Corrigendum: Construction of a circRNA–lincRNA–lncRNA–miRNA–mRNA ceRNA regulatory network identifies genes and pathways linked to goat fertility

1 INTRODUCTIONImprovement of reproductive performance is a priority in the goat industry, as it is one of the most important determinants of productivity, sustainability, and profitability. The reproductive cycle involves dynamic and complex ovarian functions, characterized by progressive emergence and development of ovarian follicles (Fatet et al., 2011) under endocrine control. Both granulosa and theca cells are involved in steroidogenesis (Qiu et al., 2013). Granulosa cells (GCs) are one of the most important cell types in ovarian follicles, with crucial roles in follicular development and atresia (Matsuda et al., 2012), especially in the late stages of oocyte development and ovulation. In addition, these cells also control cytoplasmic maturation and have a key role in nuclear maturation by responding to gonadotropins (Mori et al., 2000). Thus, reproductive function in the female is inherently complex, involving various anatomical and physiological processes (Li et al., 2021a). Furthermore, litter size, an important attribute directly related to reproductive efficiency, is controlled by multiple genes and factors (Lai et al., 2016). Hence, knowledge of the genetic basis of reproductive efficiency will provide insights into components controlling ovarian folliculogenesis and fertility in goats (de Lima et al., 2020).The candidate gene approach for fertility has been extensively studied in various livestock species (Miao et al., 2016a,b; Bahrami et al., 2017; Quan et al., 2019). Furthermore, it requires well-developed tools to detect and characterize multiple genes, pathways and networks (Ahlawat et al., 2016; Zhang et al., 2017; Ghafouri et al., 2021, Naserkheil et al., 2022).RNA sequencing (RNA-Seq) has been used to characterize transcripts and differences in gene expression, identify functional genes, and analyze regulatory networks in numerous species (Chen et al., 2015; Li et al., 2021b,c). In addition, single-cell RNA sequencing (scRNA-seq) is being used for mapping and quantifying transcriptional activity at single-cell resolution for all genes in the genome (Islam et al., 2014). It is useful for analysis of cellular heterogeneity, as it can concurrently sequence thousands of cells, as well as discover novel cell types in animals (Choi and Kim, 2019). Conversely, an integrated approach is needed to manage large-scale data generated with high-throughput technologies alongside literature mining. Integrated analyses can combine multilevel views of physiology data into a total interpretation of nonlinear regulatory molecular procedures (La et al., 2019; Sadeghi et al., 2022; Reyhan et al., 2022). Currently, various bioinformatics tools, computational approaches, and algorithms are available to identify interactions and protein functions in regulatory modules in various complex biological networks (Bugrim et al., 2004). In this regard, multi-partite networks such as CircRNA–lincRNA–lncRNA–miRNA–mRNA ceRNA regulatory networks consider various RNAs and have highlighted a new regulatory mechanism of interaction between RNAs. Circular RNAs (CircRNAs) are single-stranded, covalently closed RNA molecules without free 5' and 3' ends that act as transcriptional regulators, microRNAs, and protein templates (Zhou et al., 2020). Long intergenic non-coding RNAs (lincRNAs), RNA transcripts with >200 nucleotides, have major roles in biological processes such as gene expression control, epigenetic control, and scaffold formation (Deniz and Erman, 2017). Long non-coding RNAs (lncRNAs) are non-coding RNA transcripts involved in various biological procedures, such as cell proliferation and transcriptional regulation (Wei et al., 2016). Also, microRNAs (miRNAs), regulatory molecules with 19–25 nucleotides, have vital regulatory roles in multiple biological procedures (cell differentiation and migration, oncogenesis, and apoptosis) by suppressing mRNA (Wang et al., 2015). Therefore, this approach seems well suited to understand molecular regulatory mechanisms in polygenic traits (Hallock and Thomas, 2012).Several studies have identified important candidate genes associated with hormonal regulation of the reproductive cycle and fertility traits in goats (Li et al., 2010; Lai et al., 2016; Ahlawat et al., 2016; Zhang et al., 2018a; Li et al., 2021b). Furthermore, gene ontology and systems biology enable identification of hub genes and co-expression genes with critical roles in fertility (Ahlawat et al., 2016; Zhang et al., 2018b). For instance, in a study comparing high- versus low-fertility goats, many candidate genes were identified in each group. In an analysis of the entire genome of Chinese Laoshan dairy goats, several candidate genes (CCNB2, AR, SMAD2, AMHR2, KDM6A, SOX5, and SYCP2) were associated with both high and low fertility (Lai et al., 2016). Therefore, the purpose of the present study was to examine the