Deciphering the genetic regulation of peripheral blood transcriptome in pigs through expression genome-wide association study and allele-specific expression analysis

Polymorphism in cis-regulatory sequences can lead to different levels of expression for the two alleles of a gene, providing a starting point for the evolution of gene expression. Little is known about the genome-wide abundance of genetic variation in gene regulation in natural populations but analysis of allele-specific expression (ASE) provides a means for investigating such variation. We performed RNA-seq of multiple tissues from population samples of two closely related flycatcher species and developed a Bayesian algorithm that maximizes data usage by borrowing information from the whole data set and combines several SNPs per transcript to detect ASE. Of 2,576 transcripts analyzed in collared flycatcher, ASE was detected in 185 (7.2%) and a similar frequency was seen in the pied flycatcher. Transcripts with statistically significant ASE commonly showed the major allele in >90% of the reads, reflecting that power was highest when expression was heavily biased toward one of the alleles. This would suggest that the observed frequencies of ASE likely are underestimates. The proportion of ASE transcripts varied among tissues, being lowest in testis and highest in muscle. Individuals often showed ASE of particular transcripts in more than one tissue (73.4%), consistent with a genetic basis for regulation of gene expression. The results suggest that genetic variation in regulatory sequences commonly affects gene expression in natural populations and that it provides a seedbed for phenotypic evolution via divergence in gene expression.

Single-cell RNA-seq (scRNA-seq) is emerging as a powerful tool for understanding gene function across diverse cells. Recently, this has included the use of allele-specific expression (ASE) analysis to better understand how variation in the human genome affects RNA expression at the single-cell level. We reasoned that because intronic reads are more prevalent in single-nucleus RNA-Seq (snRNA-Seq), and introns are under lower purifying selection and thus enriched for genetic variants, that snRNA-seq should facilitate single-cell analysis of ASE. Here we demonstrate how experimental and computational choices can improve the results of allelic imbalance analysis. We explore how experimental choices, such as RNA source, read length, sequencing depth, genotyping, etc., impact the power of ASE-based methods. We developed a new suite of computational tools to process and analyze scRNA-seq and snRNA-seq for ASE. As hypothesized, we extracted more ASE information from reads in intronic regions than those in exonic regions and show how read length can be set to increase power. Additionally, hybrid selection improved our power to detect allelic imbalance in genes of interest. We also explored methods to recover allele-specific isoform expression levels from both long- and short-read snRNA-seq. To further investigate ASE in the context of human disease, we applied our methods to a Parkinson’s disease cohort of 94 individuals and show that ASE analysis had more power than eQTL analysis to identify significant SNP/gene pairs in our direct comparison of the two methods. Overall, we provide an end-to-end experimental and computational approach for future studies.

The degree to which the level of genetic variation for gene expression is shared across multiple tissues has important implications for research investigating the role of expression on the etiology of complex human traits and diseases. In the last few years, several studies have been published reporting the extent of overlap in expression quantitative trait loci (eQTL) identified in multiple tissues or cell types. Although these studies provide important information on the regulatory control of genes across tissues, their limited power means that they can typically only explain a small proportion of genetic variation for gene expression. Here, using expression data from monozygotic twins (MZ), we investigate the genetic control of gene expression in lymphoblastoid cell lines (LCL) and whole blood (WB). We estimate the genetic correlation that represents the combined effects of all causal loci across the whole genome and is a measure of the level of common genetic control of gene expression between the two RNA sources. Our results show that, when averaged across the genome, mean levels of genetic correlation for gene expression in LCL and WB samples are close to zero. We support our results with evidence from gene expression in an independent sample of LCL, T-cells, and fibroblasts. In addition, we provide evidence that housekeeping genes, which maintain basic cellular functions, are more likely to have high genetic correlations between the RNA sources than non-housekeeping genes, implying a relationship between the transcript function and the degree to which a gene has tissue-specific genetic regulatory control.

