Skip to main content
NCBI home page
Poultry Science logoLink to Poultry Science
. 2026 Jul 2;105(10):107377. doi: 10.1016/j.psj.2026.107377

Full-length transcriptomic profiling of chicken liver metabolism under different feeding states using nanopore sequencing

Bingjie Xu a, Hui Wang a, Xin Shu a, Li Liu a, Yutong Wang a, Yimeng Li a, Ziqiu Niu a, Ziwei Chen a, Xiaotong Zheng a,b, Jianfei Chen a,b,⁎
PMCID: PMC13356752  PMID: 42419215

Abstract

Feeding governs energy acquisition and drives state-specific hepatic metabolism. As the central metabolic and homeostatic organ in poultry, hepatic molecular mechanisms driving metabolic shifts pre- and post-feeding remain uncharacterized. We performed Oxford Nanopore full-length transcriptome sequencing on liver tissue from 6-week-old broilers under three energy states: ad libitum feeding (AL), fasting (F), and refeeding after fasting (RF). A total of 34,271 genes, 79,282 transcripts (including 6,605 annotated novel transcripts), 287 fusion transcripts, 74,892 simple sequence repeats, and 2,327 long non-coding RNAs (lncRNAs) were identified. Alternative polyadenylation (APA) analysis showed that AL and RF groups preferentially used proximal APA sites relative to the F group. Functional enrichment highlighted lipid and carbohydrate metabolism, autophagy, and mitophagy pathways. Differentially expressed transcripts (DETs) showed more differential events than differentially expressed genes (DEGs), and functional enrichment of DEGs, DETs, and differentially expressed lncRNAs in F vs AL and F vs RF comparisons mainly highlighted lipid, protein, carbohydrate, and xenobiotic metabolism pathways. Notably, enrichment analysis of APA, DEGs, and DETs between AL and RF identified the NOD-like receptor signaling pathway, a core innate immune signaling cascade, as one of the top altered pathways. Consistently, both F and RF groups displayed reduced expression of antimicrobial peptide genes (AvBD1, AvBD6, AvBD7, CATH2, CATH3) relative to the AL group, implying that short-term fasting may transiently suppress hepatic immune defense capacity. This study identifies numerous novel hepatic transcripts and demonstrates that, under distinct energy states, the liver maintains energy homeostasis via transcript-level expression shifts and APA site selection. These findings provide molecular insights into the metabolic responses and hepatic immune to fasting stress, laying a foundation for optimizing feeding regimens and improving liver health in broiler production.

Keywords: Chicken, Full-length transcriptome, Hepatic metabolism, Antimicrobial peptide

Introduction

As the central hub of metabolic homeostasis, the liver performs core physiological functions including substance metabolism, detoxification, bile secretion, and immune regulation, thereby maintaining organismal homeostasis (Trefts et al., 2017; Kawashima et al., 2024; Dukewich et al., 2025). Feeding constitutes the initial step of nutrient acquisition in animals, and the body exhibits distinct energy states before and after food intake (Lambert and Parks, 2012; Zhang et al., 2022). To sustain systemic energy homeostasis, the liver precisely modulates the synthesis and catabolism of nutrients such as glycogen and lipids in response to different energy states (Trefts et al., 2017; Jiang et al., 2025). For instance, cholesterol is a vital lipid regulator essential for biological activities, and the balance of its biosynthesis and metabolism is critical for normal physiological functions (Repa and Mangelsdorf, 2000; Luo et al., 2020). Notably, cholesterol biosynthesis occurs predominantly postprandially, primarily in the liver and intestinal tissues (Edwards et al., 1972; Lu et al., 2020).

Systematic studies over the past decade have demonstrated that intermittent fasting confers beneficial effects on multiple health parameters, lifespan, and cognition (Sutton et al., 2018; Reddy et al., 2024; Li et al., 2025). Intermittent fasting may even exert protective and therapeutic effects against certain cancers (Clifton et al., 2021). In avian research, intermittent fasting initiated in the second week post-hatch promotes vascular development, increases the number of subcutaneous adipocytes, and prevents white leg syndrome in broilers without significantly compromising growth performance (Ayansola et al., 2022). During summer heat stress, a 6-hour daily feed withdrawal from 10:00 to 16:00 during the final week before slaughter effectively alleviates heat stress, regulates rectal temperature, and maintains final body weight and feed conversion ratio in broilers (Ozkan et al., 2003). A 24-hour fasting period elevates serum free fatty acids and glucagon while reducing triglycerides, leptin, and insulin levels in geese, with these parameters largely restored upon refeeding (Liu et al., 2024). Although short-term fasting studies in poultry have indicated certain production benefits, relevant investigations remain limited.

In recent years, third-generation sequencing technology has developed rapidly. Compared with second-generation sequencing, which requires DNA fragmentation, third-generation sequencing features ultra-long reads and de novo assembly-free analysis, enabling accurate identification of alternative splicing (AS) and genetic variations (Ament et al., 2025). To date, third-generation full-length transcriptome sequencing has been applied in several agricultural animal studies, including cattle (Wang et al., 2024; Cao et al., 2025), sheep (Wang et al., 2024), pig (Song et al., 2024), and silkworm (Tang et al., 2024). Relevant reports also exist in poultry research, such as full-length transcriptome analysis of seven tissues in Lushi chickens (Tian et al., 2025), as well as studies on follicular development in chickens (Li et al., 2023; Zhong et al., 2023) and Muscovy ducks (Lin et al., 2021; Niu et al., 2025) and duck embryonic fibroblasts (Wu et al., 2024), all based on third-generation sequencing.

