Advertisement
Browse Subject Areas
?

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

  • Loading metrics

Open Access

Peer-reviewed

Research Article

Fat-tail allele-specific expression genes may affect fat deposition in tail of sheep

  • Hossein Mansourizadeh,

    Roles Formal analysis, Investigation, Visualization, Writing – original draft

    Affiliation Department of Animal and Poultry Science, College of Aburaihan, University of Tehran, Tehran, Iran

    ⨯
  • Mohammad Reza Bakhtiarizadeh ,

    Roles Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Visualization, Writing – original draft, Writing – review & editing

    mrbakhtiari@ut.ac.ir

    Affiliation Department of Animal and Poultry Science, College of Aburaihan, University of Tehran, Tehran, Iran

    ⨯
  • Luciana Correia de Almeida Regitano,

    Roles Methodology, Writing – review & editing

    Affiliation Embrapa Southeast Livestock, São Carlos, Brazil

    ⨯
  • Jennifer Jessica Bruscadin

    Roles Methodology, Writing – review & editing

    Affiliations Embrapa Southeast Livestock, São Carlos, Brazil, Department of Genetics and Evolution, Federal University of São Carlos (UFSCar), São Carlos, Brazil

    ⨯

Abstract

Different sheep breeds show distinct phenotypic plasticity in fat deposition in the tails. The genetic background underlying fat deposition in the tail of sheep is complex, multifactorial, and may involve allele-specific expression (ASE) mechanism to modulate allelic expression. ASE is a common phenomenon in mammals and refers to allelic imbalanced expression modified by cis-regulatory genetic variants that can be observed at heterozygous loci. Therefore, regulatory processes behind the fat-tail formation in sheep may be to some extent explained by cis- regulatory variants, through ASE mechanism, which was investigated in the present study. An RNA-Seq-based variant calling was applied to perform genome-wide survey of ASE genes using 45 samples from seven independent studies comparing the transcriptome of fat-tail tissue between fat- and thin-tailed sheep breeds. Using a rigorous computational pipeline, 115 differential ASE genes were identified, which were narrowed down to four genes (LPL, SOD3, TCP1 and LRPAP1) for being detected in at least two studies. Functional analysis revealed that the ASE genes were mainly involved in fat metabolism. Of these, LPL was of greater importance, as 1) observed in five studies, 2) reported as ASE gene in the previous studies and 3) with a known role in fat deposition. Our findings implied that complex physiological traits, like fat-tail formation, can be better explained by considering various genetic mechanisms, which can be more finely mapped through ASE analyses. The insights gained in this study indicate that biallelic expression may not be a common mechanism in sheep fat-tail development. Hence, allelic imbalance of the fat deposition-related genes can be considered a novel layer of information for future research on genetic improvement and increased efficiency in sheep breeding programs.

Citation: Mansourizadeh H, Bakhtiarizadeh MR, de Almeida Regitano LC, Bruscadin JJ (2024) Fat-tail allele-specific expression genes may affect fat deposition in tail of sheep. PLoS ONE 19(12): e0316046. https://doi.org/10.1371/journal.pone.0316046

Editor: Sayed Haidar Abbas Raza, South China Agricultural University, CHINA

Received: October 16, 2024; Accepted: December 4, 2024; Published: December 27, 2024

Copyright: © 2024 Mansourizadeh et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Data Availability: All relevant data are within the article and its Supporting Information files.

Funding: The author(s) received no specific funding for this work.

Competing interests: The authors have declared that no competing interests exist.

Introduction

Sheep were likely the first livestock species to be domesticated in the Fertile Crescent about 11,000 years ago [1]. Throughout a long history of evolution and breeding and based on their tail phenotype, sheep have been classified into one of the following main categories: thin- and fat-tailed breeds. The most probable wild ancestors of domestic sheep were thin-tailed. Hence, it is believed that the fat tail phenotype evolved after domestication as a valuable energy reservoir as well as an adaptive response to harsh environmental conditions [2]. However, fat-tail importance is decreasing in modern- or semi-intensive sheep production systems since it decreases the feed efficiency. Furthermore, most consumers prefer lean meat as a healthier choice [3, 4]. In this context, elucidating the genetic mechanisms underlying fat deposition in tail of sheep is of great importance to improve fat-tailed breeds.

The current transcriptome studies based on RNA-Seq datasets measure gene expression by counting the number of mRNA copies of the genes and compare the expression patterns between two or more conditions to identify differentially expressed genes (DEGs). While in diploid mammalian cells, most of the autosomal genes are typically expressed similarly from both alleles, cis-regulatory genetic variants or epigenetic modifications may induce imbalanced expression of the alleles, resulting in allele-specific expression (ASE) [5]. Therefore, ASE, or allelic imbalance, refers to the imbalanced expression of the alleles of a gene in a heterozygous locus, that can increase the diversity of transcript and protein isoforms abundance from one gene [6]. The extreme form of ASE is genomic imprinting or parent-of-origin specific gene expression, which results in the systematic monoallelic expression of the copy inherited from only one parent [7]. It is well known that a significant proportion of phenotypic diversity in mammals is mediated by transcriptional regulation [8, 9] of genes, which directly or indirectly is influenced by the genetic variants. Thus, the integration of genetic variants with transcriptome data through ASE analysis enables us to detect cis-regulatory variants that may underlie phenotypic differences in a population. In other words, ASE analysis provides an opportunity to identify genes whose expressions are affected by genetic variants at the same time it gives information for mapping these cis-regulatory variants. Furthermore, ASE analysis reduces the effect of prevalent technical bias compared to DEG analysis, as ASE analysis calculates the relative abundance of alleles at heterozygous loci per individual, while in DEG analysis [10] the difference in expression of a gene is an across-samples comparison.

In addition to the classical genetic mechanisms, ASE also provides a regulatory approach to modulate gene regulation patterns in mammals [11, 12] and contributes significantly to phenotypic diversity within and between species [13, 14]. These findings indicate that genomic variants have a greater impact on transcriptome variance than previously thought [15]. Hence, it can be noted that the identified causal SNPs by ASE analysis, that regulate gene expression, may help to narrow the gap between phenotype and genotype. For instance, in a recent study, several genes with ASE were detected across adipose and muscle tissues of F1 hybrids of thin- and fat-tailed sheep breeds. Of these, two important genes (PDGFD and IRF2BP2) were reported to be related to different phenotypes between the investigated sheep breeds [16]. Cao et al. compared the ASE of genes between two diverse goat breeds and 524 genes were identified, being involved in bone development, muscle cell differentiation and lipid metabolic processes. They reported these genes as potential candidates associated with the phenotypic differences between breeds [17]. It has been demonstrated that ASE genes are involved in adipogenesis and lipid metabolism in pigs [18]. Moreover, transcriptome datasets were applied to investigate ASE in different tissues in a wide range of species, including cattle [6], pig [19], mouse [20] and chicken [21]. However, in spite of these studies, the functional role of ASE mechanism is not yet clearly elucidated, especially in non-model animals such as sheep. Accordingly, ASE analysis facilitates a deeper comprehension of the molecular mechanisms behind complex traits, such as fat deposition in sheep. To date, no comprehensive study has been conducted to identify the SNPs with a cis-regulatory effect on the expression patterns of genes involved in fat-tail formation in sheep. Previous studies have shown that genetic variants derived from RNA-Seq data are reliable and can be used for downstream analysis [4, 22] such as ASE analysis [23]. Here, by taking advantage of RNA-Seq datasets, a comprehensive ASE analysis was performed, assuming that this approach can highlight important genes associated with fat-tail development in sheep. This study attempted to provide a detailed portrait of the ASE profile that may contribute to fat-tail formation in sheep breeds. To this end, seven RNA-Seq-based studies that have compared the different fat- and thin-tailed sheep breeds, were analyzed with the same pipeline. The identified candidate genes were selected by considering the genes that showed significant differential ASE in at least two datasets. Our findings shed new light on the underlying genetic mechanisms that might influence the shape and size of fat-tail tissue in different sheep breeds.