BackgroundStudies have demonstrated that inflammation has a key role in the pathogenesis of atherosclerosis due to the abnormal gene expressions of multiple cytokines. We established an accurate and precise method to observe gene expression in whole blood that might provide specific diagnostic information for coronary artery disease (CAD) and other related diseases.MethodsThe fifteen selected CAD-related genes (IL1B, IL6, IL8, IFNG, MCP-1, VWF, MTHFR, SELL, TNFalpha, ubiquitin, MCSF, ICAM1, ID2, HMOX1 and LDLR) and two housekeeping genes (ACTB and GK) as internal references have been measured simultaneously with a newly developed multiplex polymerase chain reaction (multi-PCR) method. Moreover, the precision was evaluated, and a procedure for distinguishing patients from the normal population has been developed based upon analyses of peripheral blood. A total of 148 subjects were divided into group A (control group without plaques), group B (calcified plaques) and group C (non-calcified plaques, and combination group) according dual-source CT criteria. Gene expression in blood was analyzed by multi-PCR, and levels of glucose and lipids measured in 50 subjects to explore the relationship among them.ResultsThe precision results of the multi-PCR system revealed within-run and between-run CV values of 3.695–12.537% and 4.405–13.405%, respectively. The profiles of cytokine gene expression in peripheral blood were set: a positive correlation between glucose and MCSF, HMOX1 or TNFalpha were found. We also found that triglyceride levels were negatively correlated with SELL gene expression in 50 subjects. Compared with controls, gene expression levels of IL1B, IL6, IL8 and MCP-1 increased significantly in group C.ConclusionsA new multiple gene expression analysis system has been developed. The primary data suggested that gene expression was related to CAD. This system might be used for risk assessment of CVDs and other related diseases.

BackgroundGenetic analysis of gene expression level is a promising approach for characterizing candidate genes that are involved in complex economic traits such as meat quality. In the present study, we conducted expression quantitative trait loci (eQTL) and allele-specific expression (ASE) analyses based on RNA-sequencing (RNAseq) data from the longissimus muscle of 189 Duroc × Luchuan crossed pigs in order to identify some candidate genes for meat quality traits.ResultsUsing a genome-wide association study based on a mixed linear model, we identified 7192 cis-eQTL corresponding to 2098 cis-genes (p ≤ 1.33e-3, FDR ≤ 0.05) and 6400 trans-eQTL corresponding to 863 trans-genes (p ≤ 1.13e-6, FDR ≤ 0.05). ASE analysis using RNAseq SNPs identified 9815 significant ASE-SNPs in 2253 unique genes. Integrative analysis between the cis-eQTL and ASE target genes identified 540 common genes, including 33 genes with expression levels that were correlated with at least one meat quality trait. Among these 540 common genes, 63 have been reported previously as candidate genes for meat quality traits, such as PHKG1 (q-value = 1.67e-6 for the leading SNP in the cis-eQTL analysis), NUDT7 (q-value = 5.67e-13), FADS2 (q-value = 8.44e-5), and DGAT2 (q-value = 1.24e-3).ConclusionsThe present study confirmed several previously published candidate genes and identified some novel candidate genes for meat quality traits via eQTL and ASE analyses, which will be useful to prioritize candidate genes in further studies.

