Developmental series of gene expression clarifies maternal mRNA provisioning and maternal-to-zygotic transition in a reef-building coral

Abstract

Background: Maternal mRNA provisioning of oocytes regulates early embryogenesis. Maternal transcripts are degraded as zygotic genome activation (ZGA) intensifies, a phenomenon known as the maternal-to-zygotic transition (MZT). Here, we examine gene expression over nine developmental stages in the Pacific rice coral, Montipora capitata, from eggs and embryos at 1, 4, 9, 14, 22, and 36 h-post-fertilization (hpf), as well as swimming larvae (9d), and adult colonies.

Results: Weighted Gene Coexpression Network Analysis revealed four expression peaks, identifying the maternal complement, two waves of the MZT, and adult expression. Gene ontology enrichment revealed maternal mRNAs are dominated by cell division, methylation, biosynthesis, metabolism, and protein/RNA processing and transport functions. The first MZT wave occurs from ~4-14 hpf and is enriched in terms related to biosynthesis, methylation, cell division, and transcription. In contrast, functional enrichment in the second MZT wave, or ZGA, from 22 hpf-9dpf, includes ion/peptide transport and cell signaling. Finally, adult expression is enriched for functions related to signaling, metabolism, and ion/peptide transport. Our proposed MZT timing is further supported by expression of enzymes involved in zygotic transcriptional repression (Kaiso) and activation (Sox2), which peak at 14 hpf and 22 hpf, respectively. Further, DNA methylation writing (DNMT3a) and removing (TET1) enzymes peak and remain stable past ~4 hpf, suggesting that methylome programming occurs before 4 hpf.

Conclusions: Our high-resolution insight into the coral maternal mRNA and MZT provides essential baseline information to understand parental carryover effects and the sensitivity of developmental success under increasing environmental stress.

Keywords: Embryo; Epigenetics; Oocyte; Resilience; Time course; WGCNA.

Fig. 1

Timeline of sample collection including example photographs of life each stage

Fig. 2

Principal coordinates analysis based on sample-to-sample distance. Sample-to-sample distance was computed from all genes passing a low counts filter, wherein a gene must have a count of 10 or greater in at least 2 out of the 24 samples (pOverA 0.083, 10)

Fig. 3

Clustered heatmap of WGCNA module-trait correlations. Correlation values range from -1 (dark blue) to 0 (white) to +1 (red). Developmental expression clusters are shown on top, while sampled life stages are indicated on bottom. Module clusters 1-9 were identified via a distance matrix computed with hclust using module eigengenes

Fig. 4

Clustered (binary cut) semantic similarity matrix of Biological Process gene ontology terms for all genes in the 9 modules identified as the A) maternal mRNA complement (n=92 terms), the 11 modules of the B) first wave of the MZT (n=55 terms), and the 10 modules of the C) second wave of the MZT transition (n=22 terms) and the 7 modules of D)“adult” expression (n=31 terms). Terms representing each cluster are shown in a word cloud, with size of term representing frequency of occurrence in the terms. The generic term, “Process” (identified by * in each graphic), was put as a footnote for visual clarity

Fig. 5

Clustered (binary cut) semantic similarity matrix of Molecular Function gene ontology terms for all genes in the 9 modules identified as the (A) maternal mRNA complement (n=92 terms), the 11 modules of the (B) first wave of the MZT (n=60 terms), the 10 modules of the (C) second wave of the MZT transition (n=47 terms), the 7 modules of the (D) “adult” expression (n=54 terms). Terms representing each cluster are shown in a word cloud, with size of term representing frequency of occurrence in the terms. The generic term, “Activity” (identified by * in each graphic), was put as a footnote for visual clarity

Fig. 6

Expression of key biomarkers (M. capitata gene id) in the maternal-to-zygotic transition: (i) Cyclin-B (g71356), (ii) Smaug (g4639), (iii) Kaiso (g60350), (iv) Sox2 (g53225), (v) Wnt8 (g33149) and (vi) Brachyury (TBXT; g68947). Expression is plotted as counts transformed with variance-stabilization, which is logarithmic in nature. Points and error bars display mean±standard error of the mean. Red and blue arrows indicate significant (padj<0.05) negative and positive changes in gene expression between time points with log2FoldChange greater than absolute 1 (|log2FoldChange| > 1). M=maternal expression, Z=zygotic expression, ZGA=Zygotic genome activation

Fig. 7

Expression of enzymes (M. capitata gene id) involved with transcriptional regulation during the maternal-to-zygotic transition: (i) DNMT3a (g25804), (ii) DNMT1 (g53952) (iii) TET (adi2mcaRNA27872_R0), (iv) MBD2 and MBD (both g53132), (v) UHRF1 (adi2mcaRNA19502_R1) and (vi) BRG1 (g68733). Expression is plotted as counts transformed with variance-stabilization, which is logarithmic in nature. Points and error bars display mean±standard error of the mean value of the VST normalized data. Red and blue arrows indicate significant (padj<0.05) negative and positive changes in gene expression between time points with log2FoldChange greater than absolute 1 (|log2FoldChange| > 1)