Genome-wide identification of RNA editing in seven porcine tissues by matched DNA and RNA high-throughput sequencing

Role of ADAR1 on Proliferation and Differentiation in Porcine Preadipocytes

Recent research has identified ADAR1 as a participant in the regulation of lipid accumulation in mice. However, there are no reports on the roles of ADAR1 in proliferation, apoptosis and differentiation of porcine preadipocytes. In this study, we investigated the role of ADAR1 in differentiation, proliferation and apoptosis of porcine preadipocytes using CCK-8, EdU staining, cell cycle detection, RT-qPCR, Western blot, a triglyceride assay and Oil Red O staining. The over-expression of ADAR1 significantly promoted proliferation but inhibited the differentiation and apoptosis of porcine preadipocytes. The inhibition of ADAR1 had the opposite effect on the proliferation, differentiation and apoptosis of porcine preadipocytes with over-expressed ADAR1. Then, the regulation mechanisms of ADAR1 on preadipocyte proliferation were identified using RNA-seq, and 197 DEGs in response to ADAR1 knockdown were identified. The MAPK signaling pathway is significantly enriched, indicating its importance in mediating fat accumulation regulated by ADAR1. The study's findings will aid in uncovering the mechanisms that regulate fat accumulation through ADAR1.

Imprinting at the KBTBD6 locus involves species-specific maternal methylation and monoallelic expression in livestock animals

BACKGROUND

The primary differentially methylated regions (DMRs) which are maternally hypermethylated serve as imprinting control regions (ICRs) that drive monoallelic gene expression, and these ICRs have been investigated due to their implications in mammalian development. Although a subset of genes has been identified as imprinted, in-depth comparative approach needs to be developed for identification of species-specific imprinted genes. Here, we examined DNA methylation status and allelic expression at the KBTBD6 locus across species and tissues and explored potential mechanisms of imprinting.

RESULTS

Using whole-genome bisulfite sequencing and RNA-sequencing on parthenogenetic and normal porcine embryos, we identified a maternally hypermethylated DMR between the embryos at the KBTBD6 promoter CpG island and paternal monoallelic expression of KBTBD6. Also, in analyzed domesticated mammals but not in humans, non-human primates and mice, the KBTBD6 promoter CpG islands were methylated in oocytes and/or allelically methylated in tissues, and monoallelic KBTBD6 expression was observed, indicating livestock-specific imprinting. Further analysis revealed that these CpG islands were embedded within transcripts in porcine and bovine oocytes which coexisted with an active transcription mark and DNA methylation, implying the presence of transcription-dependent imprinting.

CONCLUSIONS

In this study, our comparative approach revealed an imprinted expression of the KBTBD6 gene in domesticated mammals, but not in humans, non-human primates, and mice which implicates species-specific evolution of genomic imprinting.

Characterization of A-to-I Editing in Pigs under a Long-Term High-Energy Diet

Long-term high-energy intake has detrimental effects on pig health and elevates the risk of metabolic disease. RNA editing modifying RNA bases in a post-transcriptional process has been extensively studied for model animals. However, less evidence is available that RNA editing plays a role in the development of metabolic disorders. Here, we profiled the A-to-I editing in three tissues and six gut segments and characterized the functional aspect of editing sites in model pigs for metabolic disorders. We detected 64,367 non-redundant A-to-I editing sites across the pig genome, and 20.1% correlated with their located genes' expression. The largest number of A-to-I sites was found in the abdominal aorta with the highest editing levels. The significant difference in editing levels between high-energy induced and control pigs was detected in the abdominal aorta, testis, duodenum, ileum, colon, and cecum. We next focused on 6041 functional A-to-I sites that detected differences or specificity between treatments. We found functional A-to-I sites specifically involved in a tissue-specific manner. Two of them, located in gene SLA-DQB1 and near gene B4GALT5 were found to be shared by three tissues and six gut segments. Although we did not find them enriched in each of the gene features, in correlation analysis, we noticed that functional A-to-I sites were significantly enriched in gene 3'-UTRs. This result indicates, in general, A-to-I editing has the largest potential in the regulation of gene expression through changing the 3'-UTRs' sequence, which is functionally involved in pigs under a long-term high-energy diet. Our work provides valuable knowledge of A-to-I editing sites functionally involved in the development of the metabolic disorder.

