Figure 7. NF-Y proteins regulate TALE GRN expression and form complexes with TALE factors.

See also Figure 7—figure supplement 1. (A) Sequence logo and localization relative to Prep peak summits of motifs identified by DREME at Prep_3.5hpf peaks. (B) Enrichment of motifs in Prep_3.5hpf and Prep_12hpf-only peaks as defined by AME. p-values for enrichment above random occurrence (3.5hpf and 12hpf-only columns) or between two Prep peak populations (3.5hpf vs 12hpf-only and 12hpf-only vs 3.5hpf columns) were calculated using the ranksum test in AME. Motifs are represented in IUPAC code (M = A or C; R = A or G). (C) ChIP-qPCR showing NF-YB binding at CCAAT motif-containing MPADs in 9hpf embryos. MPADs are labeled with the name of the nearest gene. Data of three independent biological replicates are presented as mean fold change ± SEM of NF-YB IP vs control IgG. Statistical test: unpaired t-test. (D) Venn diagram illustrating the overlap of Prep and NF-YB peaks in mESCs. Two peaks are considered to overlap if their summits are within 500 bp. (E) RT-qPCR analysis of gene expression in 12hpf NF-YDN injected embryos. Results are shown as gene expression fold-change in NF-YDN vs control for select TALE-dependent genes. Data of three independent experiments are presented as mean fold change ± SEM of NF-YDN injected vs control embryos. Statistical test = unpaired t-test. (F) H3K27ac/Histone H3 signal ratio at MPADs (labeled with the name of the nearest gene) as determined by ChIP-qPCR in 9hpf control vs NF-YDN injected embryos. Data of three independent biological replicates are presented as mean fold change ± SEM of NF-YDN vs control. Statistical test: unpaired t-test. (G) Co-IP experiments showing interaction of Myc-Prep (left panels) and HA-Pbx4 (right panels) with Flag-NF-YB in transfected HEK293 cells. HC = Ig heavy chain. Asterisks indicate non-specific signal. (H) Model diagram. At blastula stages (left side) TALE binds DECA motifs (TGATTGACAG) near NF-Y motifs (CCAAT). At this stage, most binding sites are occupied by nucleosomes and those associated with developmental control genes are marked by H3K27me3 (red lollipops). Binding of TALE and NF-Y leads to deposition of H3K27ac (green lollipops) and improved accessibility. At segmentation stages (right side), TALE continues to bind DECA motifs near NF-Y motifs, but Prep also binds HEXA motifs (TGACAG) near PBX:HOX motifs (TGATTTAT). Most of the HEXA motifs lack nucleosomes and are found within 40 kb of a DECA/NF-Y site (indicated by dashed connecting line). At this stage, developmental control genes are marked by H3K27ac and are expressed.

Figure 7—figure supplement 1. NF-Y TF regulates anterior embryonic structures and interacts with Prep and Pbx.

(A) Left panel: Expression levels of maternal and paternal nf-yb transcripts during early zebrafish embryogenesis (using data from [Harvey et al., 2013]). At 3.5hpf, transcripts from the paternal allele (which requires zygotic activation) are not detected, demonstrating that only maternal nf-yb transcripts are present at this time point. Right panel: RT-PCR analysis on 0–1.25hpf, 3.5hpf and 9hpf zebrafish embryos demonstrating that nf-ya, nf-yb and nf-yc transcripts are maternally deposited. (B) Images of 4 dpf zebrafish following injection with mRNA encoding a NF-Y dominant negative construct (NFY-DN). Control gfp mRNA injected fish show normal wild type morphology, while 21/38 and 11/38 NF-YDN injected fish have severe or mild developmental defects, respectively. Six NF-YDN injected embryos died. Note that some developmental defects are shared with TALE KD injected embryos; for example smaller head, eyes and cardiac edema (see Figure 1—figure supplement 1A). (C) Average H3K27ac, H3K27me3 and nucleosome signal at TALE sites with (red line) or without (black line) adjacent NF-Y (CCAAT) motifs. (D) Distribution of the various binding motifs (DECA, HEXA, TALE:Hox, NF-Y) at 12hpf-only and 3.5hpf Prep-occupied sites (3.5hpf sites are further broken down into MPAD classes), as well as at Prep-occupied sites associated with TALE-GRN genes (DEG = differentially expressed genes identified by RNA-seq after TALE-KD). Numbers in boxes indicate percent of all sequences in each class that contain each of the motifs. (E) Reporter assays in HEK293 cells testing enhancer activity. Each element was tested in triplicate using two concentrations (100 and 400 ng) of reporter plasmid in the presence (+TF) or absence (-TF) of co-transfected TALE and NF-Y factors. Data is presented as mean ± SD fold induction over control plasmid. (F) Co-IP experiments showing interaction of Myc-Prep (left panels) and HA-Pbx4 (right panels) with Flag-NF-YA in transfected HEK293 cells. HC = Ig heavy chain. Asterisks indicate non-specific signal. (G) Localization of early zygotic genes (Lee et al., 2013) relative to Prep ChIP-seq peak summits. The number of genes within 30 kb of Prep binding sites is indicated above each bar. p-values for enrichment above a random set of genes were calculated using the Pearson correlation test. (H) Localization of first-wave zygotic genes relative to MPADs and non-MPADs. Numbers above bars indicate the number of genes within 30 kb of each type of Prep binding site. p-values for enrichment above a random set of genes were calculated using the Pearson correlation test.