Material and methods

RNA-Seq datasets

In a recent study, we performed a meta-analysis on six independent RNA-Seq datasets that compared tail fat tissue of fat- and thin-tailed sheep breeds [3]. These six datasets along with one more related RNA-Seq dataset were applied in the present study to perform ASE analysis. All samples were derived from the tail fat tissue of adult male sheep with a minimum of three and a maximum of four biological replicates per breed. All studies employed transcriptomic strategies using various platforms, with most read lengths being 150 bp paired-end. All the samples were retrieved from the GEO database (Table 1) and were analyzed based on the same bioinformatics pipeline (Fig 1).

thumbnail
Download:
Fig 1. A schematic representation of the used bioinformatics pipeline to perform ASE analysis using RNA-Seq data.

Datasets of the seven studies were obtained from GEO database.

https://doi.org/10.1371/journal.pone.0316046.g001

thumbnail
Download:
Table 1. Characteristics of the used studies that compared fat- and thin-tailed sheep breeds.

https://doi.org/10.1371/journal.pone.0316046.t001

Quality control and trimming

Quality control of the raw RNA-Seq reads accessed using FastQC (v0.11.9) [31]. Then, Trimmomatic (v0.35) [32] was applied to eliminate adaptor contamination and remove low-quality bases/reads using the options: TRAILING:20, MAXINFO:120:0.9, MINLEN:120. After filtering, clean reads were re-assessed by FastQC to ensure the reliability of trimming steps.

Mapping and SNP calling

To align the clean reads against Ensembl Ovis aries reference genome assembly (rambouillet_v1.0.104), STAR software (v2.7.9a) was used in two-pass mode [33] with following parameters: -outFilterType BySJout, -outFilterMismatchNoverLmax 0.04, -outFilterMultimapNmax 10, -outFilterMismatchNmax 999, -outSAMmapqUnique 60 and -outFilterScoreMin 0.66. The two-pass mode was used to increase the accuracy and robustness of the alignment by improving the novel splice junction discovery. In this regard, in the first pass, STAR discovers novel splice junctions with high stringency across all the samples and inserts them into the junction database. In the second pass, it uses the novel splice sites along with known junctions to align the reads against the genome [34]. The known splice site information was obtained from the ENSEMBL GTF file (rambouillet_v1.0.104, release 106). Only concordantly and uniquely aligned reads were kept for variant calling according to the workflows and parameters recommended by GATK (v4.2.6.1) best-practice guidelines, explained as follow:

Potential PCR duplicates were removed using “MarkDuplicates” module of GATK software. Since, STAR is an RNA aligner, the alignments spanning intronic regions need to be reformatted before variant calling. To do this, SplitNCigarReads module of GATK was applied to hard clip the reads that extend into these regions. Hereafter, base recalibration to known variants based on Ensembl ovine SNP database (ovis aries, rambouillet) was done using BaseRecalibrator module, followed by ApplyBQSR module. Finally, variant calling was performed using GATK’s HaplotypeCaller module with a stand_call_conf and stand_emit_conf value of 30 and mbq of 25. Only biallelic SNPs were kept for ASE analysis and were filtered based on a number of standard quality metrics (Total depth of coverage<10, HomopolymerRun>5, RMSMappingQuality<40, QualitybyDepth<2, MappingQualityRankSum<-12.5 and ReadPosRankSum<-8). Spurious SNPs may result from aligning the reads to the problematic genomic regions [4]. Hence, further filtering processes were employed as follows, in an attempt to guarantee the reliability and quality of the identified SNPs for ASE analysis. Only the known variants (based on dbSNP database) with at least three supporting reads for the SNP were retained. These SNPs should not be located in problematic genomic regions including splice signal regions (within 5 bp intronic flanking), low-complexity regions (simple sequence repeats regions, ± 3 bases) and regions with bidirectional transcription.

Remapping by WASP

RNA-Seq read depths at heterozygous loci are commonly used to measure the magnitude of the allelic imbalance in ASE analysis. This approach is prone to technical artifacts resulting from mapping biases, as reads containing non-reference alleles at heterozygous loci may not map as efficiently as reads containing reference alleles at the same site [35]. WASP tool is one of the powerful and commonly used pipelines to efficiently correct the ambiguous mapping bias for accurate ASE analysis [36]. To this end, the aligned reads were remapped following the WASP (v0.3.4) pipeline. First, uniquely mapped reads overlapping heterozygous SNPs were detected (based on the identified variants in the previous step) and the allele that is present in the read was changed to match the SNP’s other allele; if a read contains a reference allele, it is switched to the alternative allele, and if a read contains an alternative allele, it is switched to the reference allele. Then, these artificial reads were remapped and were considered as biased if they did not remap to the same genomic location efficiently. All analysis was performed based on using codes provided in https://github.com/bmvdgeijn/WASP. Finally, variant calling was performed again based on our pipeline (as described above), for the concordantly and uniquely mapped reads. The identified known SNPs from this step were selected for the ASE analysis.

Haplotype phasing

It is strongly recommended to quantify read counts across genes based on phased variants, to enhance the accuracy and power of detecting ASE. Since a gene-based ASE can be performed by integrating effects across SNPs in the same gene, simply adding up allele counts may lead to double counting of reads and resulting in both false positives and false negatives. To overcome this limitation in the present study, a haplotype-based ASE analysis was conducted using phASER tool (v0.9.9.4). This software performs RNA-Seq-based read-backed phasing and aggregates counts across multiple SNPs (including single SNPs and haplotype blocks) to generate haplotype-level allelic read counts [37]. Hence, only uniquely mapped reads that were overlapped with heterozygous SNPs in the expressed regions of genes, were considered. Subsequently, these SNPs were phased with each other and their counts per locus were aggregated according to their respective haplotypes to estimate haplotype-level expression for each allele. To do this end, phaser.py of phASER used to count different haplotype reads. Then, phaser_gene_ae.py of this package was applied (with the flag—baseq 10) to quantify the sequencing reads overlapping heterozygous sites within protein coding genes based on gene bed file of Ensembl version 104 and gene level haplotypic expression matrix generated.

