Changes in RNA Splicing in Developing Soybean (Glycine max) Embryos

Abstract

Developing soybean seeds accumulate oils, proteins, and carbohydrates that are used as oxidizable substrates providing metabolic precursors and energy during seed germination. The accumulation of these storage compounds in developing seeds is highly regulated at multiple levels, including at transcriptional and post-transcriptional regulation. RNA sequencing was used to provide comprehensive information about transcriptional and post-transcriptional events that take place in developing soybean embryos. Bioinformatics analyses lead to the identification of different classes of alternatively spliced isoforms and corresponding changes in their levels on a global scale during soybean embryo development. Alternative splicing was associated with transcripts involved in various metabolic and developmental processes, including central carbon and nitrogen metabolism, induction of maturation and dormancy, and splicing itself. Detailed examination of selected RNA isoforms revealed alterations in individual domains that could result in changes in subcellular localization of the resulting proteins, protein-protein and enzyme-substrate interactions, and regulation of protein activities. Different isoforms may play an important role in regulating developmental and metabolic processes occurring at different stages in developing oilseed embryos.

Figure 1

Flowchart of bioinformatics analyses used for differential expression of splice isoforms and subsequent data mining. The initial steps of the RNA-seq pipeline are described in [53] and tools are in red boxes.

Figure 2

Distribution of known, novel, and other class isoforms in developing soybean embryos. Results are shown for 2,942 isoforms encoded by genes expressing between two and six isoforms showing changes in their transcript levels during soybean embryo development (day 10 (D10)–day 55 (D55)). Other Cuffcompare classes represent o (partial overlap with known transcript), x (antisense exon), and s (antisense intron).

Figure 3

Major metabolic processes of CCNM and relevant clusters enriched in transcripts encoding a variety of proteins involved in these processes. GO enrichment analysis was performed on isoforms in all 114 clusters (p-value < 0.05). Only clusters displaying four basic trends consistent with: (A) early (initial decrease followed by stable low or no expression; black); (B) seed filling (moderate increase followed by a moderate decrease; red); (C) maturation (stable low or no initial expression followed by a final increase; blue); and (D) early and maturation (initial decrease followed by no or low stable expression and final increase in transcript levels; gray) CCNM are shown. Clusters showing conceptually similar trends based on visual assessment have the same color that matches the corresponding trend.

Figure 4

AS of 6-phosphogluconate dehydrogenase involved in oxidative pentose phosphate pathway. (A) SpliceGrapher representations of individual isoforms detected during soybean embryo development. Only Glyma19g05120.2, N3, and N4 (corresponding to ) were differentially expressed. Alternative 5' splicing yielded these novel isoforms. (B) Proteins resulting from these three transcripts with the respective positions of NAD-binding domains (pfam03446) and 6-phosphogluconate dehydrogenase (6PGD) C-terminal domains (pfam00393). Domain-related information was obtained from NCBI [97]. Numbers in blue represent amino acid residues in the individual proteins. (C) Overlaid expression profiles of Glyma19g05120.2, N3, and N4, obtained from SplineCluster. Numbers shown in parenthesis represent the isoform clusters. Representative developing soybean embryos are shown for each time point along with a temporal representation of processes occurring during embryo development.

Figure 5

AS of 3-ketoacyl-CoA thiolase involved in fatty acid degradation during late maturation stages of soybean embryo development. (A) SpliceGrapher representations of individual isoforms detected during soybean embryo development. Alternative 3' and/or 5' splicing of many novel isoforms were identified. Glyma10g24590.1 and 3 correspond to and were the only two transcripts displaying differential expression. (B) Proteins resulting from two isoforms of interest with the respective positions of active and dimer interface sites in the N- and C-terminal thiolase domains (pfam00108 and pfam02803, respectively), showing the positions of individual amino acid residues (red triangles). Bright blue sections of the protein show the sequences that were not included in domain analysis due to an amino acid composition-related bias. (C) Overlaid expression profiles of Glyma10g24590.1 and 3 obtained from SplineCluster.

Figure 6

AS of ABA 8'-hydroxylase 1-like protein possibly involved in ABA degradation. (A) SpliceGrapher representations of individual isoforms detected during soybean embryo development. Alternative 3' and/or 5' splicing, and/or intron retention was observed in these isoforms. Glyma10g24590.1 and N7 corresponded to isoforms . The stop sign in N7 indicates a premature stop codon. (B) Proteins resulting from the two isoforms of interest, showing the respective positions of the p450 domains (pfam00067). Isoform N7 contains a truncated p450 domain. (C) Overlaid transcriptional profiles of Glyma10g24590.1 and N7 obtained from SplineCluster.

Figure 7

Tree displaying GO-enriched processes regulated by AS and associated with RNA splicing. Filtered enriched GO categories were searched for three keywords (“RNA splicing”, “dormancy”, and “response to hormone stimulus”) in all 114 clusters and the GO database [66] was used to identify the parents and children for each GO term. These were then combined to obtain a single tree representing all enriched GO terms corresponding to each keyword. This large tree was then separated into three sub-trees corresponding to these three processes and the tree corresponding to the “RNA splicing” category is shown. GO terms that were enriched significantly (p < 0.05) in genes involved in RNA splicing are shown in green. The clusters containing these RNA-splicing-related genes are shown below the corresponding category. This figure was generated in LucidChart [106].