|
Assessing Pou5f3 and Sox3 roles in genome activation.(A) Embryos injected with 40 ng of each pou5f3.2, pou5f3.3, and sox3 morpholinos have gastrulation defects compared to stage 10.5 control embryos, and fail to close the blastopore. Embryos injected with two of the three morpholinos plus 40 ng of gfp morpholino also had developmental delays, with one of two batches of pou5f3.2+pou5f3.3 morpholino embryos also failing to close the blastopore. Stage 12 phenotype counts are summed over two different rounds of injections on different days. Scale bar = 0.5 mm. (B) Heatmap showing RNA-seq intron log2 fold difference compared to control for exonic and intronic signal, with individual replicates shown. The left columns are for morpholino treatments without cycloheximide, right columns with cycloheximide comared against a cycloheximide control. Triptolide samples are shown for comparison. All samples are from embryos collected when untreated wild-type controls were at stage 9. (C) Heatmap for two morpholino samples that likely had problems with the cycloheximide treatment, based on the RNA-seq patterns. These samples are excluded from subsequent analyses. (D–F) Biplots comparing exonic expression levels in cycloheximide-treated control embryos versus embryos with deficient genome activation. Primary-activated genes with minimal maternal contribution (strictly zygotic) are purple circles, maternal-zygotic genes detected by exonic increases are orange triangles. TPM = transcripts per million. (G–I) Venn diagrams showing overlap of significantly different genes in the morpholino treatments. (J) Proportions of genes classified as significantly up-regulated in morpholino treatments detected by intronic levels. (K) log2 fold difference RNA-seq expression of pou5f3.2+pou5f3.3+sox3 morpholino treated embryos over control for different groups of genes, according to whether they were classified as significantly up-regulated in various morpholino combinations. Although some gene groups achieve significance in the double but not triple morpholino conditions, all gene groups significantly up in at least one treatment nonetheless have elevated expression in the triple morpholino conditions. Center line, median; box limits, upper and lower quartiles; whiskers, 1.5x interquartile range; points, outliers. N = 60, 778, 210, 102, 140, 215, 553, left to right. (L) Proportions of morpholino-affected genes (no cycloheximide) stratified into strictly zygotic genes (Z, maternal contribution <1 TPM) versus maternal-zygotic genes and whether wild-type activation is only in the stage 9 blastula or also observed in differentiated lineages (left, middle). Strictly zygotic genes overall are significantly more likely to be down-regulated by morpholino treatment (p=6.6 × 10–18, χ-squared test, 3 d.o.f.), and this is a stronger effect in blastula-specific genes compared to genes also up-expressed in differentiated lineages (p=2.0 × 10–34, χ-squared test, 6 d.o.f.). (Right) proportions are not significantly different when comparing in which lineages those genes are further expressed (p=0.11, χ-squared test, 6 d.o.f.). (M) Proportions of genes affected by morpholino treatment in the presence of cycloheximide, according to their homeolog expression patterns. Strictly zygotic genes encoded in both subgenomes are more likely to depend on Pou5f3 and Sox3 for their primary activation than singleton genes, regardless of the patterns of homeolog activation (p=0.0020, χ-squared test, 5 d.o.f.). Maternal-zygotic genes have a subtler trend. (M) Proportions of genes affected by morpholino treatment without cycloheximide, according to their homeolog expression patterns. Genes expressed only on the S homeolog, particularly S singletons, are somewhat less likely to be regulated by Pou5f3 and Sox3 (p=3.4 × 10–5, χ-squared test, 10 d.o.f.), though only among maternal-zygotic genes.
|
