Combinatorial microRNA activity is essential for the transition of pluripotent cells from proliferation into dormancy

Abstract

Dormancy is a key feature of stem cell function in adult tissues as well as in embryonic cells in the context of diapause. The establishment of dormancy is an active process that involves extensive transcriptional, epigenetic, and metabolic rewiring. How these processes are coordinated to successfully transition cells to the resting dormant state remains unclear. Here we show that microRNA activity, which is otherwise dispensable for preimplantation development, is essential for the adaptation of early mouse embryos to the dormant state of diapause. In particular, the pluripotent epiblast depends on miRNA activity, the absence of which results in the loss of pluripotent cells. Through the integration of high-sensitivity small RNA expression profiling of individual embryos and protein expression of miRNA targets with public data of protein-protein interactions, we constructed the miRNA-mediated regulatory network of mouse early embryos specific to diapause. We find that individual miRNAs contribute to the combinatorial regulation by the network, and the perturbation of the network compromises embryo survival in diapause. We further identified the nutrient-sensitive transcription factor TFE3 as an upstream regulator of diapause-specific miRNAs, linking cytoplasmic MTOR activity to nuclear miRNA biogenesis. Our results place miRNAs as a critical regulatory layer for the molecular rewiring of early embryos to establish dormancy.

Figure 1.

miRNAs are indispensable for the transition of mouse ESCs and embryos into dormancy. (A) White field images of wild-type and Dgcr8 KO mouse ESCs in normal (proliferative) and dormancy conditions. Cells were induced to enter a diapause-like dormant state via MTOR inhibition (MTORi). INK128, a catalytic MTOR inhibitor, was used. Scale bar, 500 µm. (B) Proliferation curves of wild-type and Dgcr8 KO ESCs in normal (proliferative) and dormancy conditions. Cells were plated at low density on six-well plates and counted on the shown days without splitting. The statistical test is a nonlinear regression (curve fit), in which Dgcr8 KO cells have been compared with the wild type for either normal or MTORi conditions. (C) Analysis of apoptosis levels via staining of Annexin V5, followed by flow cytometry analysis. The statistical test is a one-way ANOVA with multiple testing correction. (*) P-value < 0.05. (D) Schematics of morula aggregation experiments and the subsequent testing of embryo pausing efficiency. EGFP-labeled wild-type or Dgcr8 KO mouse ESCs were aggregated with wild-type morulae. These were cultured until the blastocyst stage and then treated with DMSO or MTORi. The number of expanded embryos with or without ICM was counted every day. Embryos with a blastocoel and unfragmented TE were considered intact. Dgcr8 KO ESCs contributed highly to the ICM (Supplemental Fig. S1). (E) Survival curves of chimeric wild-type or Dgcr8 KO embryos under MTORi-induced pausing conditions. (Left) All intact embryos, (right) embryos with a visible ICM. All wild-type embryos retained the ICM during pausing, whereas most Dgcr8 KO embryos lacked it. The statistical test is a Mantel–Cox test, with wild-type paused embryos as a reference data set. (F) Immunofluorescence staining of normal and MTORi-treated (day 3) blastocysts for the epiblast marker NANOG, the trophectoderm marker CDX2, and the DNA stain DAPI. Right panel shows the number of NANOG⁺ cells in each condition. Nine to 13 embryos were stained in each group. Dgcr8 KO embryos treated with MTORi either lack the ICM or contain NANOG⁻ cells. Scale bar, 20 µm. Statistical test is a one-way ANOVA with multiple testing correction. (**) P-value < 0.01.

Figure 2.

Dgcr8 KO ESCs show an aberrant transcriptome. (A) PCA plot of wild-type and Dgcr8 KO transcriptomes in normal and MTORi (day 2) conditions. (B) Volcano plots showing differentially expressed genes (Padj < 0.05, log₂FC > 1) in wild-type and Dgcr8 KO cells in normal and MTORi conditions. Dgcr8 KO shows more DE genes (738 vs. 403 in wild-type, 153 in common) and an overall higher significance level. (C) Scatter plot showing fold change (MTORi/normal) of each gene in wild-type and Dgcr8 KO cells. Dgcr8 KO cells show a wider log₂FC range. DE genes in each background and common DE genes are highlighted with colors. (D) Gene Ontology analyses of significantly down-regulated or up-regulated genes in wild-type and Dgcr8 KO cells in normal and MTORi conditions. Commonly regulated pathways in both backgrounds are shown together with those that differ based on the genetic background. Top 10 terms are shown; the full list is provided in Supplemental Table S2.

Figure 3.

Up-regulation of a set of miRNAs is associated with dormancy. (A) Schematics of sample preparation for small RNA profiling. In vivo diapaused, in vitro diapaused, and normal E4.5 mouse blastocysts were dissected via laser microdissection to separate polar and mural ends (see Supplemental Fig. S3A). ESCs and TSCs were cultured under standard conditions with or without MTORi. Small RNA expression profiles of single embryo parts and bulk stem cells were generated via low-input small RNA-seq. (B) PCA plot of small RNA-seq data sets. Stem cells and embryos are separated along PC1 (55% variance). The polar embryo, which includes pluripotent cells, clusters closer to ESCs, whereas the mural embryo clusters closer to TSCs (PC2, 21% variance), suggesting small RNA expression profiles reflective of tissue of origin. In vivo diapaused embryos show higher variability than other groups. (C) Heatmap showing miRNA expression changes in diapaused embryos (in vitro and in vivo) compared with normal blastocysts and in paused ESCs compared with normal ESCs. miRNAs were clustered into nine clusters based on their expression levels in the three samples (for corresponding silhouette, see Supplemental Fig. S3E). Clusters 4 and 1 show a concordant trend of up-regulation on average and are the focus of the rest of the study. (D) The top 15 concordantly up-regulated miRNAs and their log₂FC values in each condition.