The liver is the primary organ responsible for energy metabolism and anabolism in animals. Investigating hepatic responses under distinct energy states contributes to a better understanding of the molecular mechanisms underlying liver metabolism. Short-read transcriptomic analyses have widely explored hepatic molecular profiles in chickens (Lindholm et al., 2022; Chen et al., 2024a; Zhang et al., 2025). Nevertheless, such sequencing approaches fail to accurately capture transcriptome-wide alternative splicing, alternative polyadenylation (APA) and full-length long non-coding RNA (lncRNA) transcripts. To date, few long-read third-generation sequencing studies have characterized hepatic regulatory responses under different energy conditions. Given this research gap, we hypothesize that alternative splicing, APA, and lncRNA-mediated regulatory networks orchestrate hepatic metabolic responses across feeding states. Herein, we employ Oxford Nanopore Technology (ONT) to generate full-length transcriptomic landscapes from liver tissues of Arbor Acres (AA) broilers under three typical nutritional treatments: ad libitum (AL) feeding, fasting (F) and refeeding (RF). Our work aims to identify novel molecular regulators and provide mechanistic insights into the transcriptional control of avian hepatic metabolism.

Materials and methods

Animals and sample collection

The animal protocols were approved by the Institutional Animal Care and Use Committee (IACUC) of Jiangsu University of Science and Technology (GQ20230302, Zhenjiang, China). Animal care and handling practices were followed by the IACUC guidelines. Newly hatched male AA broilers (Gallus gallus) were purchased from a commercial supplier. They were reared under standard feeding and husbandry conditions with a 10 h light: 14 h dark photoperiod, with the light phase maintained from 08:00 to 18:00. A starter diet (metabolizable energy: 12.48 MJ/kg, crude protein: 21.06%) was provided before 21 days of age, and a grower diet (metabolizable energy: 13.02 MJ/kg, crude protein: 19.66%) was supplied after 21 days of age. At 6 weeks old, 18 broilers with consistent body weight were randomly divided into three treatment groups (n = 6 per group): the AL group with 18 h free access to feed, the 18 h F group, and the RF group that underwent 18 h fasting followed by 2 h of refeeding. All broilers had free access to water. After treatment, all broilers were euthanized via carbon dioxide asphyxiation. Liver tissues were immediately collected, snap-frozen in liquid nitrogen, and preserved at −80°C for subsequent experiments. Three randomly selected broilers in each group were used for full-length transcriptome sequencing, and four randomly selected individuals per group were applied for reverse transcription real-time PCR (RT-qPCR) verification analysis.

RNA extraction and full-length sequencing

Total RNA was isolated from each tissue sample using the Trizol method. RNA integrity was evaluated by 1% agarose gel electrophoresis, and RNA concentration was determined with a Nanodrop spectrophotometer. Qualified total RNA (250 ng per sample) was used for cDNA library construction with the ONT cDNA-PCR sequencing kit (Benagen SMART cDNA Kit, Wuhan, China). The constructed cDNA libraries were loaded onto R10.4.1 flow cells, and full-length transcriptome sequencing was performed using the PromethION platform (ONT, Oxford, UK).

Read processing and full-length transcriptome assembly

Base calling was conducted using Dorado software to convert offline fast5 data into fastq format. Low-quality reads with a quality score < 10 were removed to obtain the final raw dataset. Pychopper (v2.4.0) was applied to identify full-length sequences from valid sequencing data and trim adapters, barcodes and primer sequences. The raw fastq data were further filtered with NanoFilt (v2.8.0) to eliminate short segments and low-quality reads, and high-quality full-length sequences were retained. The filtered full-length sequences were aligned to the chicken reference genome (GRCg7b) using Minimap2 (v2.17), and alignment statistics were analyzed via Samtools (v1.11). Pinfish (v0.1.0) was used to construct a non-redundant transcript set based on full-length sequences. Subsequently, the reconstructed transcripts were further deduplicated using the merge mode of StringTie (v2.1.4) to acquire final non-redundant transcript sequences.

Annotation of novel transcripts

TransDecoder (v5.5.0) was used to predict the coding regions of novel transcripts. The predicted proteins were subjected to functional annotation across public databases. Seven databases, including Non-Redundant Protein Sequence (Nr), Protein Families (Pfam), Universal Protein Resource (UniProt), Kyoto Encyclopedia of Genes and Genomes (KEGG), EuKaryotic Orthologous Groups (KOG), Gene Ontology (GO), and PATHWAY, were adopted to annotate the predicted coding proteins of transcripts. Hmmscan (v3.3.2) was utilized to align the predicted protein sequences against the animal transcription factor (TF) database for the identification and prediction of transcription factors.

Gene structure analysis

SUPPA2 (v2.3.0) was employed to predict seven types of AS events, including exon skipping (ES), retained intron (RI), alternative 5′ splice site (A5), alternative 3′ splice site (A3), mutually exclusive exon (MX), alternative first exon (AF), and alternative last exon (AL). Combined with gene structure annotation and transcript expression levels, the ΔPSI (dPSI) value was calculated, followed by statistical testing of P-values to identify differentially AS events between groups. FusionSeeker (v1.0.1) was used to screen fusion transcripts. The screening criteria were defined as follows: a single transcript aligned to no fewer than two distinct loci in the reference genome, each locus covering over 10% of the full transcript length, the overall transcript coverage greater than or equal to 99%, and the distance between adjacent alignment loci exceeding 100 kb. Simple sequence repeat (SSR) prediction and identification were performed using MISA (v1.0). The screening thresholds were set as: a minimum of 10 repetitions for mononucleotide repeats, 6 repetitions for dinucleotide repeats, and 5 repetitions for trinucleotide and tetranucleotide repeats. Minimap2 (v2.24-r1122) and Quantifypoly(A) were applied to identify polyadenylation [poly(A)] sites and analyze their genomic distribution and functional annotation. Three tools, CNCI (v2.0), CPC2 (v1.0.1) and PLEK (v1.2), were jointly used to evaluate the coding potential of newly identified transcripts, thereby achieving lncRNA identification.

