Super-resolution imaging reveals the evolution of higher-order chromatin folding in early carcinogenesis - PubMed (original) (raw)

. 2020 Apr 20;11(1):1899.

doi: 10.1038/s41467-020-15718-7.

Hongqiang Ma # 1, Hongbin Ma 1 2 3, Wei Jiang 1 4, Christopher A Mela 1, Meihan Duan 1 5, Shimei Zhao 1 6, Chenxi Gao 7 8, Eun-Ryeong Hahm 7 8, Santana M Lardo 9, Kris Troy 9, Ming Sun 10, Reet Pai 11, Donna B Stolz 10, Lin Zhang 7 8, Shivendra Singh 7 8, Randall E Brand 8 12, Douglas J Hartman 11, Jing Hu 8 12, Sarah J Hainer 13, Yang Liu 14 15 16

Affiliations

Super-resolution imaging reveals the evolution of higher-order chromatin folding in early carcinogenesis

Jianquan Xu et al. Nat Commun. 2020.

Abstract

Genomic DNA is folded into a higher-order structure that regulates transcription and maintains genomic stability. Although progress has been made on understanding biochemical characteristics of epigenetic modifications in cancer, the in-situ higher-order folding of chromatin structure during malignant transformation remains largely unknown. Here, using optimized stochastic optical reconstruction microscopy (STORM) for pathological tissue (PathSTORM), we uncover a gradual decompaction and fragmentation of higher-order chromatin folding throughout all stages of carcinogenesis in multiple tumor types, and prior to tumor formation. Our integrated imaging, genomic, and transcriptomic analyses reveal functional consequences in enhanced transcription activities and impaired genomic stability. We also demonstrate the potential of imaging higher-order chromatin disruption to detect high-risk precursors that cannot be distinguished by conventional pathology. Taken together, our findings reveal gradual decompaction and fragmentation of higher-order chromatin structure as an enabling characteristic in early carcinogenesis to facilitate malignant transformation, which may improve cancer diagnosis, risk stratification, and prevention.

PubMed Disclaimer

Conflict of interest statement

The authors declare no competing interests.

Figures

Fig. 1

Fig. 1. Super-resolution imaging of heterochromatin structure in _Apc_Min/+ mouse model.

a, g, m, s H&E-stained pathology images and d, j, p, v STORM-based super-resolution images of H3K9me3-dependent heterochromatin from normal tissue from wild-type, histologically normal-appearing tissue from 6-week and 12-week _Apc_Min/+, and tumor (adenoma) from 12-week _Apc_Min/+ mice. Scale bar: 10 µm. b, h, n, t Conventional wide-field fluorescence images and e, k, q, w corresponding STORM images of heterochromatin from a single nucleus. (c, i, o, u) and (f, l, r, x) are progressively zoomed regions of (b, h, n, t) and (e, k, q, w). Scale bar in (e, k, q, w): 2 µm; scale bar in (f, l, r, x): 500 nm. y The box-and-whisker plots of the H3K9me3 cluster size, where the central line of the box indicates the median; the bottom/top edges of the box indicate 25th/75th percentiles; the whiskers extend to the most extreme data points without outliers from four groups. They include wild-type (n = 289 cells), normal-appearing tissue from 6-week (n = 259 cells) and 12-week _Apc_Min/+ (n = 99 cells) and tumor (adenoma) from 12-week _Apc_Min/+ mice (n = 55 cells). Cell nuclei from three mice for each category were analyzed. For each mouse, approximately five randomly selected areas of normal-appearing cells located in the villi regions were selected in wild-type and _Apc_Min/+ mice. In 12-week _Apc_Min/+ mice with tumors, we also analyzed those regions with adenoma and imaged 3−4 randomly selected tumor regions. z Average radial distribution function (RDF) for all nuclei in each group. The solid curve shows the average RDF from all measured nuclei and the shaded area shows the standard error. This definition was used throughout the entire manuscript. The P value for wild-type vs. 6-week _Apc_Min/+, for normal-appearing cells in 6-week _Apc_Min/+ vs. 12-week _Apc_Min/+, and for normal vs. tumor cells (adenoma) from 12-week _Apc_Min/+ mice are P < 10−20, P < 10−9 and P < 10−3, respectively. All P values were determined using Mann−Whitney test.

