Gene expression dynamics before and after zygotic gene activation in Drosophila early embryogenesis

Abstract

Post-transcriptional gene regulatory mechanisms are fundamental to the determination of gene expression dynamics and especially crucial for the earliest stages of animal development in which transcription is nearly silent. Here, we performed high-resolution total RNA sequencing and quantitative mass spectrometry analysis simultaneously on Drosophila maternal-to-zygotic transition (MZT). Further, this study is the first to report the proteome-wide quantitative changes in protein ubiquitination in Drosophila MZT. Our results indicate that timely ubiquitination of the distinct target proteins during MZT is essential for the downregulation of protein expression levels. Profiling of the RNA-associated proteome changes in Drosophila MZT suggested that RNA binding can be regulated without the respective change in net protein expression levels for over 200 proteins, including Pcid2, Sym, and Cpsf73. We report that highly dynamic and post transcriptionally regulated protein expression level changes can occur at the earliest stages of the Drosophila MZT.

Keywords: Biological sciences; Developmental biology; Natural sciences; Transcriptomics.

Graphical abstract

Figure 1

Overview of transcriptome- and proteome-wide analyses in Drosophila MZT (A) Schematic outline of the Drosophila embryo and unfertilized egg collection time points. The respective developmental stages are noted above the timeline, and distinct time points are represented with E1–E6 and UF1–UF4 for the embryos and egg samples, respectively. (B) Principal component analysis of the transcriptome- and proteome-wide results (N = 2; Rep1 and Rep2) at distinct time points. The proportion of variance explained by each principal component (PC) is indicated as percentage in axes labels. (C) Dynamic range of mRNA and protein level changes during 0–4.5 h AEL in Drosophila embryogenesis. “Corr” indicates the Pearson correlation coefficient between mRNA and protein fold changes.

Figure 2

Clustering analysis of mRNA and protein expression levels (A) K-means clustering analysis of the embryonic mRNA and protein expression changes yielded the groups G1–G12, with the number of the genes in each group denoted as N (number). Groups were defined based on combined mRNA and protein expression levels at time points. The color of each cell represents the mean log fold change of mRNA (left panels) and protein (right panels) for the genes at each time point of embryos (E1-E6) and unfertilized eggs (UF1-4). (B) Comparison of the embryonic proteome profile from the previous study by Becker et al. with our study. (C) The number of identified proteins common to both studies or exclusive to our study is presented for each group of gene expression pattern analysis. (D) Western blot analysis of the embryo samples for the representative genes. Their respective mRNA and protein expression fold-changes are also shown.

Figure 3

Gene Ontology (GO) terms overrepresented by each group Top three enriched GO terms for each group. The black dashed line indicates an adjusted p-value of 0.05. The enrichment p-value for each GO term was calculated and adjusted using TopGO. The small line plots on the left indicate the median (dark color) and standard deviation (light color) of RNA and protein temporal dynamics.

Figure 4

Relationship between protein and mRNA expression changes in embryos pre- and post-ZGA (A) Scatterplot comparing protein and mRNA expression changes in E1 to E4(left panel) and E4 to E6 (right panel). Proteins with increased expression levels (log2 fold change >0.5) are categorized based on corresponding mRNA expression changes: log2 fold change <0, between 0 and 0.5, and >0.5. The number and proportion of proteins in each category are shown from left to right. (B) Scatterplot of log2 protein fold changes (x axis) and log2 absolute protein quantities (y axis). The mean intensity-based absolute quantification value (iBAQ) across all embryo time points serves as a proxy for protein absolute quantities. The protein level changes from E1 to E4 (left) and E4 to E6 (right).

Figure 5

Ubiquitome analysis of the Drosophila embryos (A) Schematic representation of the ubiquitome analysis in Drosophila embryos collected at three distinct time points (N = 3). (B) Overlap between the ubiquitinated protein identification at three different time points. (C) Overlap between the protein identification in global, ubiquitome, and phospho-proteome analyses. (D) Heatmap displaying the fraction of proteins mapped to peptides with internal K-GG modifications across groups identified in the gene expression pattern analysis. (E) Mean spectral counts of ubiquitinated proteins for each cluster represented as boxplots. The boxplots show the median (center line), first and third quartiles (lower and upper box limits, respectively), and 1.5 times the interquartile range (whiskers).

Figure 6

Dynamics of the protein ubiquitination and expression level changes (A) K-means clustering analysis of the total and ubiquitinated protein expression level changes. The color of the heatmap indicates the log2 fold change relative to the median of the total protein samples for “total” proteins, and relative to the median of the ubiquitinated protein samples for “ubiquitinated” proteins. Each gene was subjected to this log2 fold change calculation separately. (B) Total and ubiquitinated protein expression levels of selected genes. Left bar plots show the change in mean total protein expression at each time point. Green and orange dots indicate the total protein quantities in each replicate (N = 2). Right bar plots represent the mean expression level of ubiquitinated proteins (N = 3). Error bars indicate the standard error of the mean.

Figure 7

Characteristics of genes whose transcripts are bound by the regulatory RNA-binding proteins (RBPs), Brat, Pum, and Smg (A) Overlap of genes previously reported to be bound by Brat, Pum, and Smg proteins based on RIP analysis.^, (B) Heat maps showing RNA and protein expression levels (left, as shown in Figure 2A), p-values of Brat, Pum, and Smg target gene enrichment (middle), and proportions of Brat, Pum, and Smg target genes (right). The p-values for enrichment were calculated using the Fisher’s exact test. The proportions in the right panel indicate the number of RBP target genes divided by the total number of genes in each group. (C) Scatterplots showing the changes in poly(A)-tail length (x axis) and translation efficiency (TE) (y axis), reported previously in Eichhorn et al. Vertical and horizontal lines indicate the mean poly(A)-tail length and TE changes, respectively, of RBP target genes (colored) and other genes (black).

Figure 8

Formaldehyde crosslinking-based RNA-interacting proteome (FAX-RNA interactome) analysis in Drosophila embryos (A) High-confidence FAX-RNA interactome at different stages in Drosophila embryos (N = 3). Volcano plots displaying the log2 fold change of FAX over no crosslinking control (x axis) and –log10 p-value (y axis). Proteins with log2 fold change >1 and statistically significant enrichment over the no crosslinking control (BH-FDR <0.01) are highlighted in red. (B) Venn diagram showing the overlap of RNA interactomes defined in different stages of embryo samples. (C) Upset plot showing the number of proteins that are unique to or shared between FAX and previously reported Drosophila RNA interactomes.^,, (D) FAX-RNA interactome and total protein expression level changes at the indicated time point in Drosophila embryo samples. Partially transparent red colored dots represent proteins with significant changes in FAX-RNA interactome enrichment (log2 fold change >1 and BH-FDR <0.05) that are attributable to corresponding changes in total protein expression levels (total protein log2 fold change >0.5 or < −0.5). Partially transparent yellow colored dots represent dynamic RNA-binding proteins with significant FAX-RNA interactome enrichment changes, despite minimal changes in total protein expression levels (−0.5 < total protein log2 fold change <0.5). Representative proteins are marked in blue, with their names labeled.