Identification of differential expression genes and transcripts

Transcript expression quantification was performed using Salmon (v1.4.0), and gene-level expression matrices were generated based on transcript quantification results. The R package DESeq2 (v1.26.0) was used to identify differentially expressed genes (DEGs) and transcripts (DETs, including lncRNAs). The screening threshold were |log2FoldChange| > 1.0 and P < 0.05.

GO, KEGG, and gene set enrichment analysis

GO, KEGG enrichment, and Gene Set Enrichment Analysis (GSEA) was implemented by the clusterProfiler (v3.8.1).

cDNA synthesis, reverse transcription PCR amplification, and reverse transcription real-time PCR

For cDNA synthesis, 1 μg of total RNA was reverse-transcribed using the HiScript III 1st Strand cDNA Synthesis Kit with gDNA wiper (Accurate Biotechnology Co., Ltd., Changsha, China). The synthesized cDNA was diluted at a ratio of 1:5 and stored at −20°C for subsequent RT-PCR and RT-qPCR assays. Primers for fusion genes were designed using Primer Premier 5.0, following the criterion that forward and reverse primers were located in different genes. SSR primers were designed via Primer3. Detailed primer information is listed in Table S1 and S2. One randomly selected sample from the F and AL groups was used for RT-PCR validation. The RT-PCR reaction system and procedures were conducted according to the manufacturer’s instructions of 2 × Taq Plus Master Mix (Vazyme Biotech Co., Ltd, Nanjing, China). Amplified products were detected by 1% agarose gel electrophoresis and visualized. Specific primers for differential expressed lncRNAs (DEIs), DEGs and DETs were designed using Primer Premier 5.0 based on the sequencing data obtained in this study or public sequences. Detailed primer sequences are shown in Table S3–S5 and synthesized by Shangya Biotechnology Co., Ltd. (Hangzhou, Zhejiang, China). TBP was selected as the reference gene (Chen et al., 2024b). RT-qPCR was performed using the SYBR Green Premix Pro Taq HS qPCR Tracking Kit II (Accurate Biotechnology). Each sample was analyzed with three technical replicates, and the RT-qPCR reaction system and experimental procedures referred to a previous study (Chen et al., 2023).

Statistical analysis

RT-qPCR data were analyzed using Bio-Rad CFX96 Manager (v4.1), and the cycle threshold (CT) values were exported to Excel for downstream calculations. The relative gene expression levels were normalized via the 2−△△CT method. Statistical differences among groups were evaluated by one-way ANOVA followed by Tukey’s multiple comparison test, and a P value < 0.05 was considered statistically significant.

Results

Summary of the sequencing results

To investigate metabolic regulation in chicken hepatocytes under distinct feeding states, we performed ONT high-throughput sequencing on chicken liver tissues sampled pre- and post-feeding. Each sample yielded 4,937,148–6,043,435 clean reads (Table S6). Full-length transcript reads numbered 4,009,992–5,024,504, exceeding 80% of clean reads with an average N50 of 1583 bp (Table S7), and the genome mapping rate reached 97.31%–98.24% (Table S8). We clustered full-length reads to generate consensus transcripts, whose N50 values ranged 1793–1976 bp with a maximum length of 10706 bp (Table S9). After deduplication, non-redundant transcripts had an N50 of 3986 bp (Fig. S1A). Comparison against reference annotations identified 6,605 novel transcripts and 4,163 novel genes (Fig. S1B). The predicted CDS of novel transcripts had an N50 of 1306 bp, and we further characterized their classification profile (Fig. S1C, D).

Functional annotation of novel transcripts

To fully characterize the functional information of novel transcripts, a total of 6,605 newly identified transcripts were annotated against eight databases, including GO, KOG, KEGG, Nr, Pfam, Uniprot, and TF (Fig. 1A). In total, 2,362 novel transcripts were successfully annotated. The Nr database presented the largest number of annotated items (2,348), while the TF database showed the fewest (268) (Table S10, 11). The dominant GO terms were enriched in cytoplasm, nucleus, metal ion binding, and ATP binding (Fig. 1B). KEGG-enriched pathways were classified into five major categories: Cellular Processes, Environmental Information Processing, Genetic Information Processing, Metabolism, and Organismal Systems (Fig. 1C). KOG functional annotation revealed that these novel transcripts were mainly assigned to general function prediction only, posttranslational modification, protein turnover and chaperones, intracellular trafficking, secretion and vesicular transport, as well as lipid transport and metabolism (Fig. 1D). Nr homologous species analysis indicated that Gallus gallus occupied the highest proportion (89.06%) (Fig. 1E). By alignment against the animal transcription factor database, 268 transcription factors were predicted from the novel transcripts. The top three transcription factor families were zf-C2H2 (15.67%), Fork_head (12.69%), and bHLH (11.19%), respectively (Fig. 1F, Table S12).

Fig. 1.

