Lola-I is a promoter pioneer factor that establishes de novo Pol II pausing during development

Fig. 1. Lola-I is required for the loading of paused Pol II to target genes in the late Drosophila embryo. a At promoters that have no Pol II occupancy in the early (2–4 h AEL) embryo but acquire paused Pol II in the late (14–17 h AEL) embryo (opening set), the Lola-I motif was identified de novo within 200 bp upstream of the TSS by MEME analysis (e-value = 8.1e−091). The Lola-I motif was enriched 6.6-fold in the opening set (n = 492) vs the constant set (n = 843) (*P = 8.75e−34, chi-squared test with multiple-testing correction). b A Western blot with antibodies specific for the Lola-I isoform shows that Lola-I increases in expression during embryogenesis. Tubulin is shown as control. Results shown are representative of at least comparable biological experiments and are consistent. Source data is provided as a Source Data file. c Single-gene example of the ChIP-seq data showing that Lola-I binding is found at the promoter of Tpi, a gene that acquires paused Pol II over time. d Heatmaps showing that the ~60% of Lola-I-bound regions found at promoters (n = 329) is associated with an increase in Pol II occupancy, RNA levels from the early (2–4 h AEL) to the late (14–17 h AEL) embryo, and…

Fig. 2. Lola-I establishes paused Pol II throughout the embryo. a Immunostaining using the Lola-I antibodies (yellow) shows that Lola-I is expressed ubiquitously in the late (14–17 h AEL) Drosophila embryo. As control, Lamin is shown (pink), which is also ubiquitous and nuclear. The brightness and contrast settings of the lookup table are linearly adjusted for clarity. The settings in the individual panels are the same as in the merge. scale bar: left - 100 µm and right - 10 µm. b ChIP-seq data shown as normalized reads per million (RPM) from isolated embryonic tissues of late-stage embryos (14–17 h AEL) using either Lola-I antibodies (turquoise) or Pol II antibodies (dark blue) reveal that Lola-I binding and paused Pol II are found in all examined tissues at Lola-I targets, even when the target gene is expressed only in a specific tissue (Gip is shown as example). Ubiquitous Pol II is not found for all genes, e.g., for the control gene Osi20, Pol II binding and expression is restricted to tracheal cells. The tissue-specific expression of Gip in crystal cells and Osi20 in tracheal cells are known from in situ hybridization shown below, courtesy of Berkeley Drosophila Genome Project93,94.…

Fig. 3. Lola-I promotes tissue-specific gene expression through an effect on the promoter. a Left panel: Boxplot of bulk RNA-seq data show that Lola-I targets are down-regulated in lola-I mutants compared to wild-type (14–17 h AEL), while control genes show overall similar expression levels. This confirms that Lola-I mediates its effect by directly binding to target promoters, and not by indirect effects that would affect all genes (Wilcoxon two-sided test, lolaORC4/ORC4: *P = 1.8e−11, lolaORC4/ORE50: *P < 2e−16). Right panel: Boxplot of bulk RNA-seq data show that genes associated with distal Lola-I binding sites are not significantly affected in lola-I mutants compared to wild-type (14–17 h AEL), while genes with Lola-I binding at the promoter (with or without distal binding) are reduced in expression, showing that Lola-I largely mediates its effect through binding promoters. b scRNA-seq analysis of wild-type and lola-I mutant embryos at 14–14.5 h AEL suggests that Lola-I is required for both tissue-specific expression (left) and basal expression levels in other tissues (right) at target genes (defined by >2-fold loss in Pol II occupancy in mutant embryos). Expression, shown as the…

Fig. 4. Single-molecule FISH reveals that Lola-I lowers the gene activation barrier. a Example of single-molecule FISH images of Gip (yellow), and PPO1 and PPO2 genes (pink) in wild-type embryos and lola-I −/− mutant embryos (DAPI - blue). Data were acquired with ×40 magnification, scale bar - 20 µm, with the same brightness and contrast settings. Top insets show the nascent Gip signal in the nuclei; bottom insets show the single Gip RNAs in the cytoplasm. The brightness and contrast in each inset was adjusted linearly for clarity. b The percentage of crystal cells (defined by high expression of PPO1 and PPO2) that express high levels of Gip is strongly reduced in lola-I mutant embryos compared to wild-type. Barplots show the mean, and the whiskers show the standard error of the mean, with the number of embryos (n): wild-type: 10–12 h n = 7, 12–14 h n = 8, 14–16 h n = 4, lola-I mutant: 10–12 h n = 7, 12–14 h n = 7, 14–16 h n = 7. Reduced Gip expression is observed for multiple time points showing that it is not due to a developmental delay and that Gip expression only mildly recovers over time. c The percentage of PPO1 and PPO2 positive crystal cells with a visible Gip nascent site of…

Fig. 5. Lola-I is a promoter pioneer factor. a Heatmap of normalized MNase-seq data shows specific loss of nucleosome occupancy at the Lola-I target promoters during the late stages of Drosophila embryogenesis (Wilcoxon two-sided test, *P wt-early vs wt-late: 3.4e−9, wt-late vs mut-late 14–17 h: 3.4e−8). This does not occur in lola-I mutant embryos, as there is no statistically significant difference between wild-type early embryos and late lola-I mutant embryos. b Box plots of the MNase signal (centered on the Lola-I peaks) are shown for promoter-proximal regions and distal Lola-I-bound regions (Supplementary Data 2), which both show a significant decrease in late wild-type embryos (Wilcoxon two-sided test, *P distal, wt-early vs wt-late: *P = 4.5e−09, promoter wt-early vs wt-late: *P < 2e−16, distal wt-late vs mut-late: *P = 1.1e−08, promoter wt-late vs mut-late: *P = 4.1e−10). c Pol II recruitment, shown as change in normalized Pol II enrichment from the early to the late stage, only occurs when the Lola-I peak is found in close proximity to the TSS, showing that Lola-I depletes nucleosomes independently of Pol II recruitment (Wilcoxon two-sided test, *P = 1.2e−8). d Lola-I binds its…