Figure 3. T-REX17 regulation at the SOX17 topological domain.

(A) Schematic of T-REX17 locus regulation in the absence (top) or presence (bottom) of a targeting dCas9-KRAB-MeCP2 complex, decorating T-REX17 promoter with an H3K9me3 mark 355 bp upstream of the TSS. (B) Time-resolved qRT-PCR showing the expression of T-REX17 during EN differentiation in the presence or absence of dCas9-KRAB-MeCP2 complex targeting T-REX17 promoter (normalized to the housekeeping gene 18s). Symbols indicate the mean and error bars indicate SD across three independent experiments. Individual data points are displayed. (C) smRNA-FISH of T-REX17 in sgCtrl (left) and sgT-REX17 (right) EN cells counter-stained with Hoechst. Scale bars, 10 µm. For an extended field of view see Figure 3—figure supplement 1B. (D) Virtual 4C analysis from capture Hi-C experiments in sgCtrl and sgT-REX17 EN cells using SOX17 promoter as viewpoint, with 2 kb resolution (upper panel). SOX17 EN ChIP-seq (RPKM) and RNA-seq (CPM) profiles in the two conditions are shown in the tracks (lower panel). eSOX17 and pT-REX17 are highlighted in grey. (E) Time-resolved qRT-PCR showing the expression of SOX17 during EN differentiation in the presence or absence of dCas9-KRAB-MeCP2 complex targeting T-REX17 promoter (normalized to the housekeeping gene 18s). Symbols indicate the mean and error bars indicate SD across three independent experiments. Individual data points are displayed. (F) Heatmap showing SOX17 binding distribution genome-wide in sgCtrl and sgT-REX17 EN. The displayed peaks represent the union of the identified peaks in the two conditions (n=61.153). (G) SOX17 ChIP-seq and RNA-seq tracks at the T-REX17 locus showing SOX17 binding at the SOX17 enhancer (eSOX17) and T-REX17 promoter (pT-REX17). SOX17 binding on pT-REX17 results in T-REX17 activation, if pT-REX17 is not targeted by dCas9-KRAB-MeCP2.

Figure 3—figure supplement 1. SOX17 and T-REX17 reciprocal gene expression regulation.

(A) IF staining for dCas9 in PSCs expressing dCas9-KRAB-MeCP2 and counter-stained with DAPI. Mock samples represent secondary antibody only controls. Scale bars, 50 µm. (B) smRNA-FISH of T-REX17 in sgCtrl (left) and sgT-REX17 (right) EN cells counter-stained with Hoechst. Scale bars, 10 µm. Magnified regions are shown in Figure 3C. (C) H3K9me3 ChIP-qPCR enrichment percentages over input is represented at different regions of the genome in sgCtrl and sgT-REX17 endoderm cells. Bars indicate mean values, error bars indicate the SD across three independent experiments. See raw data in Supplementary file 4. (D) Capture Hi-C sequencing subtraction map of the EN sgCtrl-sgT-REX17 at the SOX17 locus and at the SOX17 loop domain. eSOX17 loop interaction with SOX17 promoter is shown in the magnification and highlighted by the dotted lines (significance threshold: log₂FC ± 0.5, p<0.01). (E) Boxplots showing log₂TPM values for genes in SOX17 neighboring TAD in sgCtrl and sgT-REX17 endoderm cells as measured by RNA-seq. Boxes indicate 25th and 75th percentiles, central lines indicate the median and whiskers show min and max values. (F) Euler diagram showing the overlap between SOX17 ChIP-seq peaks identified in sgCtrl and sgLNCOSOX17. The displayed peaks represent the union of the identified peaks in the two conditions (n=61.694), also displayed in Figure 3F. (G) Genotyping PCR-products, generated by two different primer pairs to profile SOX17 gene ablation (see schematic in G). Expected amplicon sizes within a particular genetic background are shown on the side of the agarose gel picture. (H) Schematic of the Cas9 based SOX17 gene ablation strategy. Genotyping PCR products are depicted in (E). sgRNA sequences are highlighted in grey while Cas9 targeting sites are depicted by dashed lines. Sanger sequencing results are summarized below the query sequence and detected allele frequency are displayed on the side for each respective genotype. (I) Western Blot showing SOX17 levels in PSCs and EN cells for the three indicated genotypes. LAMIN-B is used as loading control. (J) qRT-PCR showing SOX17, T-REX17, GATA4 and NANOG expression in PSCs and EN cells for the three indicated genotypes. Fold change is calculated relative to the 18s housekeeping gene. Bar indicate the means, error bars represent SD across three independent replicates. (K) Schematic of the strategy to generate the SOX17-mCitrine reporter cell line. (L) Schematic of the strategy to generate the T-REX17^p(A)/p(A) cell line. (M) Genotyping PCR products, generated by two different primer pairs to profile the early poly(A) knock-in. Expected amplicon sizes within a particular genetic background are shown on the side of the agarose gel picture. (N) qRT-PCR showing T-REX17, mRuby, and SOX17 expression in PSCs and EN cells for the two indicated genotypes. Fold change is calculated relative to the 18s housekeeping gene. Bar indicate the means, error bars represent SD across three independent replicates.

Figure 3—figure supplement 2. T-REX17 interacts with HNRNPU.

(A) Schematic representation of T-REX17 RNA-pulldown experimental workflow. (B) T-REX17-pulldown qPCR showing enrichment of T-REX17 in both Even and Odd probe sets samples as compared to LacZ in endoderm cells. GAPDH RNA has been used as a negative control. (C) Ranked T-REX17 protein interaction partners as measured by mass spectrometry. Log₂LFQ (Even +Odd/LacZ) has been calculated for each measured peptide to highlight protein partners which are enriched in Even and Odd samples as compared to LacZ sample (positive values). LFQ, Label Free Quantification. (D) Schematic representation of HNRNPU RNA immunoprecipitation (RIP) experimental workflow. (E) Western Blot showing HNRNPU enrichment in the HNRNPU IP as compared to IgG control. Unbound sample represents post-IP supernatant. (F) RIP-qPCR showing enrichment of T-REX17 in HNRNPU IP as compared to IgG in endoderm cells. 18s, GAPDH, and U1 RNAs have been used as negative controls, while XIST and NEAT1 as positive controls. Exon-specific primer pairs have been used to probe T-REX17-HNRNPU interaction. Dots indicate two independent biological replicates.