Fig 1 dummy alt text

Open in a new tab

Functional annotation of the novel transcripts. (A) Upset of TF, KOG, Pathway, Pfam, KEGG, GO, Uniprot, and Nr annotated novel transcripts. (B) Distribution of GO annotated novel transcripts. (C) Distribution of KEGG annotated novel transcripts. (D) Distribution of KOG annotated novel transcripts. (E) Species distribution of the new transcripts annotated by Nr. (F) The proportion of transcription factor families identified in the new transcripts.

Characterization of AS, fusion, and SSR

To analyze AS events of precursor mRNAs, seven AS patterns were identified, namely A3, A5, AF, AL, MX, RI and SE. Among them, AF and SE were predominant across all samples, accounting for more than 28% and 32% of total AS events, respectively (Fig. S2A–R). The total AS events and the numbers of A3, A5, AF and SE events were significantly higher in the fasting group than in the other groups (Fig. 2A). Distinct AS patterns were detected in multiple key genes involved in metabolism and translational regulation, including CPT1A, CREB3L3, RBM3, EEF1D, ECI2 and PCK2. Functional enrichment analysis of differentially AS under different feeding states revealed that xenobiotic metabolism and metabolic pathways were the most significantly enriched pathways (Fig. 2B, and Fig. S3A, B). Fusion transcripts analysis identified a total of 287 fusion transcripts in all ONT samples, among which 92 fusion transcripts were specifically expressed in the F group (Fig. 2C). Four shared fusion transcripts among all groups were randomly selected for RT-PCR validation, and bands with expected sizes were detected in both F and AL samples (Fig. 2D). Sanger sequencing of PCR products further verified the reliability of ONT sequencing data. A total of 74,892 SSR loci were screened using the MISA pipeline. P1-type SSRs were the most abundant (64.77%), followed by composite SSRs (13.77%) and P3-type SSRs (12.24%), while P6-type SSRs were the rarest (0.17%) (Fig. 2E, Table S13). Four SSR types (P2, P3, P4 and c*) were validated by RT-PCR, and amplified fragments matched the expected sizes (Fig. 2F). Sanger sequencing confirmed the accuracy of the ONT-identified SSRs.

Fig. 2.

Fig 2 dummy alt text

Open in a new tab

Transcriptome structure analysis. (A) Quantitative distribution of AS events. (B) KEGG enrichment analysis of alternative splicing differences between AL and F groups. (C) Venn diagram of the predicted fusion transcripts. (D) RT-PCR validation of fusion genes. 1–4 correspond to the fusion transcripts ENSGALG00010000028ENSGALG00010000029, ENSGALG00010000384ENSGALG00010000123, ENSGALG00010003110ENSGALG00010003817, and ENSGALG00010000255ENSGALG00010014010, respectively. (E) Distribution of SSR transcript types. (F) RT-PCR validation of four types of SSRs (P2, P3, P4, and c*).

Analysis of alternative polyadenylation

Based on the identification and analysis of poly(A) sites, the majority of poly(A) sites were located in the 3′UTR of genes (Fig. 3A). Genes with one poly(A) site accounted for 59.46%, whereas those with more than five poly(A) sites occupied 3.18% (Fig. 3B). Compared with the AL group, the F group exhibited a significantly higher proportion of poly(A) sites in the 3′UTR but markedly lower proportions in the CDS and exon regions (Fig. 3C). In the comparison between the F and RF groups, the RF group showed a notably higher percentage of intronic poly(A) sites, while no significant differences were observed in other regions across pairwise comparisons (Fig. S4A, S5A). Genes in each group displayed distinct preferences for proximal or distal poly(A) site selection. Relative to the F group, the AL and RF groups preferred to utilize proximal APA sites (Fig. 3D, and Fig. S4B, S5B). In the F versus AL comparison, KEGG enrichment analysis revealed that genes with preferential proximal and distal poly(A) usage were both significantly enriched in the PPAR signaling pathway and mitophagy–animal. Moreover, genes favoring distal poly(A) sites presented stronger enrichment significance for the autophagy–animal pathway (Fig. 3E, F). For pairwise comparisons of F vs RF and AL vs RF, genes with differential proximal and distal poly(A) usage were mainly enriched in lipid metabolism-related pathways. Notably, autophagy and mitophagy pathways were more highly activated in the AL and RF groups (Fig. S4C, D). In addition, the NOD-like receptor signaling pathway was significantly enriched in the AL and RF groups (Fig. S5C, D).

Fig. 3.

Fig 3 dummy alt text

Open in a new tab

Alternative polyadenylation analysis. (A) The distribution of poly(A) sites across different regions of the gene. (B) Distribution of the number of poly(A) sites in genes. (C) Distribution of the poly(A) tag across different regions in F vs AL groups. (D) Genes that preferentially utilize proximal and distal poly(A) sites in F vs AL groups. (E) and (F) represent the KEGG enrichment analysis of genes preferentially utilizing proximal and distal poly(A) sites in F vs AL, respectively.

