Journal: bioRxiv

Article Title: Identification of cellular double-stranded RNAs in mammalian embryonic stem cells

doi: 10.64898/2026.04.13.718158

Figure Lengend Snippet: ( A ) Expression of the pluripotency markers Pou5f1 and Nanog and differentiation markers Acta2 and Tagln in ESCs and differentiated cells was measure by RT-qPCR, comparing mock and dsRNA-stimulated cells. Actb was used as a normaliser. Data represent the average of three biological replicates ± SD relative to untreated ESC. One-way ANOVA was used to calculate significant differences between dsRNA- treated samples within each cell line, followed by a Tukey HSD. *p-val ≤ 0.05, **p-val ≤ 0.01, ***p-val ≤ 0.001. ( B ) Expression of Ifnb1 and interferon stimulated genes Oas1 and Stat1 in ESCs and differentiated cells upon dsRNA or mock stimulation by RT-qPCR analysis. Actb was used as a normaliser. Data represent the average of three biological replicates ± SD relative to untreated ESC. One-way ANOVA was used to calculate significant differences between dsRNA-treated cell lines, followed by a Tukey HSD. *p-val ≤ 0.05, **p-val ≤ 0.01, ***p-val ≤ 0.001. ( C ) Western blot analysis of phosphorylated IRF3 levels in ESCs and differentiated cells. GAPDH serves as a loading control ( D ) Western blot analysis of phosphorylated PKR in ESCs and differentiated cells. Tubulin serves as a loading control ( E ) rRNA electrophoresis from ESCs and differentiated cells, mock and dsRNA-stimulated. (+) represents lower dsRNA and (++) higher dsRNA stimulation.

Article Snippet: Cells were transfected with the dsRNA analogue poly(I:C) (Invivogen, tlrl-pic) using Lipofectamine 2000 (ThermoFisher, 11668027).

Techniques: Expressing, Quantitative RT-PCR, Western Blot, Control, Electrophoresis

Journal: bioRxiv

Article Title: Identification of cellular double-stranded RNAs in mammalian embryonic stem cells

doi: 10.64898/2026.04.13.718158

Figure Lengend Snippet: ( A ) Flow cytometry analysis against dsRNA using the dsRNA-specific antibody J2 in ESC vs differentiated (diff) cells in the presence or absence of RNase III. RNase III treatment tests specificity of J2 antibody signal. Data represent the average of three biological replicates ± SD. One-way ANOVA was used to calculate significant differences amongst comparisons, followed by a Tukey HSD. *p-val ≤ 0.05, **p-val ≤ 0.01, ***p-val ≤ 0.001. ( B ) Representative examples of normalised Flow cytometry intensity in ESCs and differentiated cells after mock and RNaseIII treatment. For all replicates and gating strategy see Supplementary figure 2a ( C ) Immunofluorescence for dsRNA. J2 antibody was used to detect dsRNA, and DAPI to stain the nucleus. Scalebar is 10 µm ( D ) Setup of dsRNA-immunoprecipitation using the J2 antibody. DsRNAs bound to J2 were either mock treated, treated with RNaseI or RNaseIII prior to purification and followed high-throughput RNA sequencing. ( E ) Bioanalyzer electropherogram of purified RNAs J2-IP using the different treatments shown in ( D ). ( F ) Detailed view of Bioanalyzer electropherogram showing accumulation of shorter fragments upon treatment with RNase I and III.

Article Snippet: Cells were transfected with the dsRNA analogue poly(I:C) (Invivogen, tlrl-pic) using Lipofectamine 2000 (ThermoFisher, 11668027).

Techniques: Flow Cytometry, Immunofluorescence, Staining, Immunoprecipitation, Purification, High Throughput Screening Assay, RNA Sequencing

Journal: bioRxiv

Article Title: Identification of cellular double-stranded RNAs in mammalian embryonic stem cells

