical-induced phase transition and global conformational reorganization of chromatin
Tengfei Wang et al. Nat Commun. 2023.
Abstract
Chemicals or drugs can accumulate within biomolecular condensates formed through phase separation in cells. Here, we use super-resolution imaging to search for chemicals that induce phase transition within chromatin at the microscale. This microscopic screening approach reveals that adriamycin (doxorubicin) - a widely used anticancer drug that is known to interact with chromatin - specifically induces visible local condensation and global conformational change of chromatin in cancer and primary cells. Hi-C and ATAC-seq experiments systematically and quantitatively demonstrate that adriamycin-induced chromatin condensation is accompanied by weakened chromatin interaction within topologically associated domains, compartment A/B switching, lower chromatin accessibility, and corresponding transcriptomic changes. Mechanistically, adriamycin complexes with histone H1 and induces phase transition of H1, forming fibrous aggregates in vitro. These results reveal a phase separation-driven mechanism for a chemotherapeutic drug.
© 2023. Springer Nature Limited.
Conflict of interest statement
The authors declare no competing interests.
Figures
Fig. 1. Adriamycin induces chromatin condensation in cancer and primary cells.
a, b Screening of chemicals that alter chromatin structures at mesoscale by super-resolution microscopy. Only adriamycin, but not the other TOP2 poison etoposide, resulted in a punctate pattern of chromatin in U2OS cells (a). Similar phenomenon could also be observed in HCT116 and HeLa cells (b). Scale bar, 10 μm. Asterisk, low-density DNA regions. c Heatmaps of STED images demonstrated the differential distribution of DNA in control (randomly distributed) and adriamycin-treated samples. Scale bar, 10 μm. d Quantification of the degree of clustering by RDF and L-function (see “Methods” and Supplementary Fig. 3). Adriamycin induced significant clustering of chromatin compared to the control. The ribbon plots are employed to show means +/− SD (the width of ribbons; n = 10 cells). e Electron microscopy revealed different morphologies of chromatin condensates. Scale bar, 10 μm. White arrows indicate the condensates formed in the proximity of nuclear envelops. f Immunostaining of Lamin A/C (red) with DAPI staining (green) in U2OS cells. Scale bar, 10 μm. White arrows indicate the condensates formed in the proximity of nuclear envelops. Ctrl control, adria adriamycin. g, h Adriamycin induced significant chromatin condensation in primary cardiomyocytes, but to a much less extent in primary hepatocytes. g Primary cells isolated from P1 mice were cultured for 3–5 days, followed by 1.5 μg/ml adriamycin treatment for 4 h. Cells were immunostained with cardiomyocyte (Tnni3) and hepatocyte (Alb) markers, as well as DAPI. Cells positive for adriamycin were examined for their chromatin conformation. h The differential accumulation of adriamycin in different cell populations. Dashed circles indicate the nuclear outlines of cardiomyocytes (Tnni3+) or hepatocytes (Alb+). Yellow arrows indicate non-cardiomyocytes (Tnni3−) or non-hepatocytes (Alb-). Scale bar, 10 μm. For each group in the bar graph, three fields (n = 3) were calculated and totally 78, 79, 109, and 79 cells were counted for each group. Data are presented as means +/− SD in the bar graph. The ratios of cells showing condensed chromatin (as in Fig. 1g) in those four populations are shown above each bar. Ns not significant; ***P < 0.001 (two-tailed unpaired t test). Source data are provided as a Source data file. Experiments of (a–c, g, h) were repeated at least three times, and experiments of (e) and (f) were repeated twice with similar results.
Fig. 2. Adriamycin forms complexes with chromatin condensates.