Identification and cis- and trans-regulatory analysis of lncRNAs

Using three prediction models (CNCI, CPC2 and PLEK) and screening their intersection, we ultimately identified 2,327 lncRNAs (Fig. 4A), which were classified into 278 lncRNAs (12.39%), 407 antisense lncRNAs (18.14%), 1,551 intronic lncRNAs (69.12%), and 8 sense lncRNAs (0.36%) (Fig. 4B). DEIs were screened (Fig. 4C, Table S14–S16) to predict their cis- and trans-target genes, followed by functional enrichment analysis of these target genes. KEGG enrichment results showed that the cis-target genes of DEIs in the F vs AL comparison were significantly enriched in fatty acid metabolism, fatty acid degradation, adipocytokine signaling pathway, PPAR signaling pathway, and amino acid metabolism (Fig. 4D). The trans-target genes of DEIs were enriched in fatty acid metabolism, fatty acid degradation, PPAR signaling pathway, steroid biosynthesis, pyruvate metabolism, glycolysis/gluconeogenesis, fatty acid elongation, fatty acid biosynthesis, adipocytokine signaling pathway, biosynthesis of unsaturated fatty acids, and glycerolipid metabolism (Fig. 4E). Moreover, trans-target genes exhibited significantly higher enrichment efficiency than cis-target genes. The enrichment patterns of cis- and trans-lncRNAs in the F vs RF comparison were consistent with those in the F vs AL comparison (Fig. S6A, B). Only two pathways showed significant enrichment for cis-acting differential lncRNAs between the RF and AL groups, while more enriched pathways were observed in trans-acting lncRNAs, which were predominantly concentrated in lipid metabolism pathways (Fig. S6C, D). Nine lncRNAs, including five annotated lncRNAs from public databases and four novel lncRNAs, were randomly selected for RT-qPCR validation. The RT-qPCR results were highly consistent with the transcriptome sequencing data (Fig. 5).

Fig. 4.

Fig 4 dummy alt text

Open in a new tab

Prediction, classification and functional annotation of lncRNAs. (A) Venn diagram of lncRNA predictions by CPC2-CNCI-PLEK. (B) Distribution of lncRNA types. (C) The number of lncRNAs differing between groups. (D) and (E) represent the KEGG enrichment analysis of cis-acting and trans-acting target genes of differentially expressed lncRNAs in F vs AL, respectively.

Fig. 5.

Fig 5 dummy alt text

Open in a new tab

RT-qPCR verification of differentially expressed lncRNAs in liver tissues under three energy states: ad libitum (AL), fasting (F), and refeeding (RF). Different lowercase letters indicate significant differences (P < 0.05).

Analysis of differentially expressed genes

Principal component analysis (PCA) showed that genes under different feeding conditions exhibited distinct clustering. The gene expression patterns differed greatly between the F group and the AL/RF groups, while the AL and RF groups shared similar expression profiles (Fig. 6A, B). A total of 1582 DEGs were identified in the F vs AL comparison (Table S17), including 851 upregulated genes (e.g., ANGPTL3, PPP1R3C, MSMO1) and 731 downregulated genes (e.g., PCK1, LDHB, CPT1A) (Fig. 6C). In the F vs RF comparison, 1120 DEGs were screened (Table S18), with 554 upregulated genes (e.g., ANGPTL3, MSMO1, CYP51A1) and 566 downregulated genes (e.g., CPT1A, ALDOB, C8orf22) (Fig. S7A). Additionally, 421 DEGs were obtained between the AL and RF groups (Table S19), consisting of 270 upregulated genes (e.g., LYG2, CATH2, LECT2) and 151 downregulated genes (e.g., LDHB, HADHB, MPC1L) (Fig. S7D). KEGG enrichment analysis revealed that DEGs from F vs AL and F vs RF comparisons were significantly enriched in multiple metabolic pathways, including fatty acid metabolism, PPAR signaling pathway, fatty acid degradation, metabolism of xenobiotics by cytochrome P450, pyruvate metabolism, biosynthesis of unsaturated fatty acids, glutathione metabolism, glycolysis/gluconeogenesis, and arachidonic acid metabolism (Figs. 6D, and S7B). GSEA-KEGG analysis further demonstrated that compared with the fasting group, both AL feeding and RF activated steroid biosynthesis, steroid hormone biosynthesis, inflammatory mediator regulation of TRP channels, and drug metabolism–cytochrome P450, whereas adipocytokine signaling pathway, PPAR signaling pathway, Staphylococcus aureus infection, and glucagon signaling pathway were universally inhibited (Figs. 6E, and S7C). For the AL vs RF group, beyond metabolic pathways, the NOD-like receptor signaling pathway represented the most significantly enriched KEGG term in both KEGG and GSEA analyses (Fig. S7E, F). 12 DEGs were randomly selected for RT-qPCR validation, covering energy metabolism-related genes (PCK1, LDHB, CPT1A, HADHB, ECI2, MSMO1), xenobiotic metabolism-related genes (ADH1C, GSTA2, GSTA3), amino acid metabolism gene (PSPH), and antimicrobial peptide genes (CATH2, AvBD1). The RT-qPCR results were highly consistent with ONT sequencing data (Fig. 7).

Fig. 6.

Fig 6 dummy alt text

Open in a new tab

Transcriptomic analysis of differentially expressed genes (DEGs). (A) Principal component analysis. (B) Heatmaps of DEGs across different groups. (C) Volcano plot of DEGs between AL and F. (D) KEGG enrichment analysis of DEGs between AL and F. (E) GSEA-KEGG enrichment analysis of DEGs between AL and F.

Fig. 7.

Fig 7 dummy alt text

Open in a new tab

RT-qPCR verification of differentially expressed genes in liver tissues under three energy states: ad libitum (AL), fasting (F), and refeeding (RF). Different lowercase letters indicate significant differences (P < 0.05).