Figure 4.

The miRNA–target network of the paused pluripotent state. (A) Schematics of the computational analysis to construct the network. From the measured logFC of proteins and miRNAs, scores are assigned to edges to assess concordant (for proteins interactions) or discordant (miRNA–target targets) regulation. Comparison of actual scores to a distribution from random pairings enables the identification of the most extreme edge scores to keep in the final network. (B) Differential expression analysis of normal and paused wild-type ESC proteomes; 457 proteins are significantly differentially expressed (216 up, 241 down) using the criteria log₂FC > 0.5 and Padj < 0.1. (C) miRNA network of concordantly up-regulated miRNAs and their significantly down-regulated target proteins. A stringency cutoff is applied to keep the top 10% of miRNA–target edges and the top 1% of protein–protein interactions. P = 1.3 × 10⁻³⁰ from Fisher's exact test (one-sided) for enrichment in differentially expressed protein. A less stringent network with the first cutoff lowered to the top 30% of miRNA–target edges is in Supplemental Figure S5B, whereas an ESC-only network is provided in Supplemental Figure S6. (D) Gene Ontology analysis of in-network proteins. (E) Immunofluorescence stainings and quantifications of the in-network PML and DDX6 proteins in MTORi-paused embryos. Single-cell quantifications were performed for the ICM and TE separately by manually cropping the images. Between seven and nine embryos were stained in each group. Scale bar, 20 µm. Statistical test is an unpaired t-test. (***) P-value < 0.001; (****) P-value < 0.0001. Supplemental Figure S8 contains additional stainings in ESCs. Scale bar, 20 µm.

Figure 5.

Combinatorial miRNA activity promotes efficient transition to dormancy during in vitro diapause. (A) Schematics of the experiment. Two-cell-stage embryos were microinjected with synthetic miRNA inhibitors (antimiRs) against the miR-200 family and miR-26b-5p or with a control inhibitor (nontargeting). After injection, the embryos were cultured until the blastocyst stage under standard conditions and treated afterward with DMSO or RapaLink-1 to induce in vitro diapause. (B) Survival curves of embryos generated as described above. The number of expanded embryos was counted every day. Embryos with a blastocoel and unfragmented TE were considered intact. Statistical test is a Mantel–Cox test with control inhibitor + MTORi as the reference data set. (n) Number of embryos. (C) Immunofluorescence staining of control-injected or antimiR-injected representative embryos on day 3 of MTORi treatment for the epiblast marker NANOG, the trophectoderm marker CDX2, and the DNA stain DAPI. Right panels show the number of NANOG⁺ cells per embryo in each condition. (n) Number of embryos. Statistical test is an unpaired t-test with Welch's correction. (***) P-value < 0.001. Scale bar, 20 µm. (D) Schematics of miR92 overexpression (OE). Zygotic pronuclei were injected with the linearized miR92 overexpression construct (pLKO.1 background, U6 promoter), empty vector, or injection buffer. The embryos were cultured in standard medium without MTORi until the end of the assay. For scoring, the same procedure as in A is applied. (E) Survival curves embryos miR92 OE or control embryos. (n) Number of embryos in each group. Statistical test is a Mantel–Cox test with empty vector injection as the control data set. (F) Bright field images of miR92 OE or control embryos on day 13 of in vitro culture without MTORi. Scale bar, 100 µm.

Figure 6.

The MTOR–TFE3 axis regulates miRNA biogenesis in dormancy. (A) Schematics of TF mining at candidate miRNA promoters. Transcription start sites of miRNAs were retrieved from the FANTOM5 database and used to scan the [−1500, +500] regions with JASPAR motifs. High-confidence hits were then used to compare the fractions of promoters with a given motif, for promoters of concordant, positive logFC miRNAs against promoters of other measured miRNAs. (B) Enriched TFs at candidate miRNA promoters. TFE3, shown in bold, was chosen for experimental validation. (C) Schematics of the MTOR–TFE3 axis. MTOR phosphorylates TFE3 when active, which results in its sequestration in the cytoplasm. When MTOR is inactive, nonphosphorylated TFE3 instead translocates into the nucleus for regulation of target genes. (D) TFE3 staining in wild-type normal and paused ESCs. Bottom panels show single-cell quantifications of mean fluorescence intensity in the nucleus and cytoplasm. Scale bar, 5 µm. Statistical test is a Kolmogorov–Smirnov test, two-sided. (E) Genome browser view of TFE3 occupancy over the Gm13648 gene, which contains the miR200a, miR200b, and miR429 miRNAs. TFE3 occupancy was mapped via CUT&Tag in wild-type normal and paused ESCs. (F) Comparison of the log₂ ratio of TFE3 levels (maximum scaled CPM) at candidate (n = 14) and control (n = 88) miRNA promoters. Statistical test is a Mann–Whitney U test, one-sided.