Advancements in sequencing technology, improvements in genome annotation, and development of quantitative genetic models have been instrumental to the significant genetic gains achieved in pork production. Several quantitative trait loci (QTL) have been identified for growth, meat quality and carcass composition (GMC) phenotypes, however, the biological mechanisms underlying most QTL remain unknown. Functional genomic analysis can reveal insights on the genetic architecture of complex traits, and transcriptomic profiling of skeletal muscle during the conversion of muscle to meat can identify critical regulators of GMC phenotypes. Gene transcripts obtained with RNA-seq of longissimus muscle from 168 pigs were used to estimate gene expression variation subject to genetic control by mapping expression QTL (eQTL) and allele-specific expression (ASE). A total of 334 eQTL were mapped (FDR≤0.01) and joint association of eQTL with phenotypic QTL (pQTL) segregating in our population revealed 16 genes significantly associated with 21 pQTL for GMC phenotypes. ASE analysis facilitates the identification of cis-acting regulation of transcript abundance. We tested for ASE in 69,502 coding SNP (cSNP) and a total of 18,234 cSNP with significant ASE were identified (FDR≤0.01). A gene-wise conditional analysis fitting all ASE cSNP per gene for each phenotype identified 60 genes associated with GMC phenotypes. A comparison of eQTL with ASE cSNP showed an overlap of 136 genes. Pearson correlations of peak eQTL SNP with ASE cSNP was significant for 51% of these genes. The ASE analysis showed more precision in the identification of cis-acting effects than the genome-wide eQTL analysis; however, both approaches provide valuable information on the regulation of transcript abundance. For instance, we observed 24 genes associated with distant eQTL (trans effects) and exhibiting ASE. This study provides new information on the complex regulation of the pig longissimus muscle transcriptome and associations with measurable differences in economically important phenotypic traits.

Allele-specific expression (ASE) is a fundamental problem in studying gene regulation and diploid transcriptome profiles, with two key challenges: (i) haplotyping and (ii) estimation of ASE at the gene isoform level. Existing ASE analysis methods are limited by a dependence on haplotyping from laborious experiments or extra genome/family trio data. In addition, there is a lack of methods for gene isoform level ASE analysis. We developed a tool, IDP-ASE, for full ASE analysis. By innovative integration of Third Generation Sequencing (TGS) long reads with Second Generation Sequencing (SGS) short reads, the accuracy of haplotyping and ASE quantification at the gene and gene isoform level was greatly improved as demonstrated by the gold standard data GM12878 data and semi-simulation data. In addition to methodology development, applications of IDP-ASE to human embryonic stem cells and breast cancer cells indicate that the imbalance of ASE and non-uniformity of gene isoform ASE is widespread, including tumorigenesis relevant genes and pluripotency markers. These results show that gene isoform expression and allele-specific expression cooperate to provide high diversity and complexity of gene regulation and expression, highlighting the importance of studying ASE at the gene isoform level. Our study provides a robust bioinformatics solution to understand ASE using RNA sequencing data only.

BackgroundIntramuscular fat (IMF) content is a determining factor for meat taste. The Luchuan pig is a fat-type local breed in southern China that is famous for its desirable meat quality due to high IMF, however, the crossbred offspring of Luchuan sows and Duroc boars displayed within-population variation on meat quality, and the reason remains unknown.ResultsIn the present study, we identified 212 IMF-correlated genes (FDR ≤ 0.01) using correlation analysis between gene expression level and the value of IMF content. The IMF-correlated genes were significantly enriched in the processes of lipid metabolism and mitochondrial energy metabolism, as well as the AMPK/PPAR signaling pathway. From the IMF-correlated genes, we identified 99 genes associated with expression quantitative trait locus (eQTL) or allele-specific expression (ASE) signals, including 21 genes identified by both cis-eQTL and ASE analyses and 12 genes identified by trans-eQTL analysis. Genome-wide association study (GWAS) of IMF identified a significant QTL on SSC14 (p-value = 2.51E−7), and the nearest IMF-correlated gene SFXN4 (r = 0.28, FDR = 4.00E−4) was proposed as the candidate gene. Furthermore, we highlighted another three novel IMF candidate genes, namely AGT, EMG1, and PCTP, by integrated analysis of GWAS, eQTL, and IMF-gene correlation analysis.ConclusionsThe AMPK/PPAR signaling pathway together with the processes of lipid and mitochondrial energy metabolism plays a vital role in regulating porcine IMF content. Trait correlated expression combined with eQTL and ASE analysis highlighted a priority list of genes, which compensated for the shortcoming of GWAS, thereby accelerating the mining of causal genes of IMF.