Analysis of differentially expressed transcripts

PCA results showed that transcripts under different feeding conditions presented distinct clustering. The transcript expression patterns differed markedly between the F group and the AL/RF groups, while the AL and RF groups shared similar expression profiles (Fig. 8A, B). A total of 2,207 DETs were identified in the F vs AL comparison (Table S20), including 1198 upregulated transcripts (e.g., ANGPTL3, PPP1R3C, MSMO1) and 1,009 downregulated transcripts (e.g., PCK1, LDHB, CPT1A) (Fig. 8C). For the RF vs F comparison, 1,611 DETs were detected (Table S21), with 815 upregulated transcripts (e.g., ANGPTL3, CYP51A1, MSMO1) and 796 downregulated transcripts (e.g., CPT1A, ALDOB, MOGAT1) (Fig. S8A). In addition, 594 DETs were screened between the RF and AL groups (Table S22), consisting of 358 upregulated transcripts (e.g., LYG2, CATH2, LECT2) and 236 downregulated transcripts (e.g., LDHB, PGK2, HADHB) (Fig. S8D). KEGG and GSEA analyses of DETs from F vs AL and F vs RF comparisons revealed significant enrichment in multiple metabolic pathways, including the PPAR signaling pathway, fatty acid degradation, and metabolism of xenobiotics by cytochrome P450 (Figs. 8D, E, and S8B, C). Beyond metabolic pathways, the NOD-like receptor signaling pathway was the most significantly enriched KEGG pathway in the AL vs RF group (Fig. S8E, F). Six DETs covering energy metabolism (DECR1), xenobiotic metabolism (GSTT1), autophagy (NBR1), and other regulatory functions (PAFAH1B1, RBM3, EEF1D) were selected for RT-qPCR validation. Although the GSTT1 gene showed quantitative differences in expression values measured by RT-qPCR and ONT sequencing, the two methods yielded identical expression trends. The expression patterns of all other genes were highly consistent (Fig. 9).

Fig. 8.

Fig 8 dummy alt text

Open in a new tab

Transcriptomic analysis of differentially expressed transcripts (DETs). (A) Principal component analysis. (B) Heatmap of DETs between groups. (C) Volcano plot of DETs between AL and F. (D) KEGG enrichment analysis of DETs between AL and F. (E) GSEA-KEGG enrichment analysis of DETs between AL and F.

Fig. 9.

Fig 9 dummy alt text

Open in a new tab

RT-qPCR verification of differentially expressed transcripts (t) and their relative genes (g) in liver tissues under three energy states: ad libitum (AL), fasting (F), and refeeding (RF). Different lowercase letters indicate significant differences (P < 0.05).

Discussion

Long-term fasting protocols, such as conventional forced molting, are well-established in poultry research. Usually lasting 7–15 days and involving water restriction. Numerous avian studies have validated that forced molting remarkably elevates productive traits, including prolonging the laying cycle of aged hens (Wang et al., 2023), enhancing sperm motility in breeder roosters (Zhu et al., 2023), and promoting bone formation (Zhang et al., 2025). By contrast, research regarding short-term fasting remains scarce, and its underlying molecular mechanisms have yet to be thoroughly elucidated. The liver serves as the core metabolic organ responsible for lipids and exogenous xenobiotics in poultry (Leveille et al., 1968; Anderson and Hammes, 1985; Chen et al., 2021). In the present study, ONT was adopted to investigate hepatic transcript expression profiles of broilers under three different feeding conditions. AL feeding is a conventional rearing pattern for broilers. Three comparative analyses were conducted to explore the underlying biological characteristics. Specifically, the comparison between the F and AL groups clarified the impacts of short-term fasting on hepatic energy metabolism. The comparison between the F and RF groups uncovered the disorder of hepatic gene expression triggered by excessive energy intake during refeeding. The comparison between the AL and RF groups further demonstrated the regulatory roles of short-term fasting followed by refeeding in modulating liver gene expression in broilers.

All samples in this study yielded sufficient valid sequencing data, with the proportion of full-length reads steadily exceeding 80% and an average N50 length of 1583 bp, which confirms high data integrity and prominent long-read advantages. Compared with conventional second-generation short-read sequencing, ONT sequencing avoids transcript structural errors caused by assembly defects of short reads (< 200 bp). These findings provide a solid foundation for the subsequent accurate identification of alternative splicing isoforms and novel transcripts.

Based on full-length transcriptomic data, a total of 6,605 novel transcripts were identified in this study, among which only 2,362 genes were successfully annotated. A total of 64.2% of the identified transcripts are unannotated, predominantly corresponding to novel splice isoforms (including novel lncRNAs), APA variants and low-abundance transcripts specific to our experimental treatment. Moreover, a considerable number of novel transcripts exhibited differential expression before and after feeding. Several novel transcripts were randomly selected and verified via RT-qPCR, and their actual expression was confirmed. Collectively, the functions of these unannotated novel transcripts warrant further investigation in subsequent studies.

In AS analysis, the numbers of alternative splicing events such as A3 and A5 were significantly higher in the fasting group than in the AL and RF groups. Functional enrichment analysis of differentially alternative spliced genes revealed that the primary enriched pathway was the exogenous drug metabolism pathway, and the expression levels of these genes were largely upregulated after feeding. This suggests that the liver mediates the metabolism of exogenous substances partially by regulating alternative splicing events. In the identification of fusion genes in liver tissues, no abundant fusion transcripts were detected in the present study. A total of 38 fusion transcripts were commonly present in all nine ONT samples, among which seven were specifically expressed in the F group and exclusively induced under fasting conditions. Consistent with this finding, several alternative splicing patterns (e.g., A3 and A5 events) were also markedly elevated upon hunger stress. The biological roles of these splicing variants and fusion transcripts under low-energy conditions remain unclear and require further experimental exploration. Furthermore, a total of 74,892 SSRs were identified and partially validated by RT-qPCR. These specific SSRs can serve as potential genetic markers for AA broilers and provide valuable resources for subsequent genetic breeding research.

