Comparative omics and feeding manipulations in chicken indicate a shift of the endocrine role of visceral fat towards reproduction

Abstract

Background: The mammalian adipose tissue plays a central role in energy-balance control, whereas the avian visceral fat hardly expresses leptin, the key adipokine in mammals. Therefore, to assess the endocrine role of adipose tissue in birds, we compared the transcriptome and proteome between two metabolically different types of chickens, broilers and layers, bred towards efficient meat and egg production, respectively.

Results: Broilers and layer hens, grown up to sexual maturation under free-feeding conditions, differed 4.0-fold in weight and 1.6-fold in ovarian-follicle counts, yet the relative accumulation of visceral fat was comparable. RNA-seq and mass-spectrometry (MS) analyses of visceral fat revealed differentially expressed genes between broilers and layers, 1106 at the mRNA level (FDR ≤ 0.05), and 203 at the protein level (P ≤ 0.05). In broilers, Ingenuity Pathway Analysis revealed activation of the PTEN-pathway, and in layers increased response to external signals. The expression pattern of genes encoding fat-secreted proteins in broilers and layers was characterized in the RNA-seq and MS data, as well as by qPCR on visceral fat under free feeding and 24 h-feed deprivation. This characterization was expanded using available RNA-seq data of tissues from red junglefowl, and of visceral fat from broilers of different types. These comparisons revealed expression of new adipokines and secreted proteins (LCAT, LECT2, SERPINE2, SFTP1, ZP1, ZP3, APOV1, VTG1 and VTG2) at the mRNA and/or protein levels, with dynamic gene expression patterns in the selected chicken lines (except for ZP1; FDR/P ≤ 0.05) and feed deprivation (NAMPT, SFTPA1 and ZP3) (P ≤ 0.05). In contrast, some of the most prominent adipokines in mammals, leptin, TNF, IFNG, and IL6 were expressed at a low level (FPKM/RPKM< 1) and did not show differential mRNA expression neither between broiler and layer lines nor between fed vs. feed-deprived chickens.

Conclusions: Our study revealed that RNA and protein expression in visceral fat changes with selective breeding, suggesting endocrine roles of visceral fat in the selected phenotypes. In comparison to gene expression in visceral fat of mammals, our findings points to a more direct cross talk of the chicken visceral fat with the reproductive system and lower involvement in the regulation of appetite, inflammation and insulin resistance.

Keywords: Adipokines; Adipolin; Adipose tissue; Chickens; Mass spectrometry; PLIN1; PTEN-pathway; RNA-seq; SFTPA1; TNF; Yolk proteins.

Fig. 1

Broiler and layer females differ in growth rate and reproduction efficiency, but not in the accumulation of the visceral fat. a BW measurements were obtained in a follow-up experiment of broiler breeders and layer hens, grown together from hatch with free access to food (n = 50 for each bird’s type; P < 0.001 at all ages including the day of hatch). b The number of large yellow follicle of ≥8 mm were counted in the ovary of the broiler and layer females at the first week of lay (4 month of age; n = 10 for each group). c Percentages of visceral fat (fat weight/live BW) were analyzed in the same birds shown in (b). d Percentages of visceral weight of broiler and layer chickens were measures when reaching 1 Kg BW. n = 10 for each bird’s type. *** P < 0.001

Fig. 2

Differential gene expression in visceral abdominal fat of broiler and layer females. a Venn Diagrams depicts the number of transcripts differentially expressed in broilers (Br) and layers (La) or not differentially expressed (FDR ≤ 0.05; absolute fold change ≥1.5; n = 3 birds per strain). b The PTEN pathway was selected by Ingenuity software as the dominant pathway enriched in broilers compared to layers (Z-score = 3.6; ratio = 0.15; P < 0.001; Additional file 1: Table S3)

Fig. 3

Expression profile of selected adipokines in visceral fat. a Expression profile in broiler and layer females was depicted from our RNA-seq data (Additional file 1: Table S2) calculated as fragments per kilobase of transcript per million mapped reads (FPKM) using edgeR and FDR value with threshold 0.05. Values are expressed as means ± SD. n = 3 in each group. *, FDR ≤ 0.05; **, FDR ≤ 0.01. APLN, IFNG, Leptin (LEP), SFTPA1, TNF, and ZP3, which are missing from the RNA-seq table either due to missing annotation in Galgal5 or low expression level, were searched manually in the RNA-seq data. P values of these transcripts were calculated by student t-test or rank transformed, in cases of lack of normal distribution, and expressed as means ± SE. b and c Expression profiles obtained by blast search using GenBank available experiments in two juvenile broiler experiments: fat and lean lines or high and low growth lines, respectively. The data was calculated as reads per kilobase of transcript per million mapped reads (RPKM). P values were calculated by student t-test and expressed as means ± SE. accession no. are detailed in the Material and Methods section

Fig. 4

Expression profile of genes characterized in visceral fat for the first time and a control gene (RBM28), using the available RNA-seq dataset of red junglefowl. Full-length cDNAs of the indicated genes were used as baits for blast search. Accession no. are detailed in the Material and Methods section. Gray bars indicate expression of ZP1, black bars indicate expression of ZP3, LCAT, LECT2, SFTPA1, SERPINE2, and RBM28

Fig. 5

Response of adipokine expression to feed deprivation. a The effect of 24 h feed deprivation (Fast) on BW of the broiler and layer chickens was calculated as the percent of BW change of the initial BW (before feed deprivation). b The effect of the treatment on relative weight of the visceral fat. c. qPCR analysis of expression of the indicated adipokines. Vertical lines represent ± SE, n = 7 for each bird’s type and treatment, in each of the two biological repeats. Significance was calculated using students t test and denoted as *, P ≤ 0.05; **; P ≤ 0.01; ***, P ≤ 0.005. Long and short horizontal lines indicate comparison between broilers and layers both at free feeding, or between free and deprived feeding, respectively. Primers are listed in Additional file 2: Table S4

Fig. 6

Expression of genes coding for egg-yolk proteins in visceral fat of chicken. a Diagram showing relative abundance of yolk proteins in chicken egg [61]. b Expression profile of yolk proteins identified by MS in visceral fat of broilers (Br) and layers (La) chickens. c Expression profile of yolk proteins’ mRNAs, detected in our RNA-seq experiment (Additional file 1: Table S2)

Fig. 7

Summary of expression analyses of selected fat secreted proteins, based on values of –log[fold change] obtained in the experiments described in Figs. 3 & 5 and Additional file 3: Table S5. Green background indicates statistical significance (FDR / P ≤ 0.05). Red color indicates higher expression in mature broilers and juvenile fat and high-growth lines. Blue color indicates higher expression in mature layers. Empty red and blue boxes represent feed deprived mature broilers and layers, respectively. Doted red color represent juvenile lean line and low-growth line broilers. The figure was modified from Excel (Conditional Formatting)