doi: 10.64898/2026.04.13.718158

Figure Lengend Snippet: ( A ) Normalised counts of the four major classes of transposable elements compared to input controls. One-way ANOVA was used to calculate significant differences between input and immunoprecipitated samples, followed by a Tukey HSD. *p-val ≤ 0.05, **p-val ≤ 0.01, ***p-val ≤ 0.001. ( B ) Violin plots showing differential enrichment (log₂ fold change) of TE subfamilies within selected families for dsRNA IP (mock) versus input. Points represent individual subfamilies, coloured by differential enrichment (red: upregulated, blue: downregulated, grey: not significant; log₂FC > 1, adjusted p < 0.05). ( C ) Violin plots showing differential enrichment (log₂ fold change) of TE subfamilies within selected families after RNase I treatment versus dsRNA (mock). Points represent individual subfamilies, coloured by differential enrichment (red: upregulated, blue: downregulated, grey: not significant; log₂FC > 1, adjusted p < 0.05). ( D ) Levels of A-to-I editing in retrotransposons for input and dsRNA-IPs. One-way ANOVA was used to calculate significant differences between input and immunoprecipitated samples, followed by a Tukey HSD. *p-val ≤ 0.05, **p-val ≤ 0.01, ***p-val ≤ 0.001. ( E ) Levels of A-to-I editing by TE class. Statistical analysis as in ( D ). ( F ) Percentage of sense reads mapping to TEs with a corresponding antisense match, for input and dsRNA IPs. One-way ANOVA was used to calculate significant differences between input and immunoprecipitated samples, followed by a Tukey HSD. *p-val ≤ 0.05, **p-val ≤ 0.01, ***p- val ≤ 0.001. ( G ) Sense/antisense overlap by TE class. Statistical analysis as in ( F ). ( H ) Examples of specific TE locus. The combined normalised reads from the biological replicates mapped are shown, for inputs and dsRNA IPs. Above black line represents sense (or ‘plus’), below represents antisense reads (or ‘minus’). Chromosomal location of each locus is represented at the top

Article Snippet: Cells were transfected with the dsRNA analogue poly(I:C) (Invivogen, tlrl-pic) using Lipofectamine 2000 (ThermoFisher, 11668027).

Techniques: Immunoprecipitation

Journal: Nucleic Acids Research

Article Title: Control of retrotransposon-driven activation of the interferon response by the double-stranded RNA binding protein DGCR8

doi: 10.1093/nar/gkag190

Figure Lengend Snippet: Basal IFN activation of DGCR8 KO cells relies on the dsRNA-sensing pathway. ( A ) RT-qPCR analysis of the ISGs IFIT3, OAS3 , and IFIT2 in WT and two DGCR8 KO PA-1 clones treated with BX795, ruxolitinib (RUXO), or vehicle control (DMSO). GAPDH serves as a normaliser. Data are presented relative to the average expression of each gene in the KO2 DMSO condition. ( B ) RT-qPCR analysis of the ISGs DDX58, IFIH1, IFIT2 , and IFIT3 in WT and two independent DGCR8 KO PA-1 clones transfected with poly(I:C). Expression is normalised as 2^-ΔCt, serving GAPDH as normaliser. ( C ) Expression levels of the IFNB1 mRNA in WT and two DGCR8 KO PA-1 clones after G3-YDNA transfection by RT-qPCR. BV-2 cells were used as a positive control for DNA stimulation. GAPDH was used as the normaliser. Expression is represented relative to the average expression of the WT + G3-YDNA condition. ( D, E ) MAVS ( D ) and IFIH1 ( E ) KO clones in DGCR8 KO HEK293T cells. Western blot analysis was used to confirm complete KO of MAVS protein ( D , left). RT-qPCR analysis of the ISGs IFIT2 and MX1 in DGCR8 KO and double KO (dKO) cell lines. Expression is represented relative to DGCR8 KO, and GAPDH is used as normaliser. Bars represent the average of three biological replicates. Error bars plot mean ± SD. * P < 0.05, ** P < 0.01, *** P < 0.001, **** P < 0.0001. P -values are determined by one-way ANOVA followed by Dunnett multiple comparison post-hoc test.

Article Snippet: Cells were transfected at 70% confluency, in six-well plates, and transfected with 5 μl of Invitrogen Lipofectamine 2000 Transfection Reagent and 1 μg of Y-shaped-DNA cGAS agonist (G3-YSD) (#tlrl-ydna, Invivogen) or dsRNA analogue polyinosinic:polycytidylic acid (poly(I:C)) (HMW #tlrl-pic, Invivogen).

Techniques: Activation Assay, Quantitative RT-PCR, Clone Assay, Control, Expressing, Transfection, Positive Control, Western Blot, Comparison

Journal: Nucleic Acids Research

Article Title: Control of retrotransposon-driven activation of the interferon response by the double-stranded RNA binding protein DGCR8