Background: Genome-wide association study (GWAS) have identified over 45 susceptibility loci for lung cancer; many studies including our own group, have focused on low-frequency and rare coding variants using fine mapping and exome sequencing. This strategy, however, has met with limited success as about 90% of GWAS hits are noncoding and act primarily through altering transcriptional regulation in an allele-specific manner. The RNA-Seq based allele-specific expression (ASE) analysis affords an innovative approach to study preferential expression of an allele in direct relationship to its genotype, providing information on cis-regulatory effects for the expression of putative genes. However currently, there are no lung cancer studies that have rigorously evaluated the ASE variation in lung tumor and adjacent tissues. Methods: Leveraging The Cancer Genome Atlas (TCGA) resource, we performed transcriptomic-wide ASE analysis using existing RNA-Seq datasets of paired tumor and adjacent tissues from 54 lung adenocarcinoma patients. We first quantified the RNA read counts of Referent and Alternate alleles of heterozygous variants, then evaluated the allelic imbalance on a per-sample basis using Beta-binomial test, and explored the differential ASE between tumor and adjacent tissues using paired Wilcoxon test. Functional regulatory consequences were generated from Ensembl Variant Effect Predictor. Results: We identified total 208 significant ASEs, including 35 tissue-specific (only in tumor or only in adjacent), 28 sharing, and 145 differential variants. Of the 208 candidates, 41 were from the human leukocyte antigen (HLA) locus (primary DQA2, DQB1, DRB1, H and J), 26 were from the immunoglobulin (IG) superfamily (primary IGH, IGL, IGK and F11R). About 80% candidates were noncoding (mostly in 5’ and 3’ untranslated regions) and with regulatory features (21 promoter, seven enhancer, seven open chromatin region, two induce nonsense-mediated mRNA decay, one CTCF-binding site, and one transcription factor binding site). Other top genes included MDM2, APOL1, and CTSB. Pathway analyses revealed 27 genes involved in immune response pathway, and 12 genes involved in HLA antigen processing and presentation pathway. Conclusion: This study is the first transcriptomics ASE analysis in lung adenocarcinoma. The key somatic cis-regulatory ASE variants identified from this study, especially immunogenic allelic variations from HLA and IG genes, could be used for identifying high-risk individuals for targeted lung cancer checkpoint blockade and related immunotherapies. Citation Format: Yanhong Liu, Spiridon Tsavachidis, Farrah Kheradmand, Margaret R. Spitz, Chris Amos. Transcriptome analysis links immune genes allelic expression imbalances to lung cancer [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2019; 2019 Mar 29-Apr 3; Atlanta, GA. Philadelphia (PA): AACR; Cancer Res 2019;79(13 Suppl):Abstract nr 1584.

Catecholamine-related gene expression in blood correlates with tic severity in tourette syndrome

Differential abundance of allelic transcripts in a diploid organism, commonly referred to as allele specific expression (ASE), is a biologically significant phenomenon and can be examined using single nucleotide polymorphisms (SNPs) from RNA-seq. Quantifying ASE aids in our ability to identify and understand cis-regulatory mechanisms that influence gene expression, and thereby assist in identifying causal mutations. This study examines ASE in breast muscle, abdominal fat, and liver of commercial broiler chickens using variants called from a large sub-set of the samples (n = 68). ASE analysis was performed using a custom software called VCF ASE Detection Tool (VADT), which detects ASE of biallelic SNPs using a binomial test. On average ~ 174,000 SNPs in each tissue passed our filtering criteria and were considered informative, of which ~ 24,000 (~ 14%) showed ASE. Of all ASE SNPs, only 3.7% exhibited ASE in all three tissues, with ~ 83% showing ASE specific to a single tissue. When ASE genes (genes containing ASE SNPs) were compared between tissues, the overlap among all three tissues increased to 20.1%. Our results indicate that ASE genes show tissue-specific enrichment patterns, but all three tissues showed enrichment for pathways involved in translation.