Differential ASE

To identify significant differences in allelic imbalances between the fat-tail of the sheep breeds, a gene-based ASE analysis was performed considering the null hypothesis that ASE values for a specific gene should be roughly equivalent between two groups. To this end, all samples of each breed of the seven studies were combined and two groups including fat- and thin-tailed compared. Taking into account the count-based nature of the data, a logistic regression model using a negative binomial distribution was adapted to evaluate whether the gene had allele-specific expression as well as its allelic imbalance differed between fat- and thin-tailed sheep breeds. It is worth noting that unlike the Poisson distribution, which assumes equal mean and variance, the negative binomial distribution can be considered a generalization of a Poisson distribution to allow for overdispersion, making it a more appropriate model for count-based dataset. In fact, negative binomial distribution introduces an extra parameter to model the overdispersion. This parameter captures the variability not explained by the Poisson distribution. In cases where there is no overdispersion, the negative binomial distribution simplifies to the Poisson distribution [36]. If the read counts for the reference and alternative alleles be denoted by ref and alt, respectively, the following model was used: glm.nb(ref ~ group, offset = (log(ref + alt))). In this model, offset adjust unequal library sizes across different samples. Also, group indicate the fat- and thin-tailed sheep breeds. Multiple testing correction was performed with the false discovery rate (FDR) method, and significant ASE genes were defined as those with adjusted p-values < = 0.1. For the last filtering step, a significant ASE gene should be observed in the samples of at least two studies to be considered a functional ASE gene. All the statistical analysis and visualization of the results were performed using R software (v4.3.2).

Functional analysis

To gain insights into the potential function of the differential ASE genes, functional enrichment analysis was performed using the EnrichR tool [38]. Biological processes and KEGG pathways with an FDR < 0.05 were considered significant. The significance level was determined using the gene ratio and adjusted p-value. The term "gene ratio" refers to the proportion of genes of interest present in a particular gene set or pathway relative to the total number of genes in that set. The adjusted p-value accounts for multiple hypothesis testing, helping to control the increased chance of false positive results when performing multiple statistical tests, thus providing a more accurate measure of statistical significance. The biological processes as well as KEGG pathways with FDR<0.05 were considered as significant terms.

Results

RNA-Seq data analysis

In total, more than 1,137 million reads, with an average of 25 million reads per sample, were analyzed. Of these, ~557 and ~580 million reads were related to fat- and thin-tailed breeds, respectively. The read counts ranged from eight to 50 million reads. After trimming and cleaning the RNA-Seq datasets, ~1,132 million reads were retained, of which ~576 and ~556 reads belonged to thin- and fat-tailed breeds, respectively. Only ∼0.004% of the reads were removed after filtering, reflecting the high quality of the data. Alignment rate of the clean reads against the reference genome ranged from 77 to 99% with an average rate of 93%. Of the aligned reads, ~974 million reads (~86%) were uniquely mapped to the genome. Details of the number of raw and clean reads as well as mapping statistics of different samples are provided in S1 File.

SNP calling

According to our RNA-Seq-based SNP discovery pipeline, a total of 1,415,762 known SNPs were identified in all samples after filtering. Then, heterozygous loci were extracted and subjected to remapping the reads by WASP approach to improve the accuracy of SNP calling. SNP discovery was performed again and 442,229 known heterozygous SNPs were discovered across all samples, which were applied for differential ASE analysis. Of these, 322,468 (72.8%) and 120,042 (27.2%) SNPs were transitions and transversions, respectively. The average number of SNPs per sample was 33,822. Of these, 270,437 SNPs were detected in the fat-tailed breeds and 341,523 SNPs in the thin-tailed group. The number of SNPs per chromosome varied from 50,055 (Chr1) to 5,512 (ChrX) with an average of 16,379 per chromosome (Fig 2).

thumbnail
Download:
Fig 2. The distribution of all identified heterozygous SNPs across the chromosomes of the reference sheep genome.

https://doi.org/10.1371/journal.pone.0316046.g002

ASE analysis

After aggregating the SNPs in haplotypes within the genic regions using PhASER [37], 2,995 genes were identified and subjected to differential ASE analysis. Fig 3 shows an UpSet plot of the intersection of these genes among the different studies. Statistical analysis revealed 115 differential ASE genes with a p-value <0.05 that were observed in at least one study. To improve the reliability of results, a set of stringent filtering criteria (Material and Methods) was applied and four genes (TCP1, LPL, LRPAP1 and SOD3) were selected based on the ASE significance threshold of FDR< = 0.1, also requiring that these ASE genes were identified in at least two studies. Of these, LPL was identified presenting ASE in five studies, SOD3 in three and the other two ASE genes were identified in two studies (S2 and S3 Files). To be sure that these genes are not imprinted genes in sheep, Geneimprint database (http://www.geneimprint.com/site/home) was investigated and it was found that none of them were imprinted genes. Genome-wide distribution of ASE results along with allelic ratio of the genes in both breeds is shown in Fig 4 as a Circos plot. Furthermore, Fig 5 shows the allelic ratio of the significant ASE genes across the different studies. As it is evident from Fig 5A, allelic ratio of the genes has the similar pattern in different studies in fat- and thin-tailed sheep breeds, which emphasized the reliability of the results.

thumbnail
Download:
Fig 3. UpSet plot of the common genes that were subjected to ASE analysis among the different studies.

The bar plots indicate the number of genes shared by specific studies. The horizontal bar chart at the bottom left represents the total number of genes found in each study. The box plots indicate the distribution of allelic ratio values for each breed. Study_2 is excluded as it did not have enough genes to be displayed.

https://doi.org/10.1371/journal.pone.0316046.g003

thumbnail
Download:
Fig 4. Circos plot displaying the genome-wide distribution of the ASE analysis results.

The outermost ring shows chromosome numbers. The analyzed genes were divided into two groups based on their allelic ratio: the genes with an average allelic ratio above 0.5 in the thin-tail breeds are shown in the first inner ring (purple points). While the genes with an average allelic ratio above 0.5 in fat-tail sheep breeds are presented in the second inner ring (pink points). The heatmaps are accompanied by histograms that show the average allelic ratio of each gene in both tail fat groups. The green and red colors of heatmap indicate the higher allelic ratio in thin- and fat-tailed sheep breeds, respectively.

https://doi.org/10.1371/journal.pone.0316046.g004

thumbnail
Download:
Fig 5. Allelic ratio of the significant differentially ASE genes in fat- and thin-tailed sheep breeds.

A) Heatmap representation of the allelic ratio for the four significant ASE genes in the used studies. The color scale indicates the intensity of higher expression of the reference allele. B) Box plot related to average allelic ratio of LPL gene in the fat-tail and thin-tailed breeds (FDR = 0.0048).

https://doi.org/10.1371/journal.pone.0316046.g005

Functional enrichment analysis

To explore the potential biological function of the genes with significant ASE, functional enrichment analysis was performed. In total, 44 biological processes and three KEGG pathways were significantly enriched with FDR < 0.05. It is noteworthy that several terms were directly related to fatty acid metabolism such as cholesterol metabolism (LPL and LRPAP1), PPAR signaling pathway (LPL) and negative regulation of lipoprotein particle clearance (LRPAP1) (Fig 6). S4 File provides the complete list of the functional enrichment analysis results.

thumbnail
Download:
Fig 6. Sankey plot related to the results of functional enrichment analysis.