ovarian granulosa cell signature genomic regions of Jining Grey goats with high and low fertility, using a public repositories ScRNA-Seq dataset. In addition, we critically reviewed the literature, searching for relevant publications using keywords related to granulosa cells and fertility in goats. Genes with significant differences related to our bioinformatics analyses and candidate gene list extracted from literature mining were compiled, and the 2 gene sets were merged and used to identify protein–protein interaction networks (PPI), and gene regulatory networks (GRNs). Overall, the ceRNA regulatory network, consisting of circRNA–lincRNA–lncRNA–miRNA–mRNA, was constructed based on various interactions to explore effects of functional modules and hub differentially expressed mRNAs, miRNAs, lncRNAs, lincRNAs, and circRNAs on fertility.2 MATERIALS AND METHODSAs an overview, the general workflow for analyzing data collection and methods of identifying key genes, metabolic and signaling pathways, as well as construction of the lncRNA–miRNA–mRNA ceRNA regulatory network and Modules affecting HF and LF in domestic goats (Capra hircus) are presented in Figure 1.2.1 Data collectionSingle-cell RNA-Sequencing (ScRNA-Seq) data from ovarian granulosa cells of high- and low-fertility Ji’Ning grey goats (HF and LF, respectively) were retrieved from the National Center for Biotechnology Information (NCBI) Gene Expression Omnibus (GEO) public database under accession number GSE135897 (www.ncbi.nlm.nih.gov/geo). This dataset was produced using the GPL15473 Illumina HiSeq 2000 platform (Li et al., 2021b). A case-control study approach was designed to identify differentially expressed genes (DEGs) between HF and LF goats. A total of 30 samples from Ji'ning gray goats (first group: 15 with high litter size (HF; ≥3 offspring) and second group: 15 with low litter size (LF; ≤2 offspring)) were analyzed. The Ji’Ning grey goat is a local breed and the oldest domesticated goat species with high fertility in China that has year-round estrus and mean litter size of 2.94 (Huang et al., 2012; Miao et al., 2016a,b). All ovarian follicles from each group of these Ji’Ning grey goats (from small (<3 mm) to large (˃7 mm)) were collected. Then, oocytes and GCs were mechanically separated by repeated pipetting. Oocytes from high- and low-fertility groups were labeled. This study was performed in accordance with ARRIVE guidelines. Library preparation and sequencing were performed separately for oocytes and GCs. Details regarding animal ethics approval, GCs, total RNA extraction, and sequencing and validation of RNA-Seq data have been reported (Zhao et al., 2020; Li et al., 2021b). 2.2 Quality control and detection of differentially expressed genesFirst, quality of the raw RNA sequences was assessed using FastQC software v0.11.5 (Andrew, 2010) and FastQ Groomer software v1.1.5 (Blankenberg et al., 2010); thereafter, these sequences were pre-processed with Trimmomatic software v0.38.0 (Bolger et al., 2014) to remove adapters, low-quality reads, and PCR primers. Alignment sequences, mapping, and identification of known and novel RNAs of reads were related to the reference genome of Capra hircus (https://ftp.ensembl.org/pub/release-108/fasta/capra_hircus/dna/) using HISAT2 software v2.2.1 with default parameters to determine number of aligned and unaligned reads (Kim et al., 2015). Regarding transcript quantification, total raw counts of mapped reads were calculated using featureCounts software (v2.0.1) (Liao et al., 2014). Subsequently, to examine whether accumulation or degradation of transcripts was related to fertility, transcripts and their expression levels were compared between ovarian granulosa cell samples of high- and low-fertility Ji’Ning grey goats. Differences in gene expression were detected from reading counts using DESeq2 software (v2.11.40.7) (Love et al., 2014). The threshold for statistical significance of the differential expression of each gene was obtained with the criteria of log2FC (fold change <5 and >5) and FDR <0.05 (false discovery rate). 2.3 Literature mining to identify candidate genes for fertility in goatWe reviewed the literature up to March 12, 2023, searching for relevant publications using keywords related to granulosa cells and fertility in goats. Regulatory RNAs (i.e., circRNAs, lincRNAs, lncRNAs, miRNAs, and mRNAs) with significant differences related to candidate RNAs list extracted from literature mining were compiled as RNA sets 2-6, respectively. Finally, RNA set 1 (from our bioinformatics analyses) and RNA sets 2-6 (from literature mining) were merged and used as input files for target prediction tools, STRING website, and Cytoscape software to identify protein-protein interactions gene regulatory networks and reconstruct the circRNA–lincRNA–lncRNA–miRNA–mRNA ceRNA regulatory network, as well as modules.2.4 Gene ontology and functional enrichment analysisGene set annotation and enrichment analysis used DAVID (the Database for