A Comprehensive Comparison Of RNA-seq-based Transcriptome Analysis Between The Porcine Subcutaneous And Visceral Adipose Tissues

Adipose tissue is not merely an organ designed to passively store excess carbon in the form of fatty acids esterified to glycerol (triacylglycerols). In essence, it’s also a major secretory and endocrine organ involved in a range of adipokines, such as adiponectin, leptin and resistin, they involved in various diseases such as obesity, type Ⅱ diabetes mellitus, cardiovascular disease, hyperlipidemia, and metabolic syndrome. It is well established that the metabolic risk factors of obesity and increased body weight are related more to body adipose distribution rather than total body adipose. Adipose tissue locates within the abdominal cavity, known as visceral adipose tissue (VAT), has been suggested to be anatomically, functionally and metabolically distinct from that of the compartmental subcutaneous adipose tissue (SAT). Here we offered a comprehensive comparison of histomorphology and fatty acid composition between one SAT (Hypodermal layer of backfat, HLB) and two VATs (Greater omentum, GOM; Mesenteric adipose tissue, MAD) from pig, it exhibited a large difference. In order to reveal the molecular mechanism, we further perform a transcriptome research by using the RNA sequencing (RNA-seq) approach. A large of novel transcripts, alternative splicing, SNVs and INDELs were identified, and the differential expressed (DE) genes between SAT and VAT were also identified. It offers a foundation for explaining the difference between SAT and VAT.Morphological differences among adipose tissues were studied by histological section, and the differences among adipocyte volume could be identified by quantitative image analysis. HLB had the largest adipocyte volume (1.92×105μm3), it’s significantly bigger than GOM (1.51×105μm3) and MAD (1.36×105μm), there is no significant difference between GOM and MAD. Then we examined the expression pattern of seven inflammatory genes in three adipose tissues, we observed that six pro-inflammatory genes (CD14, MCP-1, IL-6, IL-8, IL-1B, TNF-a) all had a high expression level in GOM and MAD, and IL-10, an anti-inflammatory cytokines primarily expressed in HLB. At last, we analyzed the fatty acid composition of three adipose tissues, comparing to HLB, GOM and MAD contain more SFA (P<0.001) and less MUFA (P<0.001) and PUFA (P<0.01). These results reveal the wide difference between three adipose tissues, especially between SAT and VAT.We sequenced the transcriptome of three porcine adipose tissues by RNA-seq, a total of263million,75bp paired-end reads were generated, yielding19.7gigabases (Gb) of sequence. We detected15,004out of the22,015identified transcripts (68.2%) in our adipose tissues, with13,325,13,120and13,484transcripts were found expressed in HLB, GOM and MAD respectively. Meanwhile,4,583-4,765novel transcript units,2,585splicing genes,16,559SNVs and1,481INDELs were also explored.We identified183differential expressed transcripts between SAT and VAT, GO and Pathway were analyzed, they involved in19functional categories, especially in fatty acid synthesis, carboxylic acid metabolism and inflammatory response. Furthermore, such important DE genes were analyzed between SAT and VAT. Firstly, three fatty acid metabolism genes SCD1, SCD5and FADS2were studied, SCD1was the highest expressed in SAT, where SCD5and FADS2expressed higher in VAT. Then, the carboxylic acid metabolic genes, UGP2, LDHA, SLC6A6, PHGDH and CRABP2were expressed higher in SAT, where RBP4, HPGD, SNCA, PGM5, ART4and ATF3were found principally expressed in VAT. At last, three inflammatory-related genes C3, CE2and ZFP36were found higher expressed in VAT.Integrate with the fatty acid composition of SAT and VAT, we highlight SCD1and FADS2were crucial for altering the fatty acid composition. It exhibit that VAT deposit more SFA that mainly affected by SCD1gene expression, and FADS2play a crucial role in generating more ω-6PUFA, especially AA, which is the key link between PUFAs and inflammatory processes. These result that VAT has a more sensibilization character for development inflammatory.Reliability analysis indicated that the average coefficient of correlation between RNA-seq and microarray was0.41for the’moderate’expressed transcripts, but it exhibited a low coefficient for the ’high’(r=0.22) and’low’(r=0.12) expressed transcripts. So we further used Q-PCR to verify the results, and a good correlation was exhibited (r=0.89, P=7.75E-07). Based on these results, it demonstrates that the RNA-seq technique used in this study is accurate and reproducible.In conclusions, this study provides a comprehensive comparison between SAT and VAT transcriptome, and reveals some crucial genes and valuable GO and Pathway. It provides some data for us to reveal the molecular mechanism difference between SAT and VAT.