The plot shows the connections between the selected significant genes (left) and the enriched biological terms (right).

https://doi.org/10.1371/journal.pone.0316046.g006

Discussion

ASE is one of many genetic mechanisms that contributes to the development of complex traits and phenotypic variation among organisms [39, 40]. Hence, it is well documented that allelic imbalance of genes can be the result of selection, with the functional consequence for adaptive phenotypic plasticity between breeds [41]. With the advantage of traditional bulk transcriptome analysis in mind, ASE analysis enable us to quantify the relative expression of alleles of genes. Therefore, we can identify the genes with similar total expression patterns between two groups of samples, while the expression of their alleles is not similar. Actually, ASE analysis is a useful approach to infer cis-regulatory variants associated with this expression pattern and enables us to decipher the genotype to phenotype map underlying complex traits. However, the exploration of genes containing ASE in the context of fat-tail formation in sheep remains relatively limited in existing literature. Previous studies have predominantly focused on classical transcriptome [24–30, 42] or genomic data analysis [4, 43–46]. Therefore, there is a gap between genetic variants and the molecular genetics behind the fat-tail development in sheep, and ASE analysis may provide novel insights to help close that gap. In this regard, to better understand the genetic mechanisms related to the phenotypic differences between fat- and thin-tailed sheep breeds, seven independent RNA-Seq datasets (fat-tail tissue) were analyzed to detect differential ASE genes at the whole-genome level. This study was performed to propel future research into important cis-acting genetic SNPs that may shape the fat-tail tissue of sheep. According to our previous studies [4, 47–49], an RNA-Seq based genetic variant discovery approach was applied to identify unbiased SNPs. Generally, more genetic variants (~ 21%) were detected in thin- than fat-tailed sheep breeds. This finding is in agreement with the hypothesis that wild ancestors of domestic sheep had thin tails and that fat-tailed breeds evolved as an environmental adaptation response after domestication [24, 50–52]. Therefore, it is expected that thin-tailed breeds, being the ancestral population, have greater genetic diversity. However, it is worth noting that the more SNPs in thin-tailed sheep breeds can be attributed to the faster evolution than fat-tailed breeds, which need to be further investigated.

On the other hand, to accurately estimate the allelic expression of genes, various aspects have to be considered during ASE analysis. Here, a reliable approach, the WASP method, was applied to remove the mapping bias favoring reference alleles [36], an important aspect to be controlled when investigating ASE [10]. It is reported that mapping bias correction using the WASP approach leads to obtaining extremely low false positive results in ASE analysis [53, 54]. Another aspect to take into account is the sparse and over-dispersed nature of count data [55], which is very common in allelic expression data. Assuming a Poisson or standard binomial distribution to test differential ASE genes can cause false positive results [56]. To avoid this, we applied a negative binomial distribution that can handle the variance independent of the mean using an additional parameter to explain the overdispersion in the data [57]. However, the key to our approach is using seven independent RNA-Seq datasets that provide an ideal opportunity to identify the common ASE genes associated with fat-tail formation in different sheep breeds. In other words, observing an ASE gene in some relevant studies makes it more likely to be involved in fat-tail development. Altogether this approach led to the identification of 115 genes with evidence of differential ASE (p-value <0.05). After adjustment for multiple comparisons and considering the ASE genes that were presented in at least two studies, four genes (TCP1, LPL, LRPAP1 and SOD3) were identified, which are predominantly involved in lipid metabolism processes such as "cholesterol metabolism”, “PPAR signaling pathway” and “negative regulation of lipoprotein particle clearance”. All these terms are directly related to fat deposition. For instance, it is well known that TFs of the PPAR play important roles in the clearance of cellular lipids via the regulation of many genes involved in fatty acid oxidation [24]. Lipoprotein particles, such as LDL (low-density lipoprotein) and VLDL (very-low-density lipoprotein), transport cholesterol and triglycerides in the bloodstream. Negative regulation of lipoprotein particle clearance refers to processes that decrease the removal of these particles from the blood. This can lead to higher levels of circulating lipoproteins, contributing to fat deposition [58].

Of the identified differential ASE genes, LPL stands out for the following reasons: First, it was observed as ASE in five different studies, which reinforces its potential regulatory role as ASE gene in tail fat deposition of sheep. It is essential to emphasize that the pattern of the allelic ratio of this gene was the same in the samples of each phenotypic group (Fig 5). Second, LPL is reported as ASE in several previous studies, showing consistency with our findings. Wang et al. used male Texel and female Kazakh to generate a hybrid population (n = 15). These F1 hybrids were applied to investigate the ASE genes in adipose tissue between thin- and fat-tailed breeds within the population. Fifteen ASE SNPs were identified in LPL, reported as an important ASE gene associated with lipid metabolism [16]. Moreover, ASE of LPL has been demonstrated in bovine muscle [6] and human tissues [59]. In this regard, LPL was reported as a gene containing breed-specific SNPs in comparison of two Iranian sheep breeds with different fat-tail phenotype (Lori vs Zel) [4]. Last but not least, LPL is closely related to lipid metabolism which makes it a promising candidate involved in fat-tail development through ASE mechanism. For example, LPL is reported as the most important factor affecting chicken fat deposition [60]. Moreover, the higher LPL activity was related to the faster abdominal adipose tissue fat deposition of broilers than laying hens [61]. LPL is a rate-limiting enzyme involved in the hydrolysis of plasma triglyceride (TG) (from TG-rich lipoproteins), thereby generating free fatty acids and glycerol [62]. These free fatty acids can be used as ligands and activators of peroxisomal proliferator activated receptor (PPAR) family of nuclear hormone receptors, which plays a key role in the metabolism of fatty acids and lipoproteins. Three PPAR isoforms have been identified including PPAR-α, PPAR-δ and PPAR-γ. Of these, α and δ stimulate fatty acid oxidation and γ activates lipid storage and adipogenesis [63, 64]. It is reported that products of LPL activity are an important source of PPAR-α and -δ ligands [65]. In this context, a possible hypothesis can be that genetic divergence and possibly adaptation to different environments induced genetic variants causing the difference in expression of LPL alleles. Different alleles of LPL may result in proteins with various catalytic efficiencies that lead to phenotypic variation of sheep fat-tail. This process may operate in concert with other molecular mechanisms to shape the different fat-tail phenotypes in sheep. A point worth emphasizing is that bulk transcriptome analysis only considers the total expression of genes and genes with ASE cannot be detected, while their total expression is similar in both conditions of interest. LPL can be considered as a typical example, as we previously showed that there is no significant difference for LPL expression level in the fat-tail of sheep breeds with different fat deposition contents using real-time PCR [62]. In this regard, LPL was not reported as DEGs between fat- and thin-tailed sheep breeds in most of the previous relevant studies [24, 26, 27, 29, 30] as well as in our previous meta-analysis of RNA-Seq datasets [3]. In line with this trend, LPL was not differentially expressed in subcutaneous and omental fat depots, between two cattle breeds with different fat depositions [66]. In the present study, a significant allelic imbalance of LPL was observed in most of the investigated datasets (>75% of samples). Therefore, in spite of the same expression pattern of LPL between fat- and thin-tailed sheep breeds, differences between the expression of alleles of this gene may contribute to the phenotypic heterogeneity of fat-tail among sheep breeds. This finding indicates that different molecular mechanisms have to be considered to identify the effective factors associated with a multi-layered biological phenomenon, such as fat-tail development.