Annotation, Visualization, and Integrated Discovery; https://david.ncifcrf.gov/) v6.8 (Sherman and Lempicki, 2009), g:Profiler (https://biit.cs.ut.ee/gprofiler/gost) (Reimand et al., 2016), GeneCards (https://www.genecards.org/), and STRING database v11.0 (https://string-db.org) (Szklarczyk et al., 2019) to determine potential functions and metabolic pathways. Genes were assigned to functional categories using the Gene Ontology (GO) database under biological process (BP), molecular function (MF), and cellular component (CC).2.5 Target prediction of differentially expressed mRNAs and other types of regulatory RNAsFunctional annotation of types of regulatory RNAs i.e., circRNAs, lincRNAs, lncRNAs, and miRNAs consisted of functional annotation of their potential target mRNA genes. Predicted targeted genes and types of regulatory RNAs were predicted using: miRBase (Kozomara and Birgaoanu, 2019) (https://www.mirbase.org/); Targetscan (Grimson et al., 2007); miRanda (http://www.microrna.org/); miRWalk 3.0 (a comprehensive atlas of microRNA–target interaction tools that integrates 12 miRNA–target prediction tools; http://mirwalk.umm.uni-heidelberg.de/); NONCODE database (Volders et al., 2019) (http://www.noncode.org/); LNCipedia database (Bader et al., 2006) (https://lncipedia.org); and CircInteractome web tool (Dudekula et al., 2016) (a computational tool that enables prediction and mapping of binding sites for RBPs and miRNAs on reported circRNAs; https://circinteractome.nia.nih.gov/). Identified target genes were selected and submitted to DAVID, KEGG, Reactome pathways, and the PANTHER database for enrichment and validation of target genes for each type of RNA.2.6 CircRNA–lincRNA–lncRNA–miRNA–mRNA ceRNA regulatory network constructionBased on the ceRNA theory, global functions for all non-coding RNAs can serve as an endogenous “sponge” to regulate upregulated or downregulated circRNAs, lncRNAs, lincRNAs, miRNAs or mRNAs that are inverse relationships together in the mRNA-mRNA, mRNA-miRNA, mRNA-lncRNA, mRNA-lincRNA, mRNA-circRNA, miRNA-lncRNA, and miRNA-circRNA interaction pairs chosen to construct the circRNA–lincRNA–lncRNA–miRNA–mRNA ceRNA regulatory network (Xu et al., 2019). In this regard, protein-protein interaction (PPI) network analysis was performed using the STRING database v11.0 (Szklarczyk et al., 2019) (Search Tool for the Retrieval of Interacting Genes or Proteins; https://string-db.org), BIND (Biomolecular Interaction Network Database) (Bader et al., 2003), MIPS (Mammalian Protein-Protein Interactions Database) (Pagel et al., 2005), and BioGRID (Biological General Repository for Interaction Datasets) et al., to explore interactions between genes in Capra identifying interactions between types of regulatory RNAs and gene expression data the circRNA–lincRNA–lncRNA–miRNA–mRNA ceRNA regulatory network was and using Cytoscape software et al., National of General Also, statistical and significance of the network was assessed with the Network in Cytoscape of circRNA–lincRNA–lncRNA–miRNA–mRNA ceRNA regulatory network and identification of hub regulatory or have a significant role in the ceRNA regulatory network, as a set of with functions that biological as functional and of the circRNA–lincRNA–lncRNA–miRNA–mRNA ceRNA regulatory network, the Cytoscape et al., and et al., algorithms to and was is a to discover of a network by between and a Also, the can used for or parameters and the statistical and significance of the ceRNA regulatory modules were calculated using the Cytoscape Network with the default for the network et al., AND ScRNA-Seq analysis to identify differentially expressed genes (DEGs) In this we the of of ovarian granulosa cells in goats. provide insights into the genetic basis of fertility in Ji'ning gray goats, gene expression analyses using the ScRNA-Seq dataset with number GSE135897 from the were The selected gene expression 30 15 samples of granulosa cells from goats and 15 samples of granulosa cells from low-fertility goats. this these 30 samples were into 2 groups and to of gene and identify significant A total of significant genes were identified by the expression of granulosa cells of high- versus low-fertility goats. Finally, by the expression change threshold change <5 and FDR genes were differentially expressed in granulosa cells from goats with high versus low these genes, genes were and were many studies in molecular and biological systems have been to discover candidate genes and identify molecular pathways involved in fertility (Ahlawat et al., 2020; et al., Li et al., 2021b). identified genes were related to reproductive functions and ovarian follicular and of involved in or their expression at various can to complex biological processes (Wang et al., 2019). In this regard, in a study using RNA-Seq analysis for ovarian of and goats, genes and were identified as important roles in goat et al., 2019). Also, study that in goats have key roles in development and cell ovarian development (Li et al., Literature for identified and