Proteomic Analysis Identifies Distinct Protein Patterns for High Ovulation in FecB Mutant Small Tail Han Sheep Granulosa Cells

Abstract

The Booroola fecundity (FecB) mutation in the bone morphogenetic protein receptor type 1B (BMPR1B) gene increases ovulation in sheep. However, its effect on follicular maturation is not fully understood. Therefore, we collected granulosa cells (GCs) at a critical stage of follicle maturation from nine wild-type (WW), nine heterozygous FecB mutant (WB), and nine homozygous FecB mutant (BB) Small Tail Han sheep. The GCs of three ewes were selected at random from each genotype and consolidated into a single group, yielding a total of nine groups (three groups per genotype) for proteomic analysis. The tandem mass tag technique was utilized to ascertain the specific proteins linked to multiple ovulation in the various FecB genotypes. Using a general linear model, we identified 199 proteins significantly affected by the FecB mutation with the LIMMA package (p < 0.05). The differential abundance of proteins was enriched in pathways related to cholesterol metabolism, carbohydrate metabolism, amino acid biosynthesis, and glutathione metabolism. These pathways are involved in important processes for GC-regulated 'conservation' of oocyte maturation. Further, the sparse partial least-squares discriminant analysis and the Fuzzy-C-mean clustering method were combined to estimate weights and cluster differential abundance proteins according to ovulation to screen important ovulation-related proteins. Among them, ZP2 and ZP3 were found to be enriched in the cellular component catalog term "egg coat", as well as some apolipoproteins, such as APOA1, APOA2, and APOA4, enriched in several Gene Ontology terms related to cholesterol metabolism and lipoprotein transport. A higher abundance of these essential proteins for oocyte maturation was observed in BB and WB genotypes compared with WW ewes. These proteins had a high weight in the model for discriminating sheep with different FecB genotypes. These findings provide new insight that the FecB mutant in GCs improves nutrient metabolism, leading to better oocyte maturation by altering the abundance of important proteins (ZP2, ZP3, and APOA1) in favor of increased ovulation or better oocyte quality.

Keywords: FecB mutation; Small Tail Han sheep; fertility; follicular granulosa cells; proteome.

Figure 1

Description of the process of collecting granular cells and conducting TMT proteome analysis. A total of 27 ewes, divided into three groups based on their FecB genotype (9 WW, 9 WB, and 9 BB), were subjected to controlled internal drug release (CIDR) to synchronize estrus. Cumulus-oocyte complexes (COCs) were obtained from follicles larger than 3 mm in diameter, 45 h after CIDR removal on day 12. Following the extraction of cumulus granulosa cells, proteins were isolated and randomly distributed into three groups for TMT labeling.

Figure 2

Illustration of the comparison of protein abundance profiles obtained through the employment of PRM and TMT protein quantification techniques. In the comparison of BB ewes with WW ewes (BB vs. WW), BB ewes with WB ewes (BB vs. WB), and WB ewes with WW ewes (WB vs. WW) groups, the following proteins were chosen for validation of their abundance: alpha 2-HS-glycoprotein (AHSG), glutamate dehydrogenase (GLUD1), tubulin beta chain (TUBB), glucosidase II alpha subunit (GANAB), SUN domain-containing protein (SUN2), and thioredoxin domain containing 5 (TXNDC5). The log2(Ratio) values were computed for the compared genotypes. The protein abundance, as determined by the PRM method in comparison to the TMT method, exhibited a consistent pattern for the selected genes across all three groups.

Figure 3

The volcano plots and Venn plots depict the differentially abundant proteins (DAPs). The threshold for identifying upregulated and downregulated DAPs was set at a fold change > 1.2 or <0.83, with a p value < 0.05. The separate analysis of DAPs was performed for the (A) comparison between BB ewes and WW ewes (WW vs. BB), (B) comparison between BB ewes and WB ewes (BB vs. WB), and (C) comparison between WB ewes and WW ewes (WB vs. WW) groups. Additionally, the Venn plot in (D) illustrates the overlap of DAPs between the BB vs. WW, BB vs. WB, and WB vs. WW groups.

Figure 4

The top 30 enriched KEGG pathways and top 30 Gene Ontology (GO) terms in each category of differentially abundant proteins (DAPs) were obtained through the genotype-based factorial model. (A) KEGG enrichment pathways for DAPs. (B) GO terms in each category for DAPs.

Figure 5

Demonstration of the application of sparse partial least-squares discriminant analysis (sPLSDA) combined with the examination of differentially abundant proteins (DAPs) acquired using a genotype-based factorial model. By employing the sPLSDA method, the impact of mutations on protein abundance was predicted, and the weights of these proteins in various genotypes were calculated. Consequently, a screening process was conducted to identify biomarkers affected by FecB mutations, resulting in the identification of two components. (A) The screened proteins exhibited a notable ability to differentiate samples with distinct genotypes. (B) The Receiver Operating Characteristic Area under the Curve (ROC AUC) yields separate results of 1.00 when utilizing proteins from the two components within the predictive model. (C,D) The visualization showcases proteins with higher weights in the two components of the model. The bars represent the absolute values of the calculated loading vector weights. The varying colors indicate the level of association between each biomarker and the genotype. BB refers to sheep with the BB genotype, WB refers to sheep with the WB genotype, and WW refers to sheep with the WW genotype.

Figure 6

The clustering diagram of differentially abundant proteins (DAPs) and the protein-protein interaction network of the proteins in cluster 6. (A) The clustering diagram of DAPs was obtained using the genotype-based factorial model and the fuzzy c-means clustering algorithm. (B) The interaction network of the proteins in cluster 6 was constructed using the STRING database. The edges in the network represent both functional and physical protein associations, with the thickness of the confidence lines indicating the strength of data support.