mRNA and miRNA Transcriptome Profiling of Granulosa and Theca Layers From Geese Ovarian Follicles Reveals the Crucial Pathways and Interaction Networks for Regulation of Follicle Selection

Abstract

Follicle development is characterized by the recruitment, growth, selection, and dominance of follicles, and follicle selection determines the lifetime reproductive performance. However, in birds, the molecular mechanisms underlying follicle selection still remain elusive. This study analyzed genome-wide changes in the mRNA and miRNA expression profiles in both the granulosa and theca layers of geese ovarian follicles before selection (4-6- and 8-10-mm follicles) and after selection (F5). The sequencing results showed that a higher number of both differentially expressed (DE) mRNAs and DE miRNAs were identified between 8-10-mm and F5 follicles compared with those between the 4-6- and 8-10-mm follicles, especially in the granulosa layer. Moreover, a Short Time-series Expression Miner analysis identified a large number of DE mRNAs and DE miRNAs that are associated with follicle selection. The functional enrichment analysis showed that DE genes in the granulosa layer during follicle selection were mainly enriched in five pathways related to junctional adhesion and two pathways associated with lipid metabolism. Additionally, an interaction network was constructed to visualize interactions among protein-coding genes, which identified 53 junctional adhesion- and 15 lipid regulation-related protein-coding genes. Then, a co-expression network between mRNAs and miRNAs in relation to junctional adhesion was also visualized and mainly included acy-miR-2954, acy-miR-218, acy-miR-2970, acy-miR-100, acy-miR-1329, acy-miR-199, acy-miR-425, acy-miR-181, and acy-miR-147. Furthermore, miRNA-mRNA interaction pairs related to lipid regulation were constructed including acy-miR-107, acy-miR-138, acy-miR-130, acy-miR-128, and acy-miR-101 during follicular selection. In summary, these data highlight the key roles of junctional adhesion and lipid metabolism during follicular selection and contribute to a better understanding of the mechanisms underlying follicle selection in birds.

Keywords: Anser cygnoides; follicle selection; granulosa layer; junctional adhesion; lipid metabolism; theca layer; transcriptome profiling.

Figure 1

Experimental design (A) and sampling of geese ovarian follicles (B). Cohorts of follicles include pre-hierarchical (follicles with diameters of 4–6 and 8–10 mm, marked by the red arrowheads and white arrows, respectively) and hierarchical follicles (F5 follicles, marked by red arrows) collected from three individual geese. Both granulosa and theca layers, separated from these follicles, were used for mRNA and small RNA sequencing, followed by a series of bioinformatics analyses. GP4, GP8, and GHF5 represent granulosa layers isolated from 4–6-mm, 8–10-mm, and F5 follicles, respectively. TP4, TP8, and THF5 represent theca layers isolated from the 4–6-mm, 8–10-mm, and F5 follicles, respectively.

Figure 2

mRNA and miRNA transcriptome sequencing and identification of differentially expressed (DE) mRNAs and miRNAs in either granulosa or theca layers. (A, B): Principal component analysis of the mRNA and miRNA transcriptomes among 18 libraries. (C, D) Pearson correlation analysis of the mRNA and miRNA transcriptome among nine granulosa or theca layers. (E, F) Number of DE mRNAs and miRNAs in either granulosa or theca layers. (G–J) Heatmap analysis of DE mRNAs (G, H) and miRNAs (I, J) in granulosa or theca layers. 4–6: follicles with 4–6 mm in diameter; 8–10: follicles with 8–10 mm in diameter; F5: hierarchical F5 follicle. GP4_1, GP4_2, and GP4_3 represent three replicates of GP4, and similar definitions have been used for GP8, GHF5, TP4, TP8, and THF5.

Figure 3

Screening of significantly differentially expressed (DE) mRNAs and miRNAs among either in granulosa or theca layers of geese follicles at different developmental stages. Time-series cluster analysis for the expression profiling of DE mRNAs and miRNAs in the granulosa layer (A, B) and theca layer (C, D). Note that the transition period for follicles from 8–10 mm in diameter to F5 was used for follicle selection. Profiles 2, 3, 5, 7, ii, and iv indicated those DE mRNAs or miRNAs with unchanged expression before follicle selection, following significant fluctuation during follicle selection. Profiles 1, 4, 6, i, and iii represent DE mRNAs or miRNAs with significantly changed expression throughout follicle development.

Figure 4

Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways of differentially expressed (DE) mRNAs from different time-series modules in either granulosa or theca layers. (A, B) Top 20 KEGG pathways of DE mRNAs from profile 3 (A) and profile 2 (B) in the granulosa layer. Five pathways of tight junction, adherens junction, cell adhesion molecules, ECM–receptor interaction, and focal adhesion were identified in profile 3 (A). The two pathways of fat digestion and absorption and glycerolipid metabolism were contained in profile 2 (B). (C, D) Top 20 KEGG pathways of DE genes of profile 5 (C) and profile 7 (D) in the theca layer.

Figure 5

Network analysis of differentially expressed (DE) mRNAs in granulosa and theca layers. (A, B) Co-expression network of DE mRNAs from top 20 Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways of profiles 2 and 3 in the granulosa layer (A), as well as of profiles 5 and 7 in the theca layer (B). This interaction network includes 53 junctional adhesion-related genes (marked in yellow), 15 lipid metabolism-related genes (marked in red), and 13 genes that are involved in ovarian steroidogenesis (marked in blue) (A). Genes marked with triangles were down-regulated, and those marked with circles were up-regulated during follicle selection. Solid lines indicated direct interactions. Dark grey genes had interactions of activation or inhibition, while light grey lines had interactions without activation or inhibition. Activation →; Inhibition.

Figure 6

Co-expression network analysis of differentially expressed (DE) miRNAs and mRNAs in relation to junctional adhesion (A, B) and lipid metabolism (C, D) in the granulosa layer. The interaction networks among 31 miRNAs and 53 mRNAs associated with junctional adhesion (A), and among these miRNAs, acy-miR-2954, acy-miR-2970, acy-miR-1329, acy-miR-218, acy-miR-199, acy-miR-100, acy-miR-181, acy-miR-147, and acy-miR-425 contained 23, 12, 12,11, 6, 6, 5, 5, and 3 protein-coding genes related to junctional adhesion, respectively (B). Seven miRNAs and 13 mRNAs were related to lipid metabolisml (C), and among these miRNAs,: acy-miR-107, acy-miR-138, acy-miR-128, acy-miR-130, and acy-miR-101 interacted with 11, 10, 9, 7, and 7 protein-coding genes implication in lipid regulation, respectively (D).