types of regulatory mining for that were in our this list of to our a platform for Gene Ontology (GO) and functional enrichment Identified based on literature mining and their role in fertility in goats are as RNA set 2 Also, literature mining for and types of regulatory i.e., circRNAs, lincRNAs, lncRNAs, and miRNAs with significant compiled as RNA sets The gene list in was merged with the gene list in and used to identify intergenic interactions and reconstruct the protein-protein interaction Then, identifying interactions between types of regulatory RNAs in RNA sets together and with gene expression the circRNA–lincRNA–lncRNA–miRNA–mRNA ceRNA regulatory network was Gene ontology and enrichment functional annotation of was detected based on the biological process (BP), molecular function (MF), and cellular component to identify functions and metabolic pathways as well as of the merged list of genes sets 1 and using the DAVID, and g:Profiler of the of the for high- and low-fertility goats are presented in Figure Identified were involved in the genes from roles in cellular regulation of biological to cellular to cell regulation of cellular metabolic cellular component or reproductive signaling signaling to and protein complex for biological processes The gene genes SMAD2, SOX5, AR, and were involved in most of the molecular function that can significant genes. molecular functions were such as protein signaling molecular function and were the most significant functions associated with fertility Regarding cellular complex, and were identified In addition, analysis for identified was performed using DAVID, and The identified involved in fertility were in the signaling pathways of cells, ovarian oocyte oocyte and and and and signaling pathways female reproductive and genes are in functions of cycle ovarian follicular process involved in and cell Hence, to genes involved in the identified pathways, and were candidate genes for reproductive functions, and cell differentiation in goats. many studies have the gene as one of the genes controlling reproductive function and fertility in small especially goats et al., Ahlawat et al., 2014). In addition, the gene has roles in and maturation of follicles and reproductive function in goats and (Chen et al., 2017). study that is involved in differential expression of ovarian mRNA genes in goat fertility et al., The gene and and the genes involved in and It has also been reported that expression of this gene was in and in and 2020). the role of the gene in especially in and differentiation of ovarian follicles and it can a candidate gene for reproductive function and fertility et al., 2013). The gene a complex of that are for formation and of cell by cell and Also, have an role in and et al., 2018b). Therefore, our regarding genes and were with other and of circRNA–lincRNA–lncRNA–miRNA–mRNA ceRNA regulatory et presented the endogenous RNA et al., The ceRNA regulatory networks have a new mechanism of interaction between RNAs and have crucial roles in multiple biological processes et al., et al., et al., 2019). In this regard, many studies have been to the ceRNA roles of non-coding RNAs in important traits by endogenous RNA networks et al., et al., 2020). detect the mechanism of non-coding RNAs regulate mRNA a ceRNA regulatory network was constructed with a of predicted mRNA-miRNA, mRNA-lncRNA, mRNA-lincRNA, mRNA-circRNA, miRNA-lncRNA, and miRNA-circRNA interaction The ceRNA regulatory network for and and types of regulatory between 2 or protein molecules related to functions, is presented in Figure on knowledge of this ceRNA regulatory network consisted of and and the associated files with the networks are in in for In circRNAs, lincRNAs, lncRNAs, miRNAs, and mRNAs were in the network molecular species lincRNAs, lncRNAs, mRNAs, and in constructed networks are as and interactions between as constructed networks were in a interaction using to Cytoscape for parameters of the ceRNA regulatory network and modules such as the number of number of and network were to examine the of and of a with other of as presented in In this ceRNA regulatory network, genes SMAD2, AMHR2, AR, and the most interactions with other genes in the these hub genes, genes and were involved in molecular function and biological in with other studies et al., et al., Zhang et al., 2018a; 2020). the miRNAs, and most of the identified hub genes as of the selected Also, of the i.e., and with and hub genes, respectively. Conversely, the miRNAs such as and with most of the involved in ceRNA regulatory network and were as hub miRNAs that with this of the circRNA–lincRNA–lncRNA–miRNA–mRNA ceRNA regulatory network was performed using et al., and et al., to algorithms to determine significant or modules by an integrated were candidate modules involved in goat the interactions of each of is in 1 consisted of and and in 2 lincRNAs, 12 miRNAs, and In this SMAD2, and were hub genes. Also, and as hub miRNAs most of the involved genes such as SMAD2, KDM6A, and in