Zscan4 mediates ubiquitination and degradation of the corepressor complex to promote chromatin accessibility in 2C-like cells

Abstract

Zygotic genome activation occurs in two-cell (2C) embryos, and a 2C-like state is also activated in sporadic (~1%) naïve embryonic stem cells in mice. Elevated chromatin accessibility is critical for the 2C-like state to occur, yet the underlying molecular mechanisms remain elusive. Zscan4 exhibits burst expression in 2C embryos and 2C-like cells. Here, we show that Zscan4 mediates chromatin remodeling to promote the chromatin accessibility for achieving the 2C-like state. Through coimmunoprecipitation/mass spectrometry, we identified that Zscan4 interacts with the corepressors Kap1/Trim28, Lsd1, and Hdac1, also with H3K9me3 modifiers Suv39h1/2, to transiently form a repressive chromatin complex. Then, Zscan4 mediates the degradation of these chromatin repressors by recruiting Trim25 as an E3 ligase, enabling the ubiquitination of Lsd1, Hdac1, and Suv39h1/2. Degradation of the chromatin repressors promotes the chromatin accessibility for activation of the 2C-like state. These findings reveal the molecular insights into the roles of Zscan4 in promoting full activation of the 2C-like state.

Keywords: 2C like state; Zscan4; chromatin accessibility; ubiquitination; zygotic genome activation.

Fig. 1.

Dynamics of Zscan4 and heterochromatin remodeling in 2C mouse embryos and mESCs. (A) Representative images of Zscan4 protein expression by fluorescence microscopy of 2-cell mouse embryos using anti-Zscan4 (red) antibody. DAPI stained nucleus in blue. (Scale bar, 10 μm.) (B) H3K9me3 immunofluorescence (red) images of heterochromatin in zygotes and 2C embryos. (Scale bar, 10 μm.) (C) Quantification of the relative levels of Zscan4 and H3K9me3 as shown in B. Number of 2C embryos analyzed for Zscan4: n ≥ 7; mean ± SEM. *P < 0.05 and ***P < 0.001. (D) Zscan4 dynamics in heterochromatin remodeling. Images captured from live cell imaging microscopy video of Zscan4 promoter-driven Tomato Red ESCs and Hoechst 33342 revealing distinct dynamics of heterochromatin foci. (E) Quantification of the frequency of cells with Zscan4 reporter positive only (indicative of Zscan4 promoter activity), Zscan4 protein positive only (indicated by Zscan4 antibody IF), or overlap of each other. n = 49. (F) Immunofluorescence microscopy of Zscan4 (Green) with the Zscan4 promoter-driven Tomato Red (Red). DAPI was used to stain the nucleus (blue). Heterochromatin foci were reduced or disappeared in Zscan4⁺ cells. (G) Quantification of DAPI-dense heterochromatin foci in Tomato Red (TR)⁺/Zscan4⁺ and Tomato Red (TR)⁻/Zscan4⁻ cells (Zscan4⁺, n = 26; Zscan4⁻, n = 29). (H) Immunofluorescence of H3K9me3 (green) in Zscan4 promoter-driven Tomato Red mESCs. (I and J) Quantification of the number of H3K9me3 foci per cell (n = 62) (I) and relative average area of H3K9me3 foci (n = 50) between Zscan4⁻ and Zscan4⁺ cells (J), as shown in H. (K) Immunofluorescence of HP1α (green) in Zscan4-positive (red) ESCs. (Scale bar, 10 μm.) (L) Quantification of the number of HP1α foci per cell between Zscan4⁻ and Zscan4⁺ cells, as shown in K. ***P < 0.001.

Fig. 2.

Identification of Zscan4 interactome and functional domains in interacting with corepressors. (A) Silver staining gel image of Zscan4-interacting proteins that were immunoprecipitated from Zscan4 OE mESCs treated with MG-132 and identified by mass spectrometry. Putative Zscan4 proteins are indicated with an asterisk. (B) Western blot of endogenous Zscan4 from IP sample. (C) List of representative proteins identified in the Zscan4 Co-IP/MS. Scores of the identified peptide are indicated. (D) IP-western blot validation of Co-IP/MASS data for selected proteins. Shown are interacting proteins with Zscan4 such as Kap1, Lsd1, Hdac1, and Trim25. (E) Schematic drawings of full-length Zscan4, SCAN domain depleted (ΔSCAN) and zinc finger motifs depleted (ΔZNF) truncated proteins. (F) Western blotting of HA-Hdac1, HA-Lsd1, HA-Kap1, and HA-Trim25 proteins following exogenously expressed Flag-Zscan4 (full length), ΔSCAN or ΔZNF and IP (using an anti-Flag antibody) in HEK293T cells. (G) Protein–protein interaction docking of the predicted structure of mouse Zscan4c (red) by AlphaFold with the crystal structure of human Hdac1 (magenta), Lsd1 (blue), Kap1-Ring (sand), and Trim25-Ring (green), respectively. (Upper) Zscan4c-ZF interacts with the N-terminal SWIRM domain of Lsd1 and may help to recruit Lsd1 to the E3 ligase center of Trim25. (Bottom) The ring domain of Kap1, ring domain of E3 ligase Trim25, and full-length Hdac1 interact with adjacent surfaces of the Zinc finger domain of Zscan4c, suggesting that Zscan4c may facilitate recruitment of Hdac1 to the E3 ligase center of Trim25.