On the origin and evolution of RNA editing in metazoans

Extensive adenosine-to-inosine (A-to-I) editing of nuclear-transcribed mRNAs is the hallmark of metazoan transcriptional regulation. Here, by profiling the RNA editomes of 22 species that cover major groups of Holozoa, we provide substantial evidence supporting A-to-I mRNA editing as a regulatory innovation originating in the last common ancestor of extant metazoans. This ancient biochemistry process is preserved in most extant metazoan phyla and primarily targets endogenous double-stranded RNA (dsRNA) formed by evolutionarily young repeats. We also find intermolecular pairing of sense-antisense transcripts as an important mechanism for forming dsRNA substrates for A-to-I editing in some but not all lineages. Likewise, recoding editing is rarely shared across lineages but preferentially targets genes involved in neural and cytoskeleton systems in bilaterians. We conclude that metazoan A-to-I editing might first emerge as a safeguard mechanism against repeat-derived dsRNA and was later co-opted into diverse biological processes due to its mutagenic nature.

The Profiles and Functions of RNA Editing Sites Associated with High-Altitude Adaptation in Goats

High-altitude environments dramatically influenced the genetic evolution of vertebrates. However, little is known about the role of RNA editing on high-altitude adaptation in non-model species. Here, we profiled the RNA editing sites (RESs) of heart, lung, kidney, and longissimus dorsi muscle from Tibetan cashmere goats (TBG, 4500 m) and Inner Mongolia cashmere goats (IMG, 1200 m) to reveal RNA editing-related functions of high-altitude adaptation in goats. We identified 84,132 high-quality RESs that were unevenly distributed across the autosomes in TBG and IMG, and more than half of the 10,842 non-redundant editing sites were clustered. The majority (62.61%) were adenosine-to-inosine (A-to-I) sites, followed by cytidine-to-uridine (C-to-U) sites (19.26%), and 32.5% of them had a significant correlation with the expression of catalytic genes. Moreover, A-to-I and C-to-U RNA editing sites had different flanking sequences, amino acid mutations, and alternative splicing activity. TBG had higher editing levels of A-to-I and C-to-U than IMG in the kidney, whereas a lower level was found in the longissimus dorsi muscle. Furthermore, we identified 29 IMG and 41 TBG population-specific editing sites (pSESs) and 53 population-differential editing sites (pDESs) that were functionally involved in altering RNA splicing or recoding protein products. It is worth noting that 73.3% population-differential, 73.2% TBG-specific, and 80% IMG-specific A-to-I sites were nonsynonymous sites. Moreover, the pSESs and pDESs editing-related genes play critical functions in energy metabolisms such as ATP binding molecular function, translation, and adaptive immune response, which may be linked to goat high-altitude adaptation. Our results provide valuable information for understanding the adaptive evolution of goats and studying plateau-related diseases.

The AIRN lncRNA is imprinted and paternally expressed in pigs

Genomic imprinting plays critical roles during the development of mammalian species and underlying epigenetic mechanisms frequently involve long non-coding RNAs (lncRNAs). The paternal transcription of the antisense Igf2r RNA noncoding (Airn) is responsible for paternal silencing of the mouse insulin-like growth factor 2 receptor (Igf2r) gene and maternal Igf2r expression. Although the corresponding maternal DNA methylation imprint is conserved in humans and pigs, the orthologous AIRN lncRNA has been identified in humans but not in pigs. Here, we aimed to examine imprinted allelic expression of the porcine AIRN lncRNA along with a corresponding differentially methylated region (DMR) and to analyze allelic expression of AIRN and IGF2R in pigs. By comparing parthenogenetic and control porcine embryos, we identified a maternally methylated DMR and a significantly higher expression of AIRN lncRNA in control embryos (P < 0.05) indicating its paternal expression. Further analyses revealed that the expression of AIRN lncRNA was enriched in the pig brain and its subregions, and it was monoallelically expressed; whereas, IGF2R was expressed biallelically suggesting an absence of allele-specific transcriptional regulation. Our findings will lead to further investigations into the role of the imprinted porcine AIRN lncRNA during pig development.