The other important differential ASE gene (SOD3) is a member of a key family of enzymatic antioxidants, named superoxide dismutase family (including three isozymes SOD1, SOD2 and SOD3). SOD3 catalyzes the dismutation of two superoxide radicals into hydrogen peroxide and oxygen, which are less harmful to cells [67]. High-level expression of SOD3 in adipose tissue and adipocytes has been demonstrated in mice [68]. In this regard, it is well known that SOD3 is associated with fat deposition in animals [69, 70] and humans [71]. Furthermore, the role of SOD family, especially SOD3, in lipid metabolism and obesity has been previously explained in mice [72], chicken [73, 74], beef [75], dairy cattle [76], and pig [77]. For example, SOD knockout mice were more obese with larger adipose tissue and more TG than wild mice [68]. Here, unequal expression of SOD3 alleles was significantly detected in three studies which highlight its importance in fat deposition in the tail of sheep. It can be interesting to notice that a previous study revealed that SOD3 silencing increased the expression of genes involved in the PPAR pathway and stimulated the accumulation of TG in human adipocytes [68]. As discussed above, LPL provides the ligands of PPAR (free fatty acids), which tempts us to suggest there as a possible interaction between these genes (LPL and SOD3) to promote tissue fat deposition and affect fat-tail phenotype. Although this finding is not validated at the experimental level, it can motivate further studies to investigate the proposed mechanism.

Two other differential ASE genes (LRPAP1 and TCP1) were observed only in two datasets, however, their roles in lipid metabolism have been described in the previous studies. Interestingly, both genes are reported as genes showing ASE in ovine brown adipose tissues [78]. LDL receptor-related protein associated protein-1, LRPAP1, encodes a receptor-associated protein (RAP) that regulates circulating cholesterol in mice [79]. This gene binds to both LRP1 and LRP2 on the cell surface in the presence of calcium ions, which are the receptors that mediate the uptake of lipoproteins and various lipids [80]. In a recent study, a copy number variation screening was performed in fat- and thin-tailed sheep breeds. Association analysis detected a significant region associated with fat metabolism harboring LRPAP1 gene [81]. Also, the identified genetic variants in LRPAP1 are reported to be closely associated with high plasma lipid levels [82]. T-complex protein 1, TCP1, is a subunit of TCP1containing ring complex (TRiC) [83], which is a molecular chaperone that assists in the folding of newly synthesized proteins. Also, it plays a role in cell growth, proliferation and apoptosis in eukaryotes [84]. Important role of TCP1 in the remodeling of cytoskeletal protein (like actin and tubulin) has been documented [85]. On the other hand, it is well known that remodeling of cytoskeletal protein is one of the most important biological processes that occurred during the adipogenesis. Hence, potential role of this gene in lipid metabolism and fat deposition can be explained, as it is reported in bovine intramuscular fat accumulation [86]. Furthermore, Taga et al. investigated the key cellular and molecular events in the growth of adult adipose tissue in bovine fetal at different days post-conception (dpc). They identified proteins such as TCP1 at 110 and 180 dpc, which may regulate the proliferation of adipocyte precursors by controlling cell cycle progression and/or apoptosis or delaying PPARγ-induced differentiation [87]. It has also been reported that the subunits of TRiC activate the unfolded protein response (UPR), which enables a coordinated regulation of lipid metabolism and overall levels of cellular energy [88]. The results of the present study suggest that deviation from the general model of biallelic expression can provides an efficient way to understand the phenotypic variation as well as fat deposition pattern in tail of sheep breeds.

Conclusion

This study provides the first comprehensive analysis of ASE across fat-tail of 45 samples belonging to seven independent studies. Overall, a higher genetic diversity was observed in the thin-tailed breeds compared to the fat-tailed breeds. We identified four key candidate genes, LPL, SOD3, LRPAP1 and TCP1, showing significant differential ASE between fat- and thin-tailed sheep breeds. These genes were of great significance, as they were closely related to fat deposition. Among these genes, LPL has a well-known involvement in fat deposition, being detected as a differential ASE gene in most of the investigated studies. Moreover, LPL was reported as an ASE gene in previous studies, which makes it a promising candidate for fat-tail formation in sheep. Taken together, these results suggested that ASE can be considered as a potential regulatory mechanism behind the fat-tail development and the identified differential ASE genes may contribute to the shape of tails in sheep. Thus, the cis-regulatory mechanisms underlying these genes’ expression patterns can be further explored. Therefore, the proposed candidate genes, especially LPL, are worthy of further investigation to explain the phenotypic variation of sheep fat-tail and improve our current understanding of the process of fat-tail development.

Supporting information

S1 File. Details of the clean reads, uniquely mapped reads, multi mapped reads, percent of mapped reads per sample, after using our pipeline.

https://doi.org/10.1371/journal.pone.0316046.s001

(XLSX)

S2 File. List of differential ASE genes and their allelic ratio (Reference / Alternative) across all samples.

https://doi.org/10.1371/journal.pone.0316046.s002

(XLSX)

S3 File. The identified genotypes/haplotype of the identified four ASE genes (TCP1, LPL, LRPAP1 and SOD3) in different samples.

https://doi.org/10.1371/journal.pone.0316046.s003

(XLSX)

S4 File. List of the identified biological processes and KEGG pathways related to the genes with significant ASE.

https://doi.org/10.1371/journal.pone.0316046.s004

(XLSX)

