Transcriptome Analysis Reveals the Genes Related to Pollen Abortion in a Cytoplasmic Male-Sterile Soybean (Glycine max (L.) Merr.)

Abstract

Cytoplasmic male sterility (CMS) lays a foundation for the utilization of heterosis in soybean. The soybean CMS line SXCMS5A is an excellent CMS line exhibiting 100% male sterility. Cytological analysis revealed that in SXCMS5A compared to its maintainer SXCMS5B, its tapetum was vacuolated and abnormally developed. To identify the genes and metabolic pathways involving in pollen abortion of SXCMS5A, a comparative transcriptome analysis was conducted between SXCMS5A and SXCMS5B using flower buds. A total of 372,973,796 high quality clean reads were obtained from 6 samples (3 replicates for each material), and 840 differentially expressed genes (DEGs) were identified, including 658 downregulated and 182 upregulated ones in SXCMS5A compared to SXCMS5B. Among them, 13 DEGs, i.e., 12 open reading frames (ORFs) and 1 COX2, were mitochondrial genome genes in which ORF178 and ORF103c were upregulated in CMS lines and had transmembrane domain(s), therefore, identified as CMS candidate mitochondrial genes of SXCMS5A. Furthermore, numerous DEGs were associated with pollen wall development, carbohydrate metabolism, sugar transport, reactive oxygen species (ROS) metabolism and transcription factor. Some of them were further confirmed by quantitative real time PCR analysis between CMS lines with the same cytoplasmic source as SXCMS5A and their respective maintainer lines. The amount of soluble sugar and adenosine triphosphate and the activity of catalase and ascorbic acid oxidase showed that energy supply and ROS scavenging decreased in SXCMS5A compared to SXCMS5B. These findings provide valuable information for further understanding the molecular mechanism regulating the pollen abortion of soybean CMS.

Keywords: RNA sequencing; cytoplasmic male sterility; gene expression; pollen abortion; soybean (Glycine max (L.) Merr.).

Figure 1

Microscopic observations of anthers from the soybean cytoplasmic male sterility (CMS) line SXCMS5A and its maintainer SXCMS5B. (A–D) Transverse sections of sterile anthers; abnormal tapetum and abnormal anthers developed in SXCMS5A. (E) Mature pollen grains stained by I₂-KI in SXCMS5A. (F–I) Transverse sections of fertile anthers; normal tapetum and normal anthers developed in SXCMS5B. (J) Mature pollen grains stained by I₂-KI in SXCMS5B. MSP, microspore; T, tapetum; PG, pollen grain; Bars = 20 μm.

Figure 2

Analysis of differentially expressed genes (DEGs) between SXCMS5A and SXCMS5B. (A) Number of upregulated and downregulated DEGs. (B) Volcano plot comparing DEGs. Red dots, green dots, and black dots indicated DEGs that were significantly upregulated, significantly downregulated, or showed no significant difference in expression, respectively. (C) Relative expression level of selected DEGs. The y-axis indicated relative mRNA expression level, determined by RNA sequencing (RNA-seq) and quantitative real time PCR (qRT-PCR) analysis. The results were obtained from three biological replicates. FC, fold change; FDR, false discovery rate.

Figure 3

Top 20 Kyoto encyclopedia of genes and genome (KEGG) pathway analysis of DEGs between SXCMS5A and SXCMS5B. The x-axis indicated the rich factor corresponding to each pathway and the y-axis indicated name of the KEGG pathway. The dot color represented the Q values of the enrichment analysis. The size and color of bubbles represented the number and degree of enrichment of DEGs, respectively.

Figure 4

Analysis of DEGs in mitochondrial genome between soybean CMS lines and their maintainer lines. (A) Heat map of DEGs in mitochondrial genome between SXCMS5A and SXCMS5B. The heat map was conducted using MeV 4.9 software. Log₂(FC) values were obtained from the RNA-seq data. (B–D) Relative expression level of ORF103c, ORF151 and ORF178 between soybean CMS lines and their maintainer lines. by qRT-PCR analysis. Asterisk indicated statistical significance: *, p < 0.05; **, p < 0.01; ***, p < 0.001. (E,F) Transmembrane domain analysis of ORF103c and ORF151. The abscissa indicated the amino acid length of ORFs. The ordinate represented the probability of the predicted transmembrane domain.

