Super-resolution imaging reveals nucleolar encapsulation by single-stranded DNA

Abstract

In eukaryotic cell nuclei, specific sets of proteins gather in nuclear bodies and facilitate distinct genomic processes. The nucleolus, a nuclear body, functions as a factory for ribosome biogenesis by accumulating constitutive proteins, such as RNA polymerase I and nucleophosmin 1 (NPM1). Although in vitro assays have suggested the importance of liquid-liquid phase separation (LLPS) of constitutive proteins in nucleolar formation, how the nucleolus is structurally maintained with the intranuclear architecture remains unknown. This study revealed that the nucleolus is encapsulated by a single-stranded (ss)DNA-based molecular complex inside the cell nucleus. Super-resolution lattice-structured illumination microscopy (lattice-SIM) showed that there was a high abundance of ssDNA beyond the 'outer shell' of the nucleolus. Nucleolar disruption and the release of NPM1 were caused by in situ digestion of ssDNA, suggesting that ssDNA has a structural role in nucleolar encapsulation. Furthermore, we identified that ssDNA forms a molecular complex with histone H1 for nucleolar encapsulation. Thus, this study illustrates how an ssDNA-based molecular complex upholds the structural integrity of nuclear bodies to coordinate genomic processes such as gene transcription and replication.

Keywords: In situ imaging; DNA–protein interaction; Nucleolus; Nucleus; Single-stranded DNA; Super-resolution imaging.

Fig. 1.

Verification of in situ ssDNA imaging. (A) Fluorescence images of DAPI (intercalating into dsDNA) and SYBR Green II (binding to both ssDNA and RNA), without and with RNase A treatment. (B,C) Histograms of the number of speckles (B) and hollows (C) per nucleus in the SYBR Green II images. n=162 cells [RNase(−)], 147 cells [RNase (+), 1 h], 157 cells [RNase (+), 24 h]. (D) The mean±s.d. amount of dsDNA and ssDNA in cells quantified with the Qubit^TM fluorometer. n=5 biological replicates. *P<0.05; **P<0.01 (two-sided Mann–Whitney U test). n.d., not detected. (E) Fluorescence images without and with DNase I, using semi-intact nuclei. Of note, the exposure time for SYBR Green II was adjusted for heated sample with increased amount of ssDNA, and thereby the signals are relatively weak in the sample without heat denaturation. (F) Fluorescence images without and with heat denaturation. Images in E and F are representative of three repeats.

Fig. 2.

In situ super-resolution imaging of ssDNA. (A) Lattice-SIM images of dsDNA and ssDNA. Hollow-shaped structures are marked with asterisks in the merged image. (B) Schematics of the organization of the nucleolus, comprising three layers: the fibrillar center (FC) containing RNA polymerase I (RNAPI), the dense fibrillar compartment (DFC) containing fibrillarin (FIB), and the granular compartment (GC) containing nucleophosmin 1 (NPM1). (C) Co-staining of dsDNA, ssDNA and nucleolar markers. (D) Lattice-SIM images of ssDNA and NPM1. Images in A, C and D are representative of three repeats. (E) Intensity profile of ssDNA and NPM1. The centroid of nucleolus was determined based on 3D lattice-SIM images of NPM1, and the intensity profiles of ssDNA and NPM1 were measured along the radial axis. The radial distance d from the centroid was normalized based on the distance between the centroid and the nucleolar outline. n=47 nucleoli from 7 cells. The intensity profile is the moving average with the s.d. indicated as the shaded area. a.u., arbitrary units.

Fig. 3.

In situ digestion of ssDNA leads to nucleolar disruption. (A) Lattice-SIM images of dsDNA, ssDNA and NPM1 for semi-intact nuclei without and with S1 nuclease treatment, after treatment with 0.5% Triton X-100 in CSK buffer for 2 min on ice. For the ‘S1 buffer’ sample, semi-intact nuclei were incubated in S1 buffer for 3 h at 37°C. (B,C) Beeswarm plots of mean intensity of NPM1 in nucleoli (C) and nucleoar volume (C). n=252 nucleoli from 30 nuclei (S1 buffer), 132 nucleoli from 24 nuclei (S1 nuclease, 1 h), and 84 nucleoli from 16 nuclei (S1 nuclease, 3 h). ****P<0.0001 (two-sided Mann–Whitney U test). a.u., arbitrary units.

Fig. 4.

ssDNA and histone H1 form a molecular complex for nuclear encapsulation. (A) Lattice-SIM images of dsDNA, ssDNA and histone H1. (B) Pearson correlation coefficient between dsDNA and histone H1, and ssDNA and histone H1. n=6 regions with a 5 µm×5 µm range around each nucleolus, from six independent nuclei. **P<0.01 (two-sided Mann–Whitney U test). (C) Fluorescence images of dsDNA, histone H1 and EdU. 50 nM EdU was incorporated prior to the fixation and fluorescently detected by Click chemistry. The upper and bottom images were obtained using 20× and 100× objectives, respectively. (D) Beeswarm plots of mean intensity and total intensity of histone H1 for EdU-negative and EdU-positive cells. n=186 nuclei (EdU-negative) and 232 nuclei (EdU-positive). ****P<0.0001 (two-sided Mann–Whitney U test). (E) Lattice-SIM images of dsDNA, ssDNA and histone H1 for semi-intact nuclei without and with S1 nuclease treatment. The ‘S1 buffer’ sample was incubated in S1 buffer for 3 h at 37°C. (F) Beeswarm plots of the number of hollow patterns of histone H1 per nucleus without and with S1 nuclease treatment. n=32 nuclei (S1 buffer), 28 nuclei (S1 nuclease, 1 h), and 34 nuclei for (S1 nuclease, 3 h). ****P<0.0001 (two-sided Mann–Whitney U test). (G) Structural schematics of nucleolus encapsulation by a molecular complex of ssDNA and histone H1. a.u., arbitrary units.