References

  1. 1. Wei C, Wang H, Liu G, Wu M, Cao J, Liu Z, et al. Genome-wide analysis reveals population structure and selection in Chinese indigenous sheep breeds. BMC Genomics. 2015;16: 1–12.
  2. 2. Pan Z, Li S, Liu Q, Wang Z, Zhou Z, Di R, et al. Rapid evolution of a retro-transposable hotspot of ovine genome underlies the alteration of BMP2 expression and development of fat tails. BMC Genomics. 2019;20: 1–15. pmid:30940097
  3. 3. Hosseini SF, Bakhtiarizadeh MR, Salehi A. Meta-analysis of RNA-Seq datasets highlights novel genes/pathways involved in fat deposition in fat-tail of sheep. Front Vet Sci. 2023;10: 1159921. pmid:37252399
  4. 4. Bakhtiarizadeh MR, Alamouti AA. RNA-Seq based genetic variant discovery provides new insights into controlling fat deposition in the tail of sheep. Scientific Reports 2020 10:1. 2020;10: 1–13. pmid:32782325
  5. 5. Bruscadin JJ, Cardoso TF, da Silva Diniz WJ, de Souza MM, Afonso J, Vieira D, et al. Differential Allele-Specific Expression Revealed Functional Variants and Candidate Genes Related to Meat Quality Traits in B. indicus Muscle. Genes (Basel). 2022;13: 2336. pmid:36553605
  6. 6. Bruscadin JJ, de Souza MM, de Oliveira KS, Rocha MIP, Afonso J, Cardoso TF, et al. Muscle allele-specific expression QTLs may affect meat quality traits in Bos indicus. Scientific Reports 2021 11:1. 2021;11: 1–14. pmid:33795794
  7. 7. Ferguson-Smith AC. Genomic imprinting: the emergence of an epigenetic paradigm. Nature Reviews Genetics 2011 12:8. 2011;12: 565–575. pmid:21765458
  8. 8. Tomar A, Teperino R. Genetic control of non-genetic inheritance in mammals: state-of-the-art and perspectives. Mammalian Genome. 2020;31: 146–156. pmid:32529318
  9. 9. Ecker S, Pancaldi V, Valencia A, Beck S, Paul DS. Epigenetic and Transcriptional Variability Shape Phenotypic Plasticity. BioEssays. 2018;40: 1700148. pmid:29251357
  10. 10. Castel SE, Levy-Moonshine A, Mohammadi P, Banks E, Lappalainen T. Tools and best practices for data processing in allelic expression analysis. Genome Biol. 2015;16: 1–12.
  11. 11. Huang WC, Ferris E, Cheng T, Hörndli CS, Gleason K, Tamminga C, et al. Diverse Non-genetic, Allele-Specific Expression Effects Shape Genetic Architecture at the Cellular Level in the Mammalian Brain. Neuron. 2017;93: 1094–1109.e7. pmid:28238550
  12. 12. Chamberlain AJ, Vander Jagt CJ, Hayes BJ, Khansefid M, Marett LC, Millen CA, et al. Extensive variation between tissues in allele specific expression in an outbred mammal. BMC Genomics. 2015;16: 1–20.
  13. 13. de Souza MM, Zerlotini A, Rocha MIP, Bruscadin JJ, Diniz WJ da S, Cardoso TF, et al. Allele-specific expression is widespread in Bos indicus muscle and affects meat quality candidate genes. Scientific Reports 2020 10:1. 2020;10: 1–11. pmid:32576896
  14. 14. Skelly DA, Ronald J, Akey JM. Inherited variation in gene expression. Annu Rev Genomics Hum Genet. 2009;10: 313–332. pmid:19630563
  15. 15. Ardlie KG, DeLuca DS, Segrè A V., Sullivan TJ, Young TRGelfand ET, et al. The Genotype-Tissue Expression (GTEx) pilot analysis: Multitissue gene regulation in humans. Science (1979). 2015;348: 648–660.
  16. 16. Wang F, Shao J, He S, Guo Y, Pan X, Wang Y, et al. Allele-specific expression and splicing provide insight into the phenotypic differences between thin- and fat-tailed sheep breeds. J Genet Genomics. 2022;49: 583–586. pmid:34998977
  17. 17. Cao Y, Xu H, Li R, Gao S, Chen N, Luo J, et al. Genetic basis of phenotypic differences between Chinese Yunling black goats and Nubian goats revealed by allele-specific expression in their F1 hybrids. Front Genet. 2019;10: 421028. pmid:30891061
  18. 18. Stachowiak M, Szczerbal I, Flisikowski K. Investigation of allele-specific expression of genes involved in adipogenesis and lipid metabolism suggests complex regulatory mechanisms of PPARGC1A expression in porcine fat tissues. BMC Genet. 2018;19: 1–9.
  19. 19. Stachowiak M, Flisikowski K. Analysis of allele-specific expression of seven candidate genes involved in lipid metabolism in pig skeletal muscle and fat tissues reveals allelic imbalance of ACACA, LEP, SCD, and TNF. J Appl Genet. 2019;60: 97–101. pmid:30684136
  20. 20. Hasin-Brumshtein Y, Hormozdiari F, Martin L, van Nas A, Eskin E, Lusis AJ, et al. Allele-specific expression and eQTL analysis in mouse adipose tissue. BMC Genomics. 2014;15: 1–13.
  21. 21. Tomlinson MJ, Polson SW, Qiu J, Lake JA, Lee W, Abasht B. Investigation of allele specific expression in various tissues of broiler chickens using the detection tool VADT. Scientific Reports 2021 11:1. 2021;11: 1–13. pmid:33597613
  22. 22. Abbasabadi H, Bakhtiarizadeh MR, Moradi MH, McEwan JC. RNA-Seq based selection signature analysis for identifying genomic footprints associated with the fat-tail phenotype in sheep. Front Vet Sci. 2024;11. pmid:39403211
  23. 23. Hu YJ, Sun W, Tzeng JY, Perou CM. Proper Use of Allele-Specific Expression Improves Statistical Power for cis-eQTL Mapping with RNA-Seq Data. J Am Stat Assoc. 2015;110: 962–974. pmid:26568645
  24. 24. Bakhtiarizadeh MR, Salehi A, Alamouti AA, Abdollahi-Arpanahi R, Salami SA. Deep transcriptome analysis using RNA-Seq suggests novel insights into molecular aspects of fat-tail metabolism in sheep. Scientific Reports 2019 9:1. 2019;9: 1–14. pmid:31235755
  25. 25. Identification of the Genetic Basis for the Large-tailed Phenotypic Trait in Han Sheep Through Integrated mRNA&nbsp;and miRNA&nbsp;Analysis of Tail Fat Tissue Samples. 2020 [cited 7 May 2024].
  26. 26. Ma L, Zhang M, Jin Y, Erdenee S, Hu L, Chen H, et al. Comparative transcriptome profiling of mRNA and lncRNA related to tail adipose tissues of sheep. Front Genet. 2018;9: 355770. pmid:30250481
  27. 27. Yuan Z, Ge L, Sun J, Zhang W, Wang S, Cao X, et al. Integrative analysis of Iso-Seq and RNA-seq data reveals transcriptome complexity and differentially expressed transcripts in sheep tail fat. PeerJ. 2021;9: e12454. pmid:34760406
  28. 28. Fan H, Hou Y, Sahana G, Gao H, Zhu C, Du L, et al. A Transcriptomic Study of the Tail Fat Deposition in Two Types of Hulun Buir Sheep According to Tail Size and Sex. Animals 2019, Vol 9, Page 655. 2019;9: 655. pmid:31491862
  29. 29. Farhadi S, Ghias JS, Hasanpur K, Mohammadi SA, Ebrahimie E. Molecular mechanisms of fat deposition: IL-6 is a hub gene in fat lipolysis, comparing thin-tailed with fat-tailed sheep breeds. Arch Anim Breed. 2021;64: 53–68. pmid:34084904