APA is an essential post-transcriptional modification mechanism for mRNA maturation in eukaryotes (Turner et al., 2018; Wang et al., 2018), and it plays critical regulatory roles in multiple biological processes, including cell proliferation, differentiation, development and immune response (Sandberg et al., 2008; Jia et al., 2017; Yang et al., 2026). Compared with the F group, genes preferentially utilizing distal poly(A) sites in the AL group were significantly enriched in autophagy and mitophagy-related pathways, while those favoring proximal poly(A) sites were mainly enriched in the mitophagy pathway. Under AL feeding conditions, autophagy-related genes (including ATG7, ATG12, ATG9A, PIK3C3, WIP11, and WIP12) tended to adopt distal poly(A) sites. The resultant lengthened 3′UTR may introduce additional binding sites for microRNAs and RNA-binding proteins, thereby reducing mRNA stability or inhibiting protein translation. Previous studies have demonstrated that mild fasting can trigger autophagy in organisms (Abdalla et al., 2026). We hypothesize that fasting likely promotes the usage of proximal poly(A) sites in autophagy-related genes, shortens the 3′UTR length, and enhances mRNA stability and translation efficiency, ultimately inducing autophagy. Notably, fasting treatment is commonly involved in artificially induced molting, which shares the same stress essence with the fasting model established in this study. It has been reported that artificial molting activates the AMPK–lipophagy pathway and improves intestinal function and laying performance in laying hens (Lv et al., 2025). Accordingly, the APA-mediated autophagy regulatory mechanism identified herein may hypothetically participate in the forced molting process of poultry, providing novel insights and potential molecular targets for further exploring the mechanisms underlying improved production traits during forced molting. APA may serve as one of the key explanations for the dissociation between unchanged gene expression levels and altered protein functions. lncRNAs exert crucial regulatory effects on gene expression and cellular homeostasis (Grammatikakis and Lal, 2022). In the nucleus, lncRNAs predominantly modulate gene expression through cis- and trans-acting regulation (Gil and Ulitsky, 2020). In the present study, multiple DEIs were identified, which regulated the expression of functional target genes via cis or trans mechanisms. GPX4, an antioxidant peroxidase, reduces cholesterol peroxides, thymine peroxides and fatty acid peroxides to protect cells against oxidative damage (Xie et al., 2023). The differentially expressed lncRNA ENSGALT00010015845 may modulate GPX4 expression. Significant upregulation of GPX4 in the AL group further participates in hepatic oxidative stress responses. ACACA is a key lipogenic enzyme involved in de novo fatty acid synthesis; enhanced ACACA activity facilitates adipocyte differentiation and lipid accumulation (Dong et al., 2024). Our results indicated that ACACA expression is potentially regulated by lncRNA ENSGALT00010015845, thereby mediating hepatic fatty acid metabolism. In total, 2,327 lncRNAs were identified in this study, among which 2,236 were novel transcripts absent from the reference genome, including novel3930, novel3980, and novel2891. Bioinformatic prediction suggested that these novel lncRNAs could interact with multiple core metabolic genes, though such interactions remain to be experimentally validated. Currently, the biological functions and underlying regulatory mechanisms of these newly identified lncRNAs are largely uncharacterized, and their roles in physiological and pathological processes require further experimental exploration. DEIs also serve as essential regulators governing hepatic gene expression alterations before and after feeding. Additionally, lncRNAs participate in multiple post-transcriptional regulatory mechanisms, such as the ceRNA sponge effect. Further integration of miRNA data will help elucidate the comprehensive functions of these lncRNAs in subsequent investigations.

Functional enrichment analyses of DEGs, DETs, and differentially expressed isoforms in the liver before and after feeding were mainly enriched in lipid metabolism, carbohydrate metabolism, xenobiotic metabolism, and amino acid metabolism. The differences in energy metabolism were predictable, as the liver undergoes distinct energy states under varying feeding conditions. Representative key genes involved in lipid and glucose metabolism included CPT1A, FASN, PGK2, and LDHB. Critical regulators also covered the lipid metabolism-related transcription factor CREB3L3 and xenobiotic metabolism genes such as ADH1C, GSTA2, and GSTA3. Furthermore, LEAP2, a recently identified liver and intestine highly expressed gene closely associated with feeding states in chickens, was also differentially expressed in the present study (Zheng et al., 2022; Chen et al., 2026). Notably, comparative analyses of APA profiles, DEGs and DETs between the F and RF groups revealed significant enrichment of the NOD-like receptor signaling pathway, a canonical upstream immune regulatory cascade that activates the NF-κB and MAPK signaling pathways (Liu et al., 2019). Multiple differentially expressed genes and transcripts were enriched in this pathway, including AvBD6, AvBD7, CATH3, CASP14, and MAPK12. Both AvBD6 and AvBD7 encode typical avian antimicrobial peptides (Xu et al., 2025), whose expression is modulated by the NF-κB signaling pathway (Lee et al., 2016), while MAPK12 acts as a core component of the MAPK cascade. These results indicated that pronounced activation of the NOD-like receptor pathway substantially altered the expression of hepatic immune-related genes. Expression profiling demonstrated that refeeding after fasting markedly suppressed the NOD-like receptor signaling pathway relative to the fasting group, suggesting that refeeding following short-term fasting reshapes hepatic immune capacity in broilers. Accumulating evidence has confirmed the beneficial effects of mild intermittent fasting on physiological health. However, the present findings indicated that refeeding after food deprivation inhibits the expression of antimicrobial peptide genes, which may be attributed to the prolonged fasting duration applied in this study, as conventional mild fasting typically lasts approximately 15 h. Alternatively, mild fasting may exert short-term adverse effects on hepatic immune defense in poultry compared with AL feeding. Further investigations are required to verify this hypothesis in future research.