A Burden of Rare Variants Associated with Extremes of Gene Expression in Human Peripheral Blood

Somatic copy number variations (CNVs), including abnormal chromosome numbers and structural changes leading to gain or loss of genetic material, play a crucial role in initiation and progression of cancer. CNVs are believed to cause gene dosage imbalances and modify cis-regulatory elements, leading to allelic expression imbalances in genes that influence cell division and thereby contribute to cancer development. However, the impact of CNVs on allelic gene expression in cancer remains unclear. Allele-specific expression (ASE) analysis, a potent method for investigating genome-wide allelic imbalance profiles in tumors, assesses the relative expression of two alleles using high-throughput sequencing data. However, many existing methods for gene-level ASE detection rely on only RNA sequencing data, which present challenges in interpreting the genetic mechanisms underlying ASE in cancer. To address this issue, we developed a robust framework that integrates allele-specific copy number calls into ASE calling algorithms by leveraging paired genome and transcriptome data from the same sample. This integration enhances the interpretability of the genetic mechanisms driving ASE, thereby facilitating the identification of driver events triggered by CNVs in cancer. In this study, we utilized BASE to conduct a comprehensive analysis of ASE in high hyperdiploid acute lymphoblastic leukemia (HeH ALL), a prevalent childhood malignancy characterized by gains of chromosomes X, 4, 6, 10, 14, 17, 18, and 21. Our analysis unveiled the comprehensive ASE landscape in HeH ALL. Through a multi-perspective examination of HeH ASEs, we offer a systematic understanding of how CNVs impact ASE in HeH, providing valuable insights to guide ASE studies in cancer.

Allele-specific expression (ASE) analysis improves the understanding of transcription’s cis-regulation. Herein, we used imputed SNPs along with RNA-Seq data from the Longissiumus thoracis muscle of 190 Nelore steers to identify functional cis-regulatory variants from ASE analysis. Using a Binomial Test, we identified 38,177 SNPs in ASE regions (ASE SNPs; FDR ≤0.05). We then searched for aseQTLs (SNPs potentially regulating the ASE) by comparing their heterozygosity to the measured allelic ratio under a Wilcoxon Rank Sum test. We identified 21,543 aseQTLs potentially regulating a total of 430 ASE SNPs (FDR ≤0.05). Based on a linear model, ASE SNPs and aseQTLs were associated with transcript abundance. We identified 3,333 SNPs acting as cis-eQTLs (FDR≤0.05). Results were integrated with previous ASE, functional regions, and meat quality-related differentially expressed genes data. This study described novel SNPs potentially regulating the transcription of genes that may affect beef traits.

The metabolic health benefits of fermented milks have already been investigated using clinical biomarkers but the development of transcriptomic analytics in blood offers an alternative approach that may help to sensitively characterise such effects. We aimed to assess the effects of probiotic yoghurt intake, compared to non-fermented, acidified milk intake, on clinical biomarkers and gene expression in peripheral blood. To this end, a randomised, crossover study was conducted in fourteen healthy, young men to test the two dairy products. For a subset of seven subjects, RNA sequencing was used to measure gene expression in blood collected during postprandial tests and after two weeks daily intake. We found that the postprandial response in insulin was different for probiotic yoghurt as compared to that of acidified milk. Moreover changes in several clinical biomarkers were associated with changes in the expression of genes representing six metabolic genesets. Assessment of the postprandial effects of each dairy product on gene expression by geneset enrichment analysis revealed significant, similar modulation of inflammatory and glycolytic genes after both probiotic yoghurt and acidified milk intake, although distinct kinetic characteristics of the modulation differentiated the dairy products. The aryl hydrocarbon receptor was a major contributor to the down-regulation of the inflammatory genesets and was also positively associated with changes in circulating insulin at 2h after yoghurt intake (p = 0.05). Daily intake of the dairy products showed little effect on the fasting blood transcriptome. Probiotic yoghurt and acidified milk appear to affect similar gene pathways during the postprandial phase but differences in the timing and the extent of this modulation may lead to different physiological consequences. The functional relevance of these differences in gene expression is supported by their associations with circulating biomarkers.