Fig. 3.

Proteasome degradation of Lsd1 and Hdac1 mediated by Zscan4. (A) Immunofluorescence microscopy showing Lsd1 (green) expression in Zscan4 promoter-driven Tomato Red (Zscan4⁺) mESCs. DNA was stained with DAPI. (Right) Quantitative analysis of the relative fluorescence intensity of Lsd1 in Zscan4⁺ and Zscan4⁻ cells (Zscan4⁺, n = 49; Zscan4⁻, n = 52). (B) Immunofluorescence microscopy showing Hdac1 (green) expression in Zscan4 promoter-driven Tomato Red (Zscan4⁺) mESCs. (Right) Quantitative analysis of the relative fluorescence intensity of Hdac1 in Zscan4⁺ and Zscan4⁻ cells (Zscan4⁺, n = 64; Zscan4⁻, n = 64). (C) Western blots of Zscan4, Lsd1, Hdac1, Kap1, H3ac, and H3K9ac levels in Zscan4⁺ cells compared with Zscan4⁻ cells. (Right) mRNA expression level (FPKM) of Lsd1 and Hdac1 between Zscan4⁺ and Zscan4⁻ cells (44). (D) Immunofluorescence of Lsd1 in Zscan4 promoter-driven tomato Red mESCs between DMSO (Control) and MG-132 (10 μM) treatment for 12 h. Arrows indicate Lsd1 protein levels in Zscan4⁺ cells. (Scale bar, 10 μm.) (E) Quantification of the data in D. (F) Quantitative analysis of the relative fluorescence intensity of Lsd1 in Zscan4⁺ cells. (G) Immunofluorescence of Hdac1 in Zscan4 promoter-driven tomato Red mESCs between DMSO (Control) and MG-132 (10 μM) treatment for 12 h. Arrows indicate Hdac1 protein levels in Zscan4+ cells. (H) Quantification of the data in G. The total number of Zscan4+ cells counted between DMSO and MG132 treatments are indicated, and the results are shown as percentages. When the fluorescence intensity in Zscan4+ cells is similar to or stronger than neighbor Zscan4− cells, those Zscan4+ are referred to Hdac1-high or Lsd1-high groups, otherwise Hdac1-medium or Lsd1-low groups (E and H). (I) Quantitative analysis of the relative fluorescence intensity of Hdac1 in Zscan4+ cells. (J) Immunofluorescence of Lsd1 and Hdac1 with Flag (fused with Zscan4) after Zscan4 transient transfection. Arrows indicate Lsd1 and Hdac1 lower levels in Flag-Zscan4 highly expressed cells. (Scale bar, 10 μm.) (K) Quantitative estimate of the relative fluorescence intensity of Lsd1 (n = 47) and Hdac1 (n = 35) between Zscan4 highly OE and control (CON). (L) Western blots of Suv39h1, Lsd1, Hdac1, Kap1, Flag, and Zscan4 in Zscan4 stable OE compared with control (CON) mESCs. *P < 0.05, **P < 0.01, and ***P < 0.001.

Fig. 4.