doi: 10.1093/nar/gkag190

Figure Lengend Snippet: 3′ UTR-localised transposable elements form dsRNA in DGCR8 KO cells. ( A ) Read distribution of the J2-based dsRNA immunoprecipitation (dsRNA-seq) in PA-1 WT and two independent DGCR8 KO clones. Reads mapping to > 1 category are counted in each. ( B, C ) Differential TE loci expression in dsRNA-seq IPs from DGCR8 KO1 ( B ) and DGCR8 KO2 ( C ) cells versus WT. Significantly downregulated (log2FC < -1 and p-adj < 0.05) and upregulated (log2FC > 1 and p-adj < 0.05) TE loci are shown for each clone. ( D ) Fraction of upregulated TE loci (log2FC > 1 and p-adj < 0.05) in DGCR8 KO1 and KO2 dsRNA-Seq on each TE family. ( E ) Representation of TE classification; ‘self-expressed’ TEs are those annotated in intergenic regions or introns of a non-expressed protein-coding gene. Gene-dependent TEs are those inserted within exons, 5′ UTRs, 3′ UTRs, extended 3′ UTRs, introns of an expressed gene, or expressed on antisense orientation. ( F ) Fraction of self-expressed versus gene-dependent dsRNA-forming TE loci in both DGCR8 KO clones. ( G ) Fraction of intragenic versus intergenic dsRNA-forming TE loci in both DGCR8 KO clones. ( H ) Genomic distribution of dsRNA-forming TE loci in WT, DGCR8 KO1 and KO2 cells, versus those enriched (up) in KO1 or KO2 versus WT (log2FC > 1 and p-adj < 0.05) and depleted (down) in KO1 or KO2 versus WT (log2FC > -1 and p-adj < 0.05). Proportions represent the average of three biological replicates.

Article Snippet: Cells were transfected at 70% confluency, in six-well plates, and transfected with 5 μl of Invitrogen Lipofectamine 2000 Transfection Reagent and 1 μg of Y-shaped-DNA cGAS agonist (G3-YSD) (#tlrl-ydna, Invivogen) or dsRNA analogue polyinosinic:polycytidylic acid (poly(I:C)) (HMW #tlrl-pic, Invivogen).

Techniques: Immunoprecipitation, Clone Assay, Expressing

Journal: Nucleic Acids Research

Article Title: Control of retrotransposon-driven activation of the interferon response by the double-stranded RNA binding protein DGCR8

doi: 10.1093/nar/gkag190

Figure Lengend Snippet: DGCR8 binds SINE-rich mRNAs. ( A ) Absolute number of TEs per gene in: mRNAs bound by DGCR8 with at least one accumulated dsRNA-TE (log2FC > 1 and p-adj < 0.05) in either of the DGCR8 KO clones, mRNAs bound by DGCR8 without any accumulated dsRNA-TEs (yellow) and unbound mRNAs, with binding defined by Macias et al. (see Materials and methods). ( B ) Density of TEs per gene in the three gene categories defined in ( A ). ( C ) Average number of SINE, LINE and LTR elements per 3′ UTR of the three gene categories defined in ( A ). ( D ) Minimum Free Energy (MFE) violin plots for the three gene categories defined in ( A ). ( E ) Aggregate plot of mean pA ± RNA 3′ end-seq from HeLa cells (Gockert et al. 2022). Signal was plotted over 300 nt TSS-upstream to 300 nt TES-downstream regions of annotated SINE elements in 3′ UTRs of mRNAs with logFC > 1 & padj < 0.05 in DGCR8 KO2/WT dsRNA-Seq ( n = 1044). The transcriptional unit (TU) body area (TSS to TES) was normalised to 300 nt for each SINE. Mean values and 90% confidence intervals (lighter shade around the curves) of log2-transformed coverage data are displayed in the aggregate plot with binning of 5 and a \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{upgreek} \usepackage{mathrsfs} \setlength{\oddsidemargin}{-69pt} \begin{document} ${{10}^{ - 1}}$\end{document} multiplication factor. ( F ) Aggregate plot of mean pA ± RNA 3′ end-seq from HeLa cells, as in ( E ). Signal was plotted over 25 nt upstream to 25 nt downstream regions of expressed mature miRNA ( n = 63). Mean values and 90% confidence intervals (lighter shade around the curves) of log2-transformed coverage data are displayed in the aggregate plot with binning of 5 and a \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{upgreek} \usepackage{mathrsfs} \setlength{\oddsidemargin}{-69pt} \begin{document} ${{10}^{ - 1}}$\end{document} multiplication factor. ( G ) Immunoprecipitation of overexpressed Flag-DGCR8 in HeLa cells and co-immunoprecipitated DROSHA, ZCCHC8, and DHX9 levels by western blot. Benzonase (benz) treatment serves to test RNA-dependent interactions. ( H ) Proximity labelling assay of WT and mTurboID-DGCR8 mESCs. Biotinylated proteins were enriched using streptavidin beads, followed by western blot against DGCR8, DROSHA, ZCCHC8, and DHX9. Vinculin was used as a negative control. ( I ) Enrichment of SINE and LINE elements in genes co-bound by pairs of proteins (DGCR8, DROSHA, DHX9) compared with genes bound by only one protein. Fold change was calculated by comparing the median TE count in co-bound genes to the higher median TE count of singly-bound genes. ( A–D ) P -values are determined by Wilcoxon rank sum test. * P < 0.05, ** P < 0.01, *** P < 0.001, **** P < 0.0001.