  30. 30. Jin M, Fei X, Li T, Lu Z, Chu M, Di R, et al. Oar-miR-432 Regulates Fat Differentiation and Promotes the Expression of BMP2 in Ovine Preadipocytes. Front Genet. 2022;13: 844747. pmid:35559046
  31. 31. Brown J, Pirrung M, Mccue LA. FQC Dashboard: integrates FastQC results into a web-based, interactive, and extensible FASTQ quality control tool. Bioinformatics. 2017;33: 3137–3139. pmid:28605449
  32. 32. Bolger AM, Lohse M, Usadel B. Trimmomatic: a flexible trimmer for Illumina sequence data. Bioinformatics. 2014;30: 2114–2120. pmid:24695404
  33. 33. Dobin A, Davis CA, Schlesinger F, Drenkow J, Zaleski C, Jha S, et al. STAR: ultrafast universal RNA-seq aligner. Bioinformatics. 2013;29: 15–21. pmid:23104886
  34. 34. Veeneman BA, Shukla S, Dhanasekaran SM, Chinnaiyan AM, Nesvizhskii AI. Two-pass alignment improves novel splice junction quantification. Bioinformatics. 2016;32: 43–49. pmid:26519505
  35. 35. Stevenson KR, Coolon JD, Wittkopp PJ. Sources of bias in measures of allele-specific expression derived from RNA-seq data aligned to a single reference genome. BMC Genomics. 2013;14: 1–13.
  36. 36. Van De Geijn B, Mcvicker G, Gilad Y, Pritchard JK. WASP: allele-specific software for robust molecular quantitative trait locus discovery. Nature Methods 2015 12:11. 2015;12: 1061–1063. pmid:26366987
  37. 37. Castel SE, Mohammadi P, Chung WK, Shen Y, Lappalainen T. Rare variant phasing and haplotypic expression from RNA sequencing with phASER. Nature Communications 2016 7:1. 2016;7: 1–6. pmid:27605262
  38. 38. Chen EY, Tan CM, Kou Y, Duan Q, Wang Z, Meirelles G V., et al. Enrichr: Interactive and collaborative HTML5 gene list enrichment analysis tool. BMC Bioinformatics. 2013;14: 1–14.
  39. 39. Cleary S, Seoighe C. Perspectives on Allele-Specific Expression. Annu Rev Biomed Data Sci. 2021;4: 101–122. pmid:34465174
  40. 40. Lim KS, Chang SS, Choi BH, Lee SH, Lee KT, Chai HH, et al. Genome-Wide Analysis of Allele-Specific Expression Patterns in Seventeen Tissues of Korean Cattle (Hanwoo). Animals 2019, Vol 9, Page 727. 2019;9: 727. pmid:31561539
  41. 41. St. Pierre CL, Macias-Velasco JF, Wayhart JP, Yin L, Semenkovich CF Lawson HA. Genetic, epigenetic, and environmental mechanisms govern allele-specific gene expression. Genome Res. 2022;32: 1042–1057. pmid:35501130
  42. 42. Bakhtiarizadeh MR. Deciphering the role of alternative splicing as a potential regulator in fat-tail development of sheep: a comprehensive RNA-seq based study. Scientific Reports 2024 14:1. 2024;14: 1–11. pmid:38287039
  43. 43. Guo Y, Bai F, Wang J, Fu S, Zhang Y, Liu X, et al. Design and characterization of a high-resolution multiple-SNP capture array by target sequencing for sheep. J Anim Sci. 2023;101. pmid:36402741
  44. 44. Li X, Yang J, Shen M, Xie XL, Liu GJ, Xu YX, et al. Whole-genome resequencing of wild and domestic sheep identifies genes associated with morphological and agronomic traits. Nature Communications 2020 11:1. 2020;11: 1–16. pmid:32499537
  45. 45. Zhang Y, Han D, Dong X, Wang J, Chen J, Yao Y, et al. Genome-wide profiling of RNA editing sites in sheep. J Anim Sci Biotechnol. 2019;10: 1–9.
  46. 46. Moghaddar N, Khansefid M, Van Der Werf JHJ, Bolormaa S, Duijvesteijn N, Clark SA, et al. Genomic prediction based on selected variants from imputed whole-genome sequence data in Australian sheep populations. Genetics Selection Evolution. 2019;51: 1–14. pmid:31805849
  47. 47. Behroozi J, Shahbazi S, Bakhtiarizadeh MR, Mahmoodzadeh H. Genome-Wide Characterization of RNA Editing Sites in Primary Gastric Adenocarcinoma through RNA-seq Data Analysis. Int J Genomics. 2020;2020. pmid:33415135
  48. 48. Shafiei H, Bakhtiarizadeh MR, Salehi A. Large-scale potential RNA editing profiling in different adult chicken tissues. Anim Genet. 2019;50: 460–474. pmid:31355950
  49. 49. Bakhtiarizadeh MR, Salehi A, Rivera RM. Genome-wide identification and analysis of A-to-I RNA editing events in bovine by transcriptome sequencing. PLoS One. 2018;13: e0193316. pmid:29470549
  50. 50. Dong K, Yang M, Han J, Ma Q, Han J, Song Z, et al. Genomic analysis of worldwide sheep breeds reveals PDGFD as a major target of fat-tail selection in sheep. BMC Genomics. 2020;21: 1–12.
  51. 51. Zhu C, Fan H, Yuan Z, Hu S, Ma X, Xuan J, et al. Genome-wide detection of CNVs in Chinese indigenous sheep with different types of tails using ovine high-density 600K SNP arrays. Scientific Reports 2016 6:1. 2016;6: 1–9. pmid:27282145
  52. 52. Muigai AWT, Hanotte O. The Origin of African Sheep: Archaeological and Genetic Perspectives. African Archaeological Review. 2013;30: 39–50.
  53. 53. Castel SE, Aguet F, Mohammadi P, Aguet F, Anand S, Ardlie KG, et al. A vast resource of allelic expression data spanning human tissues. Genome Biol. 2020;21: 1–12. pmid:32912332
  54. 54. Miao Z, Alvarez M, Pajukanta P, Ko A. ASElux: an ultra-fast and accurate allelic reads counter. Bioinformatics. 2018;34: 1313–1320. pmid:29186329
  55. 55. Payne EH, Gebregziabher M, Hardin JW, Ramakrishnan V, Egede LE. An empirical approach to determine a threshold for assessing overdispersion in Poisson and negative binomial models for count data. Commun Stat Simul Comput. 2018;47: 1722–1738. pmid:30555205
  56. 56. Degner JF, Marioni JC, Pai AA, Pickrell JK, Nkadori E, Gilad Y, et al. Effect of read-mapping biases on detecting allele-specific expression from RNA-sequencing data. Bioinformatics. 2009;25: 3207–3212. pmid:19808877
  57. 57. Lindén A, Mäntyniemi S. Using the negative binomial distribution to model overdispersion in ecological count data. Ecology. 2011;92: 1414–1421. pmid:21870615
  58. 58. Qiao YN, Zou YL, Guo SD. Low-density lipoprotein particles in atherosclerosis. Front Physiol. 2022;13: 931931. pmid:36111155
  59. 59. Richardson K, Nettleton JA, Rotllan N, Tanaka T, Smith CE, Lai CQ, et al. Gain-of-function lipoprotein lipase variant rs13702 modulates lipid traits through disruption of a MicroRNA-410 seed site. Am J Hum Genet. 2013;92: 5–14. pmid:23246289
  60. 60. Kersten S. Physiological regulation of lipoprotein lipase. Biochimica et Biophysica Acta (BBA)—Molecular and Cell Biology of Lipids. 2014;1841: 919–933. pmid:24721265
  61. 61. Griffin HD, Butterwith SC, Goddard C. Contribution of lipoprotein lipase to differences in fatness between broiler and layer‐strain chickens. Br Poult Sci. 1987;28: 197–206. pmid:3607546