Granulosa Cell Specific Loss of Adar in Mice Delays Ovulation, Oocyte Maturation and Leads to Infertility

Adenosine deaminases acting on RNA-(ADAR) comprise one family of RNA editing enzymes that specifically catalyze adenosine to inosine (A-to-I) editing. A granulosa cell (GC) specific Adar depleted mouse model [Adar flox/flox:Cyp19a1-Cre/+ (gcAdarKO)] was used to evaluate the role of ADAR1 during the periovulatory period. Loss of Adar in GCs led to failure to ovulate at 16 h post-hCG, delayed oocyte germinal vesicle breakdown and severe infertility. RNAseq analysis of GC collected from gcAdarKO and littermate control mice at 0 and 4 h post-hCG following a super-ovulatory dose of eCG (48 h), revealed minimal differences after eCG treatment alone (0 h), consistent with normal folliculogenesis observed histologically and uterine estrogenic responses. In contrast, 300 differential expressed genes (DEGs; >1.5-fold change and FDRP < 0.1) were altered at 4 h post-hCG. Ingenuity pathway analysis identified many downstream targets of estrogen and progesterone pathways, while multiple genes involved in inflammatory responses were upregulated in the gcAdarKO GCs. Temporal expression analysis of GCs at 0, 4, 8, and 12 h post-hCG of Ifi44, Ifit1, Ifit3b, and Oas1g and Ovgp1 confirmed upregulation of these inflammatory and interferon genes and downregulation of Ovgp1 a glycoprotein involved in oocyte zona pellucida stability. Thus, loss of ADAR1 in GCs leads to increased expression of inflammatory and interferon response genes which are temporally linked to ovulation failure, alterations in oocyte developmental progression and infertility.

Loss of Monoallelic Expression of IGF2 in the Adult Liver Via Alternative Promoter Usage and Chromatin Reorganization

In mammals, genomic imprinting operates via gene silencing mechanisms. Although conservation of the imprinting mechanism at the H19/IGF2 locus has been generally described in pigs, tissue-specific imprinting at the transcript level, monoallelic-to-biallelic conversion, and spatio-temporal chromatin reorganization remain largely uninvestigated. Here, we delineate spatially regulated imprinting of IGF2 transcripts, age-dependent hepatic mono- to biallelic conversion, and reorganization of topologically associating domains at the porcine H19/IGF2 locus for better translation to human and animal research. Whole-genome bisulfite sequencing (WGBS) and RNA sequencing (RNA-seq) of normal and parthenogenetic porcine embryos revealed the paternally hypermethylated H19 differentially methylated region and paternal expression of IGF2. Using a polymorphism-based approach and omics datasets from chromatin immunoprecipitation sequencing (ChIP–seq), whole-genome sequencing (WGS), RNA-seq, and Hi-C, regulation of IGF2 during development was analyzed. Regulatory elements in the liver were distinguished from those in the muscle where the porcine IGF2 transcript was monoallelically expressed. The IGF2 transcript from the liver was biallelically expressed at later developmental stages in both pigs and humans. Chromatin interaction was less frequent in the adult liver compared to the fetal liver and skeletal muscle. The duration of genomic imprinting effects within the H19/IGF2 locus might be reduced in the liver with biallelic conversion through alternative promoter usage and chromatin remodeling. Our integrative omics analyses of genome, epigenome, and transcriptome provided a comprehensive view of imprinting status at the H19/IGF2 cluster.

Preliminary identification and analysis of differential RNA editing between higher and lower backfat thickness pigs using DNA-seq and RNA-seq data

RNA editing is an essential post-transcriptional regulatory mechanism. However, few studies have investigated the functional RNA edits in the economic traits of livestock on a genome-wide scale. Pigs are one of the most important livestock species and their fat is the principal organ involved in the regulation of adipose deposition. Here, we used three full-sibling pairs, with each pair comprising a pig with higher backfat (BF) thickness and lower backfat thickness, to identify RNA editing events based on whole-genome and transcriptome sequencing data. A total of 60,903 non-redundant RNA editing sites with 59,472 (97.7%) A-to-G edits were detected using a revised bioinformatics pipeline. A specific sequence context with G preference was found one base downstream of the edited site, and the editing level was associated with the distribution of nucleotides across nearly sites. Moreover, the A-to-G editing sites mostly occurred in the pig-special short interspersed nuclear elements, Pre0_SS. Comparing the difference between pigs with higher BF and lower BF, we found 211 differentially edited sites (DESites). Functional enrichment analyses revealed a significant enrichment of genes containing DESites in terms of adipose deposition. The DESites located in the six adipose-related genes (SKP1, GSK3B, COL5A3, MDM4, NT5C2, and DENND2A) were selected as candidate RNA editing sites associated with adipose deposition, and thus require further evaluation. This study mined the potentially functional RNA editing sites in pig adipose tissue and indicated that RNA editing may play an important role in adipose deposition, which provides a new insight into the post-transcriptionally mediated regulation mechanism of fat development.