Article Snippet: Cells were transfected at 70% confluency, in six-well plates, and transfected with 5 μl of Invitrogen Lipofectamine 2000 Transfection Reagent and 1 μg of Y-shaped-DNA cGAS agonist (G3-YSD) (#tlrl-ydna, Invivogen) or dsRNA analogue polyinosinic:polycytidylic acid (poly(I:C)) (HMW #tlrl-pic, Invivogen).

Techniques: Clone Assay, Binding Assay, Transformation Assay, Immunoprecipitation, Western Blot, Negative Control

Journal: Nucleic Acids Research

Article Title: Control of retrotransposon-driven activation of the interferon response by the double-stranded RNA binding protein DGCR8

doi: 10.1093/nar/gkag190

Figure Lengend Snippet: Loss of DGCR8 leads to immune system activation in 22qDS cells. ( A ) Western blot analyses of DGCR8 protein levels in PA-1 WT, DGCR8 HET, and DGCR8 KO cells. GAPDH serves as loading control (left). Quantification of three independent biological replicates (right). The dashed line represents the WT mean. ( B ) Quantification of IFNB1 mRNA levels by RT-qPCR in PA-1 WT, DGCR8 HET, and DGCR8 KO cells, both untreated and after stimulation with poly(I:C). Expression is represented as 2^-ΔCt, using GAPDH as a normaliser. ( C ) Influenza A Virus RNA levels by RT-qPCR after infection of PA-1 WT, DGCR8 HET, and DGCR8 KO. Expression was normalised to 18S rRNA and plotted relative to WT cells. ( D ) DGCR8 mRNA levels by RT-qPCR in control (healthy cells) and 22qDS immortalised fibroblasts (GM07215F). Data are normalised to GAPDH and expressed relative to WT cells. ( E ) IFNB1 expression levels by RT-qPCR in control (healthy cells) versus 22qDS immortalised fibroblasts (GM07215F), after poly(I:C) stimulation. Data are normalised to RN7SK and represented relative to stimulated healthy (WT) fibroblasts. ( F ) IFNB1 expression levels by RT-qPCR in control (healthy cells) versus 22qDS-derived primary fibroblasts (GM02944), after poly(I:C) stimulation. Data are normalised to GAPDH and represented relative to healthy stimulated cells. ( G ) Percentage of ISGs with a log2FC (22qDS/control) > 0 (up) or log2FC (22qDS/control) < 0 (down) from the following datasets: blood (Lin et al. ), T-lymphocytes (Raje et al. 2022), iPSCs-derived neurons (Lin et al. ), and cerebral cortex organoids from 22qDS-iPSCs (Khan et al. ). ( H ) Correlation between the log2FC (22qDS/control) of 22q11.2 deleted genes and ISGs. The log2FC of the mean of ISGs is plotted on the y -axis. The log2FC of the mean of 22q11.2 genes or the DGCR8 gene is plotted on the x -axis. Each dot represents one dataset, these being: blood ( n = 79 patients, n = 68 controls) (Lin et al. 2021), T-lymphocytes ( n = 13 patients, n = 6 controls) (Raje et al. 2022), iPSCs-derived neurons ( n = 8 patients, n = 7 controls) (Lin et al. 2016), hPSCs, NPCs and neurons ( n = 20 patients, n = 29 controls) (Nehme et al. ) and cerebral cortex organoids differentiated from 22qDS-iPSCs after 25, 50, 50, and 100 days ( n = 15 patients, n = 14 controls) (Khan et al. 2020). On the right, a table representing the top15 22q11.2 genes with the highest correlation index (all genes in ). ( I ) In homeostasis, DGCR8 and dsRNA-BPs such as DHX9 bind and resolve dsRNA structures of TE-rich mRNAs. After unwinding, RNAs are exported to the cytoplasm, where they are unseen by dsRNA receptors (left). In the absence of DGCR8, dsRNA structures remain unresolved. In the cytoplasm, these TE-rich RNAs are recognised by the MDA5 sensor, triggering IFN activation (right). Figure was created in BioRender ( https://BioRender.com/jt7uab7 ). In RT-qPCR, bars represent the average of three biological replicates. Error bars plot mean ± SD. * P < 0.05, ** P < 0.01, *** P < 0.001, **** P < 0.0001. P -values are determined by one-way ANOVA followed by Dunnett multiple comparison post-hoc test.

Article Snippet: Cells were transfected at 70% confluency, in six-well plates, and transfected with 5 μl of Invitrogen Lipofectamine 2000 Transfection Reagent and 1 μg of Y-shaped-DNA cGAS agonist (G3-YSD) (#tlrl-ydna, Invivogen) or dsRNA analogue polyinosinic:polycytidylic acid (poly(I:C)) (HMW #tlrl-pic, Invivogen).

Techniques: Activation Assay, Western Blot, Control, Quantitative RT-PCR, Expressing, Virus, Infection, Derivative Assay, Comparison