a Comparison of dynamics of two DNA-intercalating reagents, adriamycin and cisplatin, in mouse embryonic fibroblasts (MEFs). As revealed by Airyscan2 microscope, adriamycin entered nuclei exhibited the fibrous morphology, followed by a transition towards dense aggregates. Cisplatin was conjugated with Texas Red for visualization while adriamycin has intrinsic excitation wavelength at 514 nm. Arrows, large puncta which could possibly be heterochromatin. Scale bar, 10 μm. b FRAP analysis showed the material exchange of adriamycin within condensates. Scale bar, 10 μm. Signals were corrected for photobleaching using a similarly sized unbleached area and then normalized to the ratio between the average intensity of the pre-bleach images and the lowest post-bleach intensity. The signal intensities are presented as means +/− standard deviation (SD) (n = 10–15 cells per condition). c Adriamycin co-localized with condensed chromatin in U2OS and HCT116 cells. Adriamycin appeared fibrous or dotted at the 0.5 h time point, while the chromatin appeared homogenous. The co-localization became apparent at the 4 h time point. Scale bar, 10 μm. d Live imaging of the adriamycin–DNA condensates by the SIM super-resolution microscope. U2OS cells were incubated with 0.5 μM SiR-Hoechst overnight and treated with 1.5 μg/ml for 2 h before imaging. Scale bar, 10 μm. e Co-aggregated adriamycin and chromatin were reversible. U2OS cells were treated with 1.5 μg/ml adriamycin for 4 h, followed by drug removal for 12 and 24 h. The punctate pattern of adriamycin–DNA was almost completely disappeared after 24 h reversal. Scale bar, 10 μm. Quantification of % of punctate cells in each condition was performed on 3 separate fields (n = 3), each containing ~50 cells. Data are presented as means +/− SD. Ctrl, control; adria, adriamycin. Source data are provided as a Source data file. Experiments of (a) and (b) were repeated at least three times, and experiments of (c) and (e) were repeated twice with similar results.
Fig. 3. Adriamycin–chromatin condensates co-localize with heterochromatin.
a, b U2OS cells were pulse-labeled with dUTP-Cy5 to label heterochromatin (a) or euchromatin (b), depending on the status of cell cycle. Cells were live-imaged at ×63 upon 1.5 μg/ml adriamycin treatment on Leica Thunder Imager microscope. As heterochromatin regions indicated by arrows, adriamycin accumulated primarily in heterochromatin, not euchromatin regions. Scale bar, 10 μm. c, d Adriamycin–chromatin condensates were co-localized with heterochromatin marker H3K9me3 (c) and transfected HP1-CFP (d). Scale bar, 10 μm. e, f Adriamycin condensates were partially co-localized with euchromatin marker H3K27ac (promoters and enhancers), but were localized mutual exclusively with H3K4me3 (promoters). Arrows indicate adriamycin condensates. Scale bar, 10 μm. g Adriamycin-induced nuclear p53 and adriamycin mainly exhibited mutually exclusive occupancy in nuclei. Arrows indicate the minority p53 that located within adriamycin condensates. Scale bar, 10 μm. Source data are provided as a Source data file. All experiments were repeated twice with similar results.
Fig. 4. Adriamycin induces genome-wide changes in chromatin accessibility.
a Volcano plots showed the changes of ATAC-Seq peaks (top row) or gene expression (bottom row) upon etoposide (left) or adriamycin (right) treatment. Adriamycin, but not etoposide, induced a large amount of gain/loss peaks. b Averaged line graph and heatmaps show the ATAC-Seq signal intensities surrounding the TSSs of genes in control and adriamycin-treated cells. Adriamycin induced a general loss of chromatin accessibility surrounding TSSs. c Adriamycin-specific gene repression was correlated with the loss of chromatin accessibility surrounding regions.
Fig. 5. Adriamycin induces genome-wide 3D chromatin conformational change.
a Hi-C heatmaps binned at 10 kb show chromatin interactions patterns of representative regions on Chromosome 1 in control and adriamycin-treated U2OS cells. Multiple TADs and loops are diminished in adriamycin-treated cells. Prob. probability. b P(s) curves indicate relationships between chromatin contact probability and genomic distances for chromatin interactions on autosomes in control and adriamycin-treated cells. c Box plots quantify the insulation scores and boundary strength for TAD boundaries in control (ctrl) and adriamycin (adria)-treated cells. Boxes, middle 50% of TAD boundary strength. Center bars, medians of boundary strength. Whiskers, 1.5× interquartile range. p values are calculated from one-tailed Mann–Whitney _U_-test. n = 2 replicates (independently cultured cells were harvested, processed and sequenced separately) for each condition. d Venn diagram depicts the overlap between TAD boundaries identified in control (red) and adriamycin-treated (blue) cells. e Averaged insulation profiles in control (red) and adriamycin-treated (blue) cells for 1 Mb genomic regions centered at the 10 kb genomic bins containing control-specific (left), shared (middle), or adria-specific TAD boundaries (right). f Insulation profiles for a representative genomic region in control (red) and adriamycin-treated (blue) cells. Bars below insulation profiles indicate control-specific (red), shared (black), or adria-specific (blue) TAD boundaries identified in this region. g–i An example region (g) and overall statistic (h) demonstrated the switch of compartment A/B upon adriamycin treatment. i The change of contact frequencies between compartments upon adriamycin treatment. Genomic regions belong to the B compartment (B–B) exhibited a notable increase. j, k Relationship between gene expression and genomic A/B compartments. j Left, the boxplots showing E1 scores of all differentially expressed genes. Boxes, middle 50% of E1 scores. Center bars, medians of E1. Whiskers, 1.5× of inter-quartile range. Right, KEGG analysis of the pathways enriched in down-regulated genes with decreased E1. p values were calculated by two-sided Wilcoxon test. n = 2 technical replicates for each condition. Examples were showed in (k). l Histone modification-based learning and annotation of compositions of A/B compartments in U2OS cells performed by ChromHMM analysis (see “Methods” for the detail). The correlations between compartment switches and different chromatin states were analyzed and shown in the heatmap (right).