Figure 5

Analysis of DEGs related to pollen wall development between soybean CMS lines and their maintainer lines. (A) Heat map of the DEGs related to pollen wall development between SXCMS5A and SXCMS5B. The heat map was created using MeV 4.9 software. Log₂(FC) values were obtained from RNA-seq data in this study. (B) Heat map of the DEGs related to pollen wall development in four different tissues. The color scale represented the relative transcript abundance of the DEGs in four soybean tissues. The heat map was created using MeV 4.9 software. Fragments per kilobase of transcript per million mapped reads (FPKM) values were obtained from RNA-seq data in Phytozome v12.0. R, root; S, stem; L, leaf; F, flower. (C,D) Relative expression level of GmPME (Glyma.02G008300) and GmPL (Glyma.13G064700) between soybean CMS lines and their maintainer lines by qRT-PCR analysis. PME, Pectin methylesterase; PL, pectate lyase. Asterisk indicated statistical significance: ***, p < 0.001.

Figure 6

Analysis of DEGs related to carbohydrate metabolism and sugar transport between soybean CMS lines and their maintainers. (A) Heat map of the DEGs related to starch and sucrose metabolism, pentose and glucuronate interconversions, glycolysis/gluconeogenesis, and sugar transport between SXCMS5A and SXCMS5B. The heat map was created using MeV 4.9 software. Log₂(FC) values were obtained from RNA-seq data. (B–D) Relative expression level of UDP-glucuronic acid decarboxylase 2-like (UDP-GAD2, Glyma.07G246600), exopolygalacturonase (exoPG, Glyma.07G245100) and sugar transport protein 11 (STP11, Glyma.20G103900) between soybean CMS lines and their maintainers by qRT-PCR analysis. (E–G) Soluble sugar, starch, and adenosine triphosphate (ATP) contents analysis between SXCMS5A and SXCMS5B. Asterisk indicated statistical significance: *, p < 0.05; ***, p < 0.001.

Figure 7

Analysis of DEGs related to reactive oxygen species (ROS) metabolism between soybean CMS lines and their maintainers. (A) Heat map of the DEGs related to glutathione metabolism and ascorbate and aldarate metabolism between SXCMS5A and SXCMS5B. The heat map was created using MeV 4.9 software. Log₂(FC) values were obtained from RNA-seq data. (B,C) Relative expression level of Glutathione S-transferase-like (GST, Glyma.02G154400) and L-ascorbate oxidase homolog (L-AO, Glyma.07G225400) between soybean CMS lines and their maintainers by qRT-PCR analysis. (D–F) Activity assays of catalase (CAT), ascorbic acid oxidase (AAO) and glutathione peroxidase (GPX) between SXCMS5A and SXCMS5B. Asterisk indicated statistical significance: *, p < 0.05; ***, p < 0.001.

Figure 8

Analysis of DEGs related to transcription factors (TFs) between soybean CMS lines and their maintainers. (A) Heat map of the DEGs related to TFs between SXCMS5A and SXCMS5B. The heat map was created using MeV 4.9 software. Log₂(FC) values were obtained from RNA-seq data. (B–E) Relative expression level of GmMYB35 (Glyma.06G188400), GmMYB (Glyma.16G218900), GmbHLH118 (Glyma.09G150000) and GmWRKY43 (Glyma.18G238600) between soybean CMS lines and their maintainers by qRT-PCR analysis. Asterisk indicated statistical significance: *, p < 0.05; **, p < 0.01; ***, p < 0.001.

Figure 9

A proposed model for the mechanism of male sterility in soybean CMS line SXCMS5A. The upregulated and downregulated genes or metabolite contents are in red and green backgrounds, respectively.