Morphological Characteristics and Transcriptome Landscapes of Chicken Follicles during Selective Development

Abstract

Ovarian follicle selection largely depends on the transition of granulosa cells from an undifferentiated to a fully differentiated state, which is accompanied by morphological and functional changes in follicles. The processes and transcriptional regulation of follicles during follicle selection are unclear; we thus used follicles from the prehierarchal to the hierarchal stage to investigate histology, reproductive endocrinology, and transcription. The morphology of follicles changed markedly during follicle selection. The numbers of large white, small yellow, and large yellow follicles (LWF, SYF, and LYF, respectively) were 11.83 ± 2.79, 6.83 ± 2.23, and 1.00, respectively, per ovary. LYF showed thicker granulosa cell layers than those of other prehierarchal follicles. Progesterone concentrations were significantly higher in LYF than that in LWF and SYF. In total, 16,823 genes were positively expressed in LWF, SYF, and LYF. Among follicle types, 1290 differentially expressed genes were enriched regarding cell differentiation, blood vessel morphogenesis, and response to steroid hormones. Candidate genes associated with follicle selection participated in the Wnt signaling pathway, steroid hormone biosynthesis, and the TGF-β signaling pathway. We produced insights into crucial morphological characteristics of transcriptional regulation in follicle development. Our results provide an important basis for revealing the mechanism of follicle selection and potential impact on the poultry industry.

Keywords: chicken; follicle selection; granulosa cells; morphological characteristics; transcriptome sequencing.

Figure 1

Ovarian follicles of laying hens. Asterisks indicate prehierarchal follicles; SWF: small white follicle; LWF: large white follicle; SYF: small yellow follicle; F5–F1 represent hierarchal follicles, which are sorted from smallest to largest in diameter. F5 represents the most recently selected follicle (LYF: large yellow follicle).

Figure 2

Number and size of follicles at development stages. (A–C) Diameter and weight of large white follicles (LWF) (A), small yellow follicles (SYF) (B), and large yellow follicles (LYF) (C). Each number on the x-axis represents an individual hen, and each dot represents one follicle. (D,E) Average diameter (D) and weight (E) of follicles around follicle selection. Data represent the mean ± standard error (n = 6), and different letters indicate significant difference (p < 0.05).

Figure 3

Histological observation and progesterone concentration of follicles. (A–C) Paraffin sections of large white follicles (LWF) (A), small yellow follicles (SYF) (B), and large yellow follicles (LYF) (C). Sections of 5 μm thickness were cut and stained with hematoxylin and eosin. The red rectangles indicate the granulosa cell layer, and the red arrows indicate the blood vessels; scale bar = 90 μm. (D) The thickness of granulosa cell layers. (E) The area of single granulosa cells. Indexes were calculated using ImageJ software. (F) Concentration of progesterone in follicles around follicle selection. (G–I) Relative mRNA expression of key genes involved in progesterone synthesis: StAR (G) CYP11A1 (H), and HSD3B1 (I). Values are as mean ± standard error (n = 6); different letters indicate significant difference (p < 0.05).

Figure 4

Experimental design of RNA-seq and differentially expressed gene (DEG) distribution. (A) Schematic illustration of the study workflow. Large white follicles (LWF), small yellow follicles (SYF), and large yellow follicles (F5/LYF) were collected for total RNA extraction and subjected to RNA-seq. (B) Venn diagram of DEGs; the criteria for filtering the DEGs were p_adj < 0.05 and $\left|{\log }_2\mathrm{fold} \mathrm{change}\right|$ ≥ 1.

Figure 5

Validating the RNA-seq data by qRT-PCR. (A) Comparison of ${\log }_2\mathrm{fold} \mathrm{change}$ in 10 differentially expressed genes (DEGs) between qRT-PCR and RNA-seq and (B) regression analysis of $\left|{\log }_2\mathrm{fold} \mathrm{change}\right|$ values between qRT-PCR and RNA-seq. A high ${\mathrm{R}}^2$ indicated that the RNA-seq data were considered to be of a high accuracy.

Figure 6

Differentially expressed genes (DEGs) in large white follicles (LWF) vs. small yellow follicles (SYF). (A) Volcano map of DEGs in LWF vs. SYF. DEGs with p_adj < 0.05 and ${\log }_2\mathrm{fold} \mathrm{change}$ ≥ 1 are marked in red, and DEGs with p_adj < 0.05 and ${\log }_2\mathrm{fold} \mathrm{change}$ ≤ −1 are marked in blue. (B) Gene Ontology functional classification of DEGs in biological process, cellular component, and molecular function categories. (C) Kyoto Encyclopedia of Genes and Genomes pathways of DEGs.

Figure 7

Differentially expressed genes (DEGs) in small yellow follicles (SYF) vs. large yellow follicles (LYF). (A) Volcano map of DEGs in SYF vs. LYF. DEGs with p_adj < 0.05 and ${\log }_2\mathrm{fold} \mathrm{change}$ ≥ 1 are marked in red, and DEGs with p_adj < 0.05 and ${\log }_2\mathrm{fold} \mathrm{change}$ ≤ −1 are marked in blue. (B) Gene Ontology functional classification of DEGs in biological process, cellular component, and molecular function categories. (C) Kyoto Encyclopedia of Genes and Genomes pathways of DEGs.

Figure 8

Differentially expressed genes (DEGs) in large white follicles (LWF) vs. large yellow follicles (LYF). (A) Volcano map of DEGs in LWF vs. LYF. DEGs with p_adj < 0.05 and ${\log }_2\mathrm{fold} \mathrm{change}$ ≥ 1 are marked in red, and DEGs with p_adj < 0.05 and ${\log }_2\mathrm{fold} \mathrm{change}$ ≤ −1 are marked in blue. (B) Gene Ontology functional classification of DEGs in biological process, cellular component, and molecular function categories. (C) Kyoto Encyclopedia of Genes and Genomes pathways of DEGs.

Figure 9

Gene expression dynamics profiles around follicle selection. (A) All mRNA expression profiles. All mRNAs could be clustered into 16 profiles, of which 6 were significant (p < 0.05). (B–D) Three downregulated patterns (profiles 2, 3, and 7, respectively). (E–G) Three upregulated patterns (profiles 13, 12, and 15, respectively).

Figure 10

Signaling pathways involved in follicle selection. (A) Wnt signaling pathway. (B) Steroid hormone biosynthesis. (C) TGF-β signaling pathway. Violin plots show the relative expression levels, ${\log }_2\left[\mathrm{FPKM}+1\right],$ of the ligands and receptors and target differentially expressed genes (DEGs) in each Kyoto Encyclopedia of Genes and Genomes pathway. The purple, green, and orange violins represent LWF, SYF, and LYF, respectively. Flow diagrams show the relationship among these genes.