Fig. 2

Fig. 2. Super-resolution imaging of DNA in _Apc_Min/+ mouse model.

a–d STORM images of DNA folding from normal cells from wild-type, histologically normal-appearing cells from 6-week and 12-week _Apc_Min/+, and tumor cells from 12-week _Apc_Min/+ mice. Scale bars: 10 µm, 2 µm and 500 nm in the original and magnified images, respectively. e–g The statistical analysis of DNA nanodomain size (e), the local density of DNA quantified by Voronoi polygon density (f) and percentage of occupied DNA domains for each nucleus (g) in normal cells from wild-type mice (n = 72 cells), from 6-week (n = 104 cells) and 12-week _Apc_Min/+ mice (n = 81 cells) and tumor cells from 12-week _Apc_Min/+ mice (224 cells). Error bars, mean ± 95% CI. Cell nuclei from three mice for each category were used in the above analysis, as described in the figure caption of Fig. 1. h, i 3D-SIM images of DAPI-stained DNA folding in normal cells from wild-type mice and tumor cells from 12-week _Apc_Min/+ mice. Scale bars: 5 and 2 µm in the original and magnified images, respectively. j The occupancy of DNA defined by the total volume of DNA over the entire 3D volume of each nucleus, calculated from the 3D-SIM images. k The average fluorescence intensity from each of the condensed region of heterochromatin (HC) foci in 3D-SIM images. Cells from wild-type mice (n = 33 cells) and tumor cells from 12-week ApcMin/+ mice (n = 61 cells) were used in the 3D-SIM image analysis. Error bars: mean ± 95% confidence interval (CI). All P values were determined using Mann−Whitney test.

Fig. 3

Fig. 3. Impact of disrupted chromatin structure on transcription and genomic stability.

a, b Average enrichment of H3K9me3 and total H3 from previously described H3K9me3 enriched genomic regions and average H3K4me3 occupancy over transcription start sites (TSSs), and c pie charts that show the genomic distribution of peaks that were enriched by H3K9me3, from normal-appearing intestinal tissue from wild-type mice and age-matched _Apc_Min/+ mice at 6 and 12 weeks, respectively. The genomic distribution in the pie chart consists of satellite repeats, genes, centromere, telomere and other genomic regions (including intergenic, other repeat regions that are unassigned to a type of repeat and non-annotated regions of the genome). d Gene ontology (GO) analysis for overlapping upregulated genes with reduced occupancy of H3K9me3 and increased occupancy of H3K4me3 of 6- and 12-week _Apc_Min/+ mice compared to age-matched wild-type mice. e γH2AX immunofluorescence staining in wild-type mice, _Apc_Min/+ mice at 6 and 12 weeks. Scale bars: 5 µm. f The number of γH2AX foci for each group (n = 50, 31, 28 cells, respectively). g The STORM images of active RNAP II from normal cells from wild-type, histologically normal-appearing cells from 6- and 12-week _Apc_Min/+ and tumor cells from 12-week _Apc_Min/+ mice. Scale bars: 10 µm, 2 µm and 500 nm in the original and magnified images, respectively. h, i Statistical analysis of the active RNAPII cluster size and number of localizations per cluster for each group (n = 301, 333, 174 and 321 cells, respectively). Cell nuclei from three mice were analyzed for each group, similar as described in Fig. 1. Error bars: mean ± 95% CI. All P values were determined using Mann−Whitney test.

Fig. 4

Fig. 4. Functional and structural consequences upon disrupting heterochromatin structure.

a, b Representative two-color STORM images showing the spatial relationship between DNA and active (phosphorylated) RNAP II in control NIH-3T3 cells and those cells with SUV39h1 knockdown (siSUV39h1). Green: phosphorylated RNAP II (labeled by Alexa 647); Red: DNA (labeled by CF568). Scale bars, 4 µm and 500 nm in the original and magnified images, respectively. c Local density map of DNA (quantified by Voronoi polygon density) of control and SUV39h1 knockdown cells. d Average histogram distribution of intra-nuclear local density of control and SUV39h1 knockdown cells. The solid curve shows the average RDF from all measured nuclei and the shaded area shows the standard error. e Western blotting of H3K9me3 and phosphorylated RNAP II with tubulin as an internal reference in control cells and those cells with SUV39h1 knockdown (siSUV39h1). f–h Quantitative analysis of DNA nanodomain size (n = 12 and 24 cells, respectively), active RNAP II (n = 24 and 24 cells, respectively) cluster size and the percentage of DNA that overlaps with active RNAP II (n = 14 and 27 cells, respectively) in control and cells with SUV39h1 knockdown. i γH2AX immunofluorescence and quantification of γH2AX foci numbers in control (n = 43 cells) and cells with SUV39h1 knockdown (n = 29 cells). Error bars: mean ± 95% CI. All P values were determined using Mann−Whitney test. j Cytogenetic analysis of chromosomal aberration in control cells or SUV39h1 knockdown cells. The enlarged regions showed chromosomes with breaks pointed by red arrows. Error bars: mean ± standard error, over 30 cells were counted per group in four randomly assigned groups. The full western blots are provided as Source data.