  62. 62. Bakhtiarizadeh MR, Moradi-Shahrbabak M, Ebrahimie E. Underlying functional genomics of fat deposition in adipose tissue. Gene. 2013;521: 122–128. pmid:23523858
  63. 63. Varga T, Czimmerer Z, Nagy L. PPARs are a unique set of fatty acid regulated transcription factors controlling both lipid metabolism and inflammation. Biochimica et Biophysica Acta (BBA)—Molecular Basis of Disease. 2011;1812: 1007–1022. pmid:21382489
  64. 64. Kliewer SA, Sundseth SS, Jones SA, Brown PJ, Wisely GB, Koble CS, et al. Fatty acids and eicosanoids regulate gene expression through direct interactions with peroxisome proliferator-activated receptors α and γ. Proceedings of the National Academy of Sciences. 1997;94: 4318–4323. pmid:9113987
  65. 65. Ziouzenkova O, Perrey S, Asatryan L, Hwang J, MacNaul KL, Moller DE, et al. Lipolysis of triglyceride-rich lipoproteins generates PPAR ligands: Evidence for an antiinflammatory role for lipoprotein lipase. Proc Natl Acad Sci U S A. 2003;100: 2730–2735. pmid:12606719
  66. 66. Ren MQ, Wegner J, Bellmann O, Brockmann GA, Schneider F, Teuscher F, et al. Comparing mRNA levels of genes encoding leptin, leptin receptor, and lipoprotein lipase between dairy and beef cattle. Domest Anim Endocrinol. 2002;23: 371–381. pmid:12206871
  67. 67. Zelko IN, Mariani TJ, Folz RJ. Superoxide dismutase multigene family: a comparison of the CuZn-SOD (SOD1), Mn-SOD (SOD2), and EC-SOD (SOD3) gene structures, evolution, and expression. Free Radic Biol Med. 2002;33: 337–349. pmid:12126755
  68. 68. Gao D, Hu S, Zheng X, Lin W, Gao J, Chang K, et al. SOD3 Is Secreted by Adipocytes and Mitigates High-Fat Diet-Induced Obesity, Inflammation, and Insulin Resistance. https://home.liebertpub.com/ars. 2019;32: 193–212. pmid:31680537
  69. 69. Vahedi SM, Salek Ardestani S, Karimi K, Banabazi MH. Weighted Single-Step GWAS for Body Mass Index and Scans for Recent Signatures of Selection in Yorkshire Pigs. Journal of Heredity. 2022;113: 325–335. pmid:35079818
  70. 70. Zhang T, Zhang LC, Wang LG, Wang LX. Study on methylation difference of fat deposition candidate genes AOC3, PPARG1 and SOD3 in Pig. China Anim Husb Vet Med. 2016;43: 2820–2825.
  71. 71. Lewandowski Ł, Kepinska M, Milnerowicz H. Alterations in Concentration/Activity of Superoxide Dismutases in Context of Obesity and Selected Single Nucleotide Polymorphisms in Genes: SOD1, SOD2, SOD3. International Journal of Molecular Sciences 2020, Vol 21, Page 5069. 2020;21: 5069. pmid:32709094
  72. 72. Ito J, Ishii N, Akihara R, Lee J, Kurahashi T, Homma T, et al. A high-fat diet temporarily renders Sod1-deficient mice resistant to an oxidative insult. J Nutr Biochem. 2017;40: 44–52. pmid:27855316
  73. 73. Association of single nucleotide polymorphisms with phenotypic production traits in broiler chickens—ProQuest. [cited 7 May 2024]. Available: https://www.proquest.com/openview/7950be2c618b83d0f08ee539f1801fae/1?pq-origsite=gscholar&cbl=18750
    • 74. Iqbal M, Pumford NR, Tang ZX, Lassiter K, Ojano-Dirain C, Wing T, et al. Compromised liver mitochondrial function and complex activity in low feed efficient broilers are associated with higher oxidative stress and differential protein expression. Poult Sci. 2005;84: 933–941. pmid:15971533
    • 75. Alexandre PA, Kogelman LJA, Santana MHA, Passarelli D, Pulz LH, Fantinato-Neto P, et al. Liver transcriptomic networks reveal main biological processes associated with feed efficiency in beef cattle. BMC Genomics. 2015;16: 1–13.
    • 76. Xia L, Zhao Z, Yu X, Lu C, Jiang P, Yu H, et al. Integrative analysis of miRNAs and mRNAs revealed regulation of lipid metabolism in dairy cattle. Funct Integr Genomics. 2021;21: 393–404. pmid:33963462
    • 77. Key Genes and Regulatory Network Analysis of Lipid Metabolism Differences between Back Fat and Abdominal Fat of Large Diqing Tibetan Pigs at Different Growth Stages. [cited 7 May 2024]. Available: https://www.xmsyxb.com/EN/10.11843/j.issn.0366-6964.2023.02.010
      • 78. Ghazanfar S, Vuocolo T, Morrison JL, Nicholas LM, McMillen IC, Yang JYH, et al. Gene expression allelic imbalance in ovine brown adipose tissue impacts energy homeostasis. PLoS One. 2017;12: e0180378. pmid:28665992
      • 79. Willnow TE, Armstrong SA, Hammer RE, Herz J. Functional expression of low density lipoprotein receptor-related protein is controlled by receptor-associated protein in vivo. Proceedings of the National Academy of Sciences. 1995;92: 4537–4541. pmid:7538675
      • 80. Thiriet M. Receptors of Cell–Matrix Mass Transfer. Signaling at the Cell Surface in the Circulatory and Ventilatory Systems. 2012; 335–359.
      • 81. Taghizadeh S, Gholizadeh M, rahimi-Mianji G, Moradi MH, Costilla R, Moore S, et al. Genome-wide identification of copy number variation and association with fat deposition in thin and fat-tailed sheep breeds. Scientific Reports 2022 12:1. 2022;12: 1–12. pmid:35614300
      • 82. Variation in the lipoprotein receptor-related protein, α2-ma…: Coronary Artery Disease. [cited 7 May 2024]. Available: https://journals.lww.com/coronary-artery/abstract/2002/08000/variation_in_the_lipoprotein_receptor_related.1.aspx
        • 83. Wang XN, Sim BR, Chen H, Zheng YX, Xue JB, Wang L, et al. Identification of sitagliptin binding proteins by affinity purification mass spectrometry: Sitagliptin binding proteins identified by affinity purification mass spectrometry. Acta Biochim Biophys Sin (Shanghai). 2022;54: 1453. pmid:36239351
        • 84. Ghozlan H, Cox A, Nierenberg D, King S, Khaled AR. The TRiCky Business of Protein Folding in Health and Disease. Front Cell Dev Biol. 2022;10: 906530. pmid:35602608
        • 85. Kim NK, Lee SH, Cho YM, Son ES, Kim KY, Lee CS, et al. Proteome analysis of the m. longissimus dorsi between fattening stages in Hanwoo steer. BMB Rep. 2009;42: 433–438. pmid:19643041
        • 86. Rajesh RV, Park MR, Heo KN, Yoon D, Kim TH, Lee HJ. Proteomic Analysis of Bovine Longissimus Muscle Satellite Cells during Adipogenic Differentiation. Asian-Australas J Anim Sci. 2011;24: 685–695.
        • 87. Taga H, Chilliard Y, Meunier B, Chambon C, Picard B, Zingaretti MC, et al. Cellular and molecular large-scale features of fetal adipose tissue: Is bovine perirenal adipose tissue Brown1685. J Cell Physiol. 2012;227: 1688–1700. pmid:21678425
        • 88. Gutiérrez‐Gutiérrez Ó, Felix DA, Salvetti A, Amro EM, Thems A, Pietsch S, et al. Regeneration in starved planarians depends on TRiC/CCT subunits modulating the unfolded protein response. EMBO Rep. 2021;22.