Zscan4 mediates polyubiquitination-induced degradation of the interacting proteins via Trim25. (A) Ubiquitination of Hdac1, Lsd1, Suv39h1, or Suv39h2, but not of Setdb1 and G9a following OE of Zscan4 in HEK293T cells. n ≥ 3 repeats. (B) Schematic representation of Trim25 knockout (KO) using CRISPR/Cas9 and Sanger sequencing of WT as well as Trim25-KO cell lines. The gRNA sequence locates in Exon 1, and Trim25-KO clone #22 displays 13 bp sequence deletion in both alleles, which results in the disruption of E3 ligase RING domain and frameshift. (C) Western blotting confirmation of Trim25-KO mESC lines with Trim25 antibody. (D) Immunofluorescence microscopy of Lsd1 and Hdac1 in WT and Trim25-KO mESCs. Arrows indicate Hdac1 and Lsd1 immunofluorescence in Zscan4⁺ cells. Since both Lsd1 and Zscan4 antibodies are rabbit polyclonal, Zscan4c promoter-driven AcGFP1 expression reporter plasmid was first transfected into WT and Trim25-KO mESCs to indicate Zscan4⁺ cells. (E) Quantitative estimate of fluorescence intensity of Lsd1 and Hdac1 in Zscan4⁺ cells between WT and Trim25-KO mESCs. The total number of Zscan4⁺ cells counted between WT and Trim25-KO mESCs are indicated, and the results are shown as percentages. (F) Quantitative analysis of the relative fluorescence intensity of Lsd1 and Hdac1 in Zscan4⁺ cells between WT and Trim25-KO mESCs. (G–I) In vitro ubiquitination assay using purified proteins of Trim25 to ubiquitinate Hdac1 in the presence of Zscan4 (G) or absence of Zscan4 (H) and Lsd1 in the absence of Zscan4 (I). Each experiment contained two biological repeats. *P < 0.05, **P < 0.01, and ***P < 0.001.

Fig. 5.

Trim25 releases Kap1 from binding to Zscan4. (A) IP/western blot of Kap1 interactions with Hdac1 and Lsd1 in HEK293T cells. (B) Ubiquitination of Hdac1 and Lsd1 following OE of Zscan4 and Kap1 in HEK293T cells. (C) In vitro ubiquitination assay using purified proteins demonstrating that Trim25 but not Kap1 can ubiquitinate Hdac1 and Lsd1. The presence of Kap1 decreases Trim25 ubiquitination activity of Hdac1 or Lsd1. (D) IP/western blot showing that Kap1 does not interact with Trim25 in HEK293T cells. (E) IP/western blot of HA (Zscan4) interaction with Flag (Kap1) in the presence or absence of Trim25. (F) Western blotting of Kap1 and Flag in Trim25 OE in control and Zscan4 OE mESC lines. Each experiment contained two biological repeats.

Fig. 6.

Zscan4 promotes chromatin accessibility and up-regulates 2C-like genes and retrotransposons in ESCs. (A) Heatmap showing expression of representative 2C-like genes according to single-cell RNA-seq in preimplantation embryos data (53). The genes including the Zscan4 family up-regulated are marked in red on the Right. (B) Scatter diagrams of TEs expression after Zscan4 OE. (C) The average profile showing the ATAC-seq signals in Zscan4 OE mESCs compared with WT control (CON). (D) The average profile and heatmap of the ATAC-seq enrichment around accessible promoters at 2C-like genes up-regulated after Zscan4 OE from A. (E) The average profile and heatmap of the ATAC-seq enrichment at the start and end sites of MERVL-int up-regulated in Zscan4 OE mESCs. (F) Z-score of 2C-like gene expression (TPM) and ATAC-seq enrichment after Zscan4 OE. (G and H) Plot of Zscan4, Lsd1, Hdac1, Kap1, and H3K9me3 binding profile and heatmap at TSS and TES of 2C-like genes (G) and retrotransposons (H) up-regulated in Zscan4 OE mESCs. The ChIP-seq signals are calculated as the ratio of normalized reads relative to input. TES, transcription end site. Data from GSE140621 (29) (Zscan4 ChIP-seq data from Zscan4⁺ cells), GSE18515 (55) (Lsd1 ChIP-seq), GSE183292 (56) (Hdac1 ChIP-seq), GSE70799 (57) (Kap1 ChIP-seq), or GSE98256 (58) (H3K9me3 ChIP-seq).

Fig. 7.

Proposed simplified model for Zscan4 function in chromatin remodeling. During the accumulation of Zscan4 protein, Zscan4 interacts with Kap1, Hdac1, and Lsd1 and with heterochromatin modifiers such as Suv39h1 and Suv39h2, transiently forming a large repressive chromatin complex (specific foci can be observed under the microscopy). The E3 ligase Trim25 is then recruited to Zscan4 through the ZNF domain, to release Kap1 from Zscan4, allowing polyubiquitination and degradation of the corepressors such as Hdac1 and Lsd1 mediated by Trim25. Decompaction of repressive chromatin promotes histone acetylation and increases chromatin accessibility for promoting the 2C-like state.