Fig. 6. Adriamycin treatment leads to suppression of TE (transposable elements) expression.
a Volcano plots show differentially regulated TE transcripts after etoposide and adriamycin treatment. TE transcripts with abs(log2(foldchange)) >1, Padj <0.05 were highlighted by cyan color. b Heatmap showing relative expression of differentially regulated TE transcripts in each treatment groups. c Normalized expression of top 50 adriamycin induced downregulated TE transcripts in each treatment groups. The boxplot shows the normalized gene expression (read counts) in control, etoposide-, and adriamycin-treated conditions. Boxes, middle 50% of normalized gene expression. Center bars, medians of normalized gene expression. Whiskers, 1.5× of inter-quartile range. p values were calculated by two-sided Wilcoxon test. n = 2 technical replicates for each condition. d Pie chart showing species classification of top 50 adriamycin induced downregulated TE transcripts. e Relative expression of selective 12 TE transcripts downregulated upon adriamycin treatment. The boxplots show the normalized expression (read counts) of TEs in control, etoposide-, and adriamycin-treated conditions. Boxes, middle 50% of normalized gene expression. Center bars, medians of normalized gene expression. Whiskers, 1.5× of inter-quartile range. f, g Examples of TE-containing non-coding RNAs (f) and TEs (g) demonstrate the correlation between compartment A to B transition and expression repression induced by adriamycin. Ctrl control, adria adriamycin.
Fig. 7. Adriamycin induces phase transition of chromatin and histone H1 in vitro.
a Isolation of native chromatin and further purification of its DNA fraction. b Adriamycin did not induce aggregation of the DNA fraction of chromatin, but 0.1 mg/ml Hoechst 33342 did. In contrast, adriamycin induced the phase transition of native chromatin which contains the linker histone H1 and nucleosome histones, suggesting this phase separation phenomenon is dependent on chromatin-associated proteins. Scale bar, 10 μm. The experiment was repeated at least three times with similar results. c, d SPR analysis revealed the affinities between adriamycin (c) and mono-nucleosomes in which H1 is absent (d). Experiment of b and c were repeated twice with similar results. e In vitro condensation of adriamycin and H1-CFP. By itself, neither adriamycin nor H1-CFP formed condensate. Mixing adriamycin and H1-CFP led to the phase separation, forming condensates containing both adriamycin and H1-CFP. Scale bar, 10 μm. BF, bright field. f The “expansion–fusion” dynamics of adriamycin condensates. Also see the text for the description. Scale bar, 1 μm. g FRAP experiment demonstrated the diffusible property of adriamycin within adriamycin-H1-CFP condensates. Scale bar, 10 μm. h The co-localization of adriamycin, histone H1 and DNA in cells. U2OS cells were treated with 1.5 μg/ml adriamycin (adria), fixed at indicated time points, and stained with anti-H1 antibody and DAPI. Scale bar, 10 μm. Source data are provided as a Source data file. i Proposed model for adriamycin-induced chromatin reorganization.
References
- Strzelecka M, et al. Coilin-dependent snRNP assembly is essential for zebrafish embryogenesis. Nat. Struct. Mol. Biol. 2010;17:403–409. -PubMed
Publication types
MeSH terms
Substances
LinkOut - more resources
Full Text Sources
Molecular Biology Databases