TASOR expression in naive embryonic stem cells safeguards their developmental potential

Abstract

The seamless transition through stages of pluripotency relies on a balance between transcription factor networks and epigenetic mechanisms. Here, we reveal the crucial role of the transgene activation suppressor (TASOR), a component of the human silencing hub (HUSH) complex, in maintaining cell viability during the transition from naive to primed pluripotency. TASOR loss in naive pluripotent stem cells (PSCs) triggers replication stress, disrupts H3K9me3 heterochromatin, and impairs silencing of LINE-1 (L1) transposable elements, with more severe effects in primed PSCs. Notably, the survival of Tasor knockout PSCs during this transition can be restored by inhibiting caspase or deleting the mitochondrial antiviral signaling protein (MAVS). This suggests that unscheduled L1 expression activates an innate immune response, leading to cell death specifically in cells exiting naive pluripotency. Our findings highlight the importance of epigenetic programs established in naive pluripotency for normal development.

Keywords: 5mC; CP: Stem cell research; DNA methylation; H3K9me3; HUSH complex; L1; LINE-1; Stem cells; TASOR; heterochromatin; naive pluripotency; primed pluripotency.

Figure 1.. TASOR loss induces cell death during ESC differentiation

(A) Diagram of Tasor mRNA and protein levels during mouse ESC-to-EpiLC transition from the Stem Cell Atlas dabase. (B) Western blot showing the expression of TASOR, P-SMAD2/3, SMAD2/3, and α-tubulin (loading control) of mouse ESCs cultured in 2i/L (naive), and EpiSCs cultured in NBFR (primed). (C) Western blot showing the expression of TASOR and histone H3 (H3; loading control) for wild type (WT), Tasor knockout (KO), and putback (PB) mouse ESCs. (D) Colony formation assay for naive ESC cells transitioned to primed EpiSCs. (E) Heatmap of differentially expressed pluripotency-related genes between mouse ESCs and EpiLCs in WT, Tasor KO, and PB cells. (F) SYTOX Green staining for cell death in mouse ESCs transitioning to the formative state (FS) and cultured for 72 h in AloXR.

Figure 2.. TASOR loss induces cell-cycle arrest, DNA damage, and DNA replication stress

(A) Flow cytometry cell cycle analysis via EdU incorporation and DNA staining of Tasor KO, PB, WT, and overexpression (OE) mouse ESCs. Bars represent standard deviation (SD) among biological replicates. Kruskal-Wallis one-way ANOVA with Tukey’s HSD from 8 biological replicates, *adjusted p < 0.05. (B) Immunofluorescence staining for the mitosis marker phospho H3 (serine 10) for Tasor KO and PB mouse ESCs. Scale bar: 100 μm. (C) Immunofluorescence staining of WT and Tasor KO mouse ESCs for phosphoserine 139 of histone H2AX (γH2AX). Scale bar: 100 μm. (D) Western blot for TASOR 8 h after addition of 2 μM 5-ph-IAA. (E) Z-slice confocal immunofluorescence image for γH2AX in control DMSO or 5-Ph-IAA-treated cells. The white arrow indicates colocalization between FLAG and γH2AX. (F) Mean segmented nuclear intensity normalized to DAPI of FLAG Alexa Fluor 488. Unpaired t test from 2 biological replicates. (G) Mean segmented nuclear intensity normalized to DAPI of γH2AX Alexa Fluor 555. Unpaired t test from 2 biological replicates. (H) Diagram of DNA fiber assay with a representative image of a chromatin fiber. (I) Replication fork symmetry quantification of chromatin fibers in DMSO or 5-ph-IAA treated cells (n = 2). Kruskal-Wallis one-way ANOVA with Tukey’s HSD from 2 biological replicates. (J) Replication fork length quantification of chromatin fibers in DMSO or 5-ph-IAA-treated cells (n = 2). Kruskal-Wallis one-way ANOVA with Tukey’s HSD from 2 biological replicates. (K) Replication fork speed quantification of chromatin fibers in DMSO or 5-ph-IAA-treated cells (n = 2). Kruskal-Wallis one-way ANOVA with Tukey’s HSD from 2 biological replicates.

Figure 3.. L1 RNA abundance and half-life is increased upon TASOR loss

(A) MA plots for repeats showing log2 fold change of Tasor KO over WT mouse ESCs. (B) MA plots for repeats showing log2 fold change of Tasor KO over WT mouse EpiLCs. (C) Normalized average counts for LINE-1 (L1) subfamilies in mouse ESCs and EpiLCs for WT, Tasor KO, and PB. (D) Heatmap for CUT&Tag of TASOR-33Flag and H3K9me3 at L1 family members. (E) Time-course normalized RNA fold change via qPCR for L1 after Auxin treatment. Kruskal-Wallis one-way ANOVA with Tukey’s HSD from 2 or more biological replicates. Bars represent standard deviation (SD) among biological replicates. (F) Experimental diagram for measuring RNA half-life after actinomycin D (ActD) treatment. (G) Relative mRNA abundance after ActD treatment for TBP measured by RT-qPCR. Two-way ANOVA with Geisser-Greenhouse correction. Bars represent standard deviation (SD) among biological replicates. (H) Relative mRNA abundance after ActD treatment for L1 measured by RT-qPCR. Two-way ANOVA with Geisser-Greenhouse correction. Bars represent standard deviation (SD) among biological replicates.

Figure 4.. Cell death induced by TASOR loss is partially mediated by a MAVS innate immune response

(A) Western blot analysis for DNMT1, TASOR, and MAVS in Dnmt33KO, Tasor KO, and Tasor/Mavs double KO (dKO) mouse ESCs. (B) Representative epifluorescence images for cell death using SYTOX Green staining (bottom) and flow cytometry analysis (top) after 48 h of AloXR-mediated formative stem cell conversion. Percentages are shown as mean +/− standard deviation (SD) (C) RT-qPCR analysis for L1 abundance in WT, Tasor KO, Dnmt33KO, Tasor/Mavs dKO mouse ESCs, and Tasor KO + TASOR^mAID mouse ESCs with and without 2 μM 5ph-IAA treatment. Kruskal-Wallis one-way ANOVA with Tukey’s HSD from 2 or more biological replicates. Bars represent standard deviation (SD) among biological replicates.