Landscape of adenosine-to-inosine RNA recoding across human tissues

RNA editing by adenosine deaminases changes the information encoded in the mRNA from its genomic blueprint. Editing of protein-coding sequences can introduce novel, functionally distinct, protein isoforms and diversify the proteome. The functional importance of a few recoding sites has been appreciated for decades. However, systematic methods to uncover these sites perform poorly, and the full repertoire of recoding in human and other mammals is unknown. Here we present a new detection approach, and analyze 9125 GTEx RNA-seq samples, to produce a highly-accurate atlas of 1517 editing sites within the coding region and their editing levels across human tissues. Single-cell RNA-seq data shows protein recoding contributes to the variability across cell subpopulations. Most highly edited sites are evolutionary conserved in non-primate mammals, attesting for adaptation. This comprehensive set can facilitate understanding of the role of recoding in human physiology and diseases.

Gabay et al. provide a highly-accurate atlas of recoding by A-to-I RNA editing in human, profiled across tissues and cell subpopulations. Most highly edited sites are evolutionary conserved in non-primate mammals, attesting for adaptation.

Genomic Imprinting at the Porcine PLAGL1 Locus and the Orthologous Locus in the Human

Implementation of genomic imprinting in mammals often results in cis-acting silencing of a gene cluster and monoallelic expression, which are important for mammalian growth and function. Compared with widely documented imprinting status in humans and mice, current understanding of genomic imprinting in pigs is relatively limited. The objectives of this study were to identify DNA methylation status and allelic expression of alternative spliced isoforms at the porcine PLAGL1 locus and assess the conservation of the locus compared to the orthologous human locus. DNA methylome and transcriptome were constructed using porcine parthenogenetic or biparental control embryos. Using methylome, differentially methylated regions between those embryos were identified. Alternative splicing was identified by differential splicing analysis, and monoallelic expression was examined using single nucleotide polymorphism sites. Moreover, topological boundary regions were identified by analyzing CTCF binding sites and compared with the boundary of human orthologous locus. As a result, it was revealed that the monoallelic expression of the PLAGL1 gene in porcine embryos via genomic imprinting was maintained in the adult stage. The porcine PLAGL1 locus was largely conserved in regard to maternal hypermethylation, tissue distribution of mRNA expression, monoallelic expression, and biallelic CTCF-binding, with exceptions on transcript isoforms produced by alternative splicing instead of alternative promoter usage. These findings laid the groundwork for comparative studies on the imprinted PLAGL1 gene and related regulatory mechanisms across species.

Genome-Wide Identification of RNA Editing Sites Affecting Intramuscular Fat in Pigs

Intramuscular fat (IMF) is essential for improving the palatability and flavor of meat, and it is strongly associated with human insulin resistance. RNA editing is a widespread regulating event in different tissues. Here, we investigated the global RNA editing difference of two groups of pig with different IMF contents to find the potential editing sites affecting IMF. In this research, RES-Scanner and REDItools were used to identify RNA editing sites based on the whole genome and transcriptome sequencing data of the high and low groups composed of three full-sib pairs with opposite IMF phenotypes. A total of 295 RNA editing sites were investigated in the longissimus dorsi muscle, and 90.17% of these sites caused A to G conversion. After annotation, most editing sites were located in noncoding regions (including five sites located on the 3' UTR regions). Five editing sites (including two sites that could lead to nonsynonymous amino acid changes) were located in the exons of genes. A total of 36 intergroup (high and low IMF) differential RNA editing sites were found in 33 genes. Some candidate editing sites, such as sites in acyl-coenzymeA: cholesterol acyltransferase 1 (ACAT1), coatomer protein, subunit alpha (COPA), and nuclear receptor coactivator 3 (NCOA3), were selected as candidate RNA editing sites associated with IMF. One site located on the 3' UTR region of growth hormone secretagogue receptor (GHSR) may regulate GHSR expression by affecting the interaction of miRNA and mRNA. In conclusion, we identified a total of 36 nonredundant RNA editing sites in the longissimus dorsi muscle, which may reveal the potential importance of RNA editing in IMF. Four were selected as candidate sites associated with IMF. Our findings provide some new insights of RNA editing function in pig longissimus dorsi muscle which useful for pig IMF breeding or human insulin resistances research.

Developmental stage-specific A-to-I editing pattern in the postnatal pineal gland of pigs (Sus scrofa)

Background

RNA editing is a widespread post-transcriptional modification mechanism in mammalian genomes. Although many editing sites have been identified in domestic pigs (Sus scrofa), little is known about the characteristics and dynamic regulation of RNA editing in the pineal gland (PG), a small neuroendocrine gland that synthesizes and secretes melatonin, which is primarily responsible to modulate sleep patterns.