this and with and respectively. All genes in this were involved in the pathways that were 2 consisted of and and in 2 lincRNAs, miRNAs, and In this SMAD2, and were hub genes. Also, and as hub miRNAs most of the genes such as SMAD2, and the gene was other genes involved in the in this and with and SMAD2, Furthermore, consisted of and and in 1 circRNAs, lncRNAs, miRNAs, and In this and the most interaction with other genes as hub genes. Also, and as hub miRNAs genes KDM6A, and most of the genes involved in the in this and circRNAs with most miRNAs such as and Finally, consisted of and and in lncRNAs, 2 miRNAs, and In this and the most interaction with other genes as hub genes. Also, all genes involved in this were by and with all of the i.e., and involved in the on the literature, genes related to reproductive function in dairy goats, AR, SMAD2, AMHR2, and genes, were selected in goats with high fertility (Lai et al., 2016; et al., 2020). Therefore, our of the between hub genes SMAD2, AMHR2, and and female reproductive functions in goats. It was reported that the gene is a involved in signaling et al., the gene is a protein that involves a binding and it has a key role in by et al., 2020). is a gene that has a major role in and protein activity et al., 2022). and genes were associated with signaling pathways of cells, and signaling pathways (Li et al., 2021b). The SMAD2, and genes have critical roles in and differentiation of ovarian cells, with of (Li et al., et al., 2014). the gene is associated with of and protein This gene is also involved in and development of ovarian follicles in and goats et al., Furthermore, genes and as candidate genes, are of the directly related to and and development of ovarian follicles in and goats et al., 2013). The gene is involved in ovarian and as well as stages of ovarian et al., is a critical gene in activity and in this gene signaling of differentiation and et al., 2020). The gene is directly involved in in the control of reproductive functions, ovarian follicular and cell and et al., The gene has been detected in various and in various especially small This is involved in many endocrine and is for reproductive function et al., Therefore, to the functions of hub genes identified in the circRNA–lincRNA–lncRNA–miRNA–mRNA ceRNA regulatory network and we that these genes have important roles in reproductive performance and fertility of goats. In that regard, are involved in endocrine cell as well as and and selected in to In addition, most genes involved in the ceRNA regulatory network signaling pathways of cells, ovarian and functions of the involved genes, especially hub and metabolic pathways by the genes involved in the circRNA–lincRNA–lncRNA–miRNA–mRNA ceRNA regulatory network and modules are presented in 2 and The signaling pathways of of cells, and are by ceRNA regulatory network and The signaling a large group of related and differentiation et al., pathways of cells are by cells have potential to all cell It is that cells are from the cell of et al., 2020). The signaling is involved in cell proliferation and and is a gene in a This gene has a major role in the signaling that various cellular processes such as and cell cycle in to a of (Miao et al., metabolic pathways such as ovarian and and are also in these with important roles in and In this a computational approach with a circRNA–lincRNA–lncRNA–miRNA–mRNA ceRNA regulatory network was performed using predicted and expression of RNAs. differential expression in various especially ovarian follicles, the potential roles of types of RNAs in the transcriptional and regulation of genes involved in A for in our were differences in molecular and gene differences in ovarian of and bioinformatics of the present studies the of a comprehensive dataset with and ovarian from goat In we that identified are important to understand functional pathways involved in fertility in female goats. are needed to the biological functional types of RNAs in and our integrated circRNAs, lincRNAs, lncRNAs, miRNAs, and mRNAs based on an integrated approach from bioinformatics analyses and literature mining to construct ceRNA regulatory this can a approach to detect insights into biological will needed to our study used a novel approach to various types of RNA as an integrated network in goat of single-cell RNA sequencing data in identification of in goats with high versus low genes were and were circRNAs, lincRNAs, lncRNAs, and miRNAs, all types of regulatory were obtained from literature mining. circRNA–lincRNA–lncRNA–miRNA–mRNA ceRNA regulatory network was constructed and these identified RNAs were associated with transcriptional regulatory and binding in of molecular functions, as well as reproductive functions such as ovarian and differentiation cells based on biological Furthermore, our are a to molecular networks and the functions of ovarian follicular development and of the genetic basis of high- versus low-fertility goats.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.