Fig. 5

Fig. 5. Super-resolution imaging of disrupted heterochromatin structure in prostate neoplasia.

a–e Representative histology and the corresponding super-resolution images of heterochromatin structure (from the blue boxes) from normal epithelial cells of the prostate from wild-type mice, low-grade prostate intraepithelial neoplasia (low-grade PIN), high-grade PIN, carcinoma in situ (CIS) and invasive prostate carcinoma from Hi-Myc mice. Scale bars in the H&E images are 200 and 10 µm, respectively. Scale bars in the STORM images are 10 µm, 2 µm and 500 nm, respectively. f Box-and-whisker plot of the H3K9me3 cluster size (n = 111, 72, 110, 86, 69 cells, respectively). Cell nuclei from three mice for each category were analyzed. g Radial distribution function (RDF) that quantifies H3K9me3-dependent heterochromatin structure, averaged over all nuclei for each group. The shaded area shows the standard error. The P value for wild-type vs. low-grade PIN, for low-grade PIN vs. high-grade PIN, for high-grade PIN vs. CIS and for CIS vs. cancer is P < 10−5, P < 10−8, P < 0.05 and P < 0.05, respectively. All P values were determined using Mann−Whitney test.

Fig. 6

Fig. 6. Super-resolution imaging of disrupted heterochromatin structure in human neoplasia.

a–c Representative histology and corresponding super-resolution images of heterochromatin structure for colorectal neoplastic lesions together with their paired normal tissue located at more than 10 cm away from the tumor. Scale bars: 2 µm. d The radial distribution functions (RDF) for characterization of H3K9me3-dependent heterochromatin structure for the normal-tumor pairs. All P values between two groups are P < 10−20. e The average H3K9me3 cluster size for a total of 19 patient samples (patient information shown in Supplementary Table 1) including five normal tissue from non-neoplastic patients, five adenoma or low-grade dysplasia (LGD), four high-grade dysplasia and five invasive colorectal cancer. Each point is the average of over 100 cell nuclei for each patient. All P values were determined using Mann−Whitney test.

Fig. 7

Fig. 7. Disrupted heterochromatin structure in precursors indistinguishable by conventional pathology.

a–f Representative histology and super-resolution images of heterochromatin structure in normal acinar cells between wild-type mice and _Kras_G12D/+ mice, between transient ADM from wild-type mice treated with cerulein and persistent ADM from _Kras_G12D/+ mice treated with cerulein, PanIN-1 lesions from _Kras_G12D/+ mice and _Kras_G12D/+ mice with cerulein treatment that underwent accelerated tumorigenesis. Scale bars in STORM images are 2 µm and 500 nm in the original and magnified images, respectively. g Statistical analysis of the H3K9me3 cluster size for each group. Acinar cell from wild-type (n = 115 cells) and _Kras_G12D/+ (n = 103 cells), transient ADM (n = 103 cells) and persistent ADM (n = 145 cells), _Kras_G12D/+ (n = 133 cells) and accelerated tumorigenesis (n = 101 cells). Data are presented as violin plots. h, i Representative histology and corresponding super-resolution images of heterochromatin structure between non-advanced adenoma and advanced adenoma without high-grade dysplasia. Note that adenoma and advanced adenoma without HGD are histologically indistinguishable, and advanced adenoma without HGD refers to those large adenoma with a tumor size of more than 1 cm (1.5 cm polyps). Scale bars in STORM images are 10 µm, 2 µm and 500 nm in the original and magnified images, respectively. j The box-and-whisker plots of the cluster size of H3K9me3-dependent heterochromatin (n = 105 and 154 cells, respectively). k The average radial distribution function (RDF) for H3K9me3, where the solid curve shows the average value from all measured nuclei and the shaded area shows the standard error. The P value between non-advanced adenoma and advanced adenoma without HGD is P < 10−15. P values were determined using Mann−Whitney test.

Fig. 8

Fig. 8. Model to depict the molecular-scale heterochromatin structure in carcinogenesis.

As normal cells transform into tumor cells in carcinogenesis, chromatin folding becomes gradually de-compacted and fragmented accompanied with enlarged heterochromatin foci, which increases enhanced formation of transcription factories and genomic instability.

References

    1. Brock MV, Herman JG, Baylin SB. Cancer as a manifestation of aberrant chromatin structure. Cancer J. 2007;13:3–8. -PMC -PubMed
    1. Zink D, Fischer AH, Nickerson JA. Nuclear structure in cancer cells. Nat. Rev. Cancer. 2004;4:677–687. doi: 10.1038/nrc1430. -DOI -PubMed
    1. Reddy KL, Feinberg AP. Higher order chromatin organization in cancer. Semin. Cancer Biol. 2013;23:109–115. doi: 10.1016/j.semcancer.2012.12.001. -DOI -PMC -PubMed
    1. Fischer AH, et al. The cytologic criteria of malignancy. J. Cell Biochem. 2010;110:795–811. doi: 10.1002/jcb.22585. -DOI -PubMed
    1. Janssen A, Colmenares SU, Karpen GH. Heterochromatin: guardian of the genome. Annu. Rev. Cell Dev. Biol. 2018;34:265–288. doi: 10.1146/annurev-cellbio-100617-062653. -DOI -PubMed

Publication types

MeSH terms

Substances

Grants and funding

LinkOut - more resources