Results

This study analyzed the expression of adenosine-to-inosine (A-to-I) editing regulators and profiled the first dynamic A-to-I RNA editome during postnatal PG development. The results identified ADAR1 as the most abundantly expressed ADAR enzyme, which was down-regulated during postnatal PG development. Furthermore, 47,284 high-confidence RNA editing sites were identified, the majority of which (93.6%) were of the canonical A-to-I editing type, followed by C-to-T editing. Analysis of its characteristics showed that the A-to-I editing sites mostly localized in SINE retrotransposons PRE-1/Pre0_SS. Moreover, a strong deficiency and preference for guanine nucleotides at positions of one base upstream or downstream were found, respectively. The overall editing level at the puberty stage was higher than at both infancy and adulthood stages. Additionally, genome-wide RNA editing was found to exhibit a dynamic stage-specific fashion (postnatally). Genes that underwent developmental changes in RNA editing were associated with catabolic processes as well as protein localization and transport functions, implying that RNA editing might be responsible for the molecular machineries of the postnatal developing PG. Remarkably, RNA editing in 3'-UTRs might regulate gene expression by influencing miRNA binding during PG development.

Conclusions

This study profiles the first comprehensive developmental RNA editome in the pig PG, which contributes to the understanding of the importance of post-transcriptionally mediated regulation during mammalian postnatal PG development. Moreover, this study widely extends RNA editome resources in mammals.

Sex-Dependent RNA Editing and N6-adenosine RNA Methylation Profiling in the Gonads of a Fish, the Olive Flounder (Paralichthys olivaceus)

Adenosine-to-inosine (A-to-I) editing and N6-methyladenosine (m6A) are two of the most abundant RNA modifications. Here, we examined the characteristics of the RNA editing and transcriptome-wide m6A modification profile in the gonads of the olive flounder, Paralichthys olivaceus, an important maricultured fish in Asia. The gonadal differentiation and development of the flounder are controlled by genetic as well as environmental factors, and the epigenetic mechanism may play an important role. In total, 742 RNA editing events were identified, 459 of which caused A to I conversion. Most A-to-I sites were located in 3′UTRs, while 61 were detected in coding regions (CDs). The number of editing sites in the testis was higher than that in the ovary. Transcriptome-wide analyses showed that more than one-half of the transcribed genes presented an m6A modification in the flounder gonads, and approximately 60% of the differentially expressed genes (DEGs) between the testis and ovary appeared to be negatively correlated with m6A methylation enrichment. Further analyses revealed that the mRNA expression of some sex-related genes (e.g., dmrt1 and amh) in the gonads may be regulated by changes in mRNA m6A enrichment. Functional enrichment analysis indicated that the RNA editing and m6A modifications were enriched in several canonical pathways (e.g., Wnt and MAPK signaling pathways) in fish gonads and in some pathways whose roles have not been investigated in relation to fish sex differentiation and gonadal development (e.g., PPAR and RNA degradation pathways). There were 125 genes that were modified by both A-to-I editing and m6A, but the two types of modifications mostly occurred at different sites. Our results suggested that the presence of sex-specific RNA modifications may be involved in the regulation of gonadal development and gametogenesis.

Systematic analyses reveal RNA editing events involved in skeletal muscle development of goat (Capra hircus)

RNA editing is a posttranscriptional molecular process involved with specific nucleic modification, which can enhance the diversity of gene products. Adenosine-to-inosine (A-to-I, I is read as guanosine by both splicing and translation machinery) is the main type of RNA editing in mammals, which manifested as AG (adenosine-to-guanosine) in sequence data. Here, we aimed to explore patterns of RNA editing using RNA sequencing data from skeletal muscle at four developmental stages (three fetal periods and one postnatal period) in goat. We found the occurrences of RNA editing events raised at fetal periods and declined at the postnatal period. Also, we observed distinct editing levels of AG editing across stages, and significant difference was found between postnatal period and fetal periods. AG editing patterns in newborn goats are similar to those of 45-day embryo compared with embryo at 105 days and embryo at 60 days. In this study, we found a total of 1415 significantly differential edited AG sites among four groups. Moreover, 420 sites were obviously clustered into six time-series profiles, and one profile had significant association between editing level and gene expression. Our findings provided some novel insights into understanding the molecular mechanism of muscle development in mammals.