The number of DETs was higher than that of DEGs. Although functional enrichment results were generally comparable between DEGs and DETs, DETs exhibited more significant enrichment. This phenomenon may be explained by the overall insignificant expression changes of genes in certain pathways, whereas specific transcripts derived from these genes were differentially expressed, thereby indirectly enhancing the enrichment significance. This viewpoint was well validated through transcript-level and gene-level quantification via RT-qPCR (Fig. 9). Collectively, alternative splicing modulation serves as a vital strategy for the chicken liver to regulate systemic energy homeostasis in response to altered feeding states. In addition, the RF group exhibited expression patterns of metabolic genes highly similar to the AL group (r = 0.98 for PPAR pathway genes and r = 0.79 for fatty acid metabolism genes). Nevertheless, significant differences were detected in antimicrobial peptide genes (AvBD1, AvBD6, AvBD7, CATH2, CATH3), F-specific fusion transcripts, and lncRNAs between F vs AL and RF vs AL comparisons. For these immune-related genes, the RF group exhibited expression profiles similar to the F group (r = 0.84), reflecting partial transcriptional consistency between RF and F. These findings suggest that the response of immune-related genes such as antimicrobial peptides lags behind that of metabolic genes.

In the present study, liver samples were harvested at 2 h post-refeeding after an 18 h fasting period. Under this acute nutritional shift, hepatic energy metabolism may be preferentially activated, whereas the transcriptional regulation of some immune-related genes is. This timing difference may partly explain why antimicrobial peptide genes did not exhibit significant upregulation at our sampling time point; extended sampling windows may allow these genes to rebound to expression levels comparable to those in AL-fed broilers. Consistent with this notion, previous in vitro data demonstrated only a marginal elevation of antimicrobial peptides at 3 h post-stimulation, followed by sustained high expression from 6 to 24 h in HD11 cells and primary monocytes (Sunkara et al., 2011). Likewise, significant induction of AvBD9 and CATHB1 was not observed until 3 h after butyrate or glucose treatment in HD11 macrophages, with expression remaining unchanged at 2 h and peaking at 6 h (Yang et al., 2021). Additional in vivo evidence further supports that immune-related genes typically require more than 3 h to be transcriptionally induced; for instance, AvBD expression was markedly increased only at 3 days post-infection in chicken cecal tonsils (Crhanova et al., 2011). Subsequent studies with serial sampling time points are required to verify the delayed transcriptional induction of hepatic antimicrobial peptide genes upon fasting and refeeding. In addition, the present work only examined hepatic transcriptomic changes triggered by a single short-term fasting challenge. Although second-generation sequencing has revealed hepatic gene expression alterations under prolonged fasting, relevant evidence from third-generation full-length transcriptome sequencing is still lacking, which merits further comprehensive exploration in future research.

In this study, several DETs were identified that were not detected as DEGs. Partial DETs were further validated by RT-qPCR. Although GSTT1 gene showed significant difference in RT-qPCR, it showed no significant changes in ONT transcriptome sequencing. This discrepancy stemmed from differing analytical thresholds rather than experimental errors. DEGs from ONT data were screened by |log2FoldChange| > 1.0 and P < 0.05, and GSTT1 failed the fold-change requirement despite having a low P value. In contrast, RT-qPCR analyzed by ANOVA adopted no fold-change cutoff, with P < 0.05 defined as significant. In fact, GSTT1 exhibited identical expression trends in both tests. The remaining genes displayed strong concordance between the two platforms, confirming the reliability of these DETs. These results indicate that gene function is not solely determined by total gene expression; the differential regulation of distinct transcripts also acts as a crucial regulatory mechanism. Under varying energy conditions, the liver remodels metabolic processes by modulating the abundance of specific transcripts. For certain genes, this modulation occurs without altering overall gene expression and is achieved through the differential expression of individual transcript isoforms. This finding provides a valuable reference for exploring molecular mechanisms underlying complex metabolic adaptation. It also demonstrates the unique advantages of full-length transcriptome sequencing in accurately characterizing transcript expression regulation.

Consistent with previous short-read transcriptome research on chicken liver under different feeding conditions (Lindholm et al., 2022), we identified similar metabolic genes and pathways using ONT long-read sequencing. Importantly, our full-length transcriptome captured numerous novel full-length transcripts, which provide new candidate targets for future research on hepatic metabolism. With the rapid advancement of the FarmGTEx initiative, the ChickenGTEx project has established a comprehensive atlas of genetic variation and gene expression in chickens (Guan et al., 2025). Herein, we generated third-generation RNA-seq data from chicken liver tissues, which serves as a valuable supplement to ChickenGTEx and can be integrated into its future updates.

In conclusion, this study systematically characterized the hepatic transcriptomic profiles of AA broilers under three different feeding conditions using ONT sequencing. Integrated analyses covering full-length transcript identification (including novel transcripts, lncRNAs, and fusion transcripts), alternative splicing events, and differential polyadenylation site usage collectively revealed the complex transcriptional regulatory network underlying hepatic adaptation to nutritional fluctuations. Moreover, genes associated with carbohydrate, lipid and xenobiotic metabolism exhibited the most prominent differential expression during feeding transitions. In contrast, immune defense genes displayed delayed expression changes after short-term fasting followed by refeeding, retaining expression patterns similar to those in the fasting group. Collectively, these findings provide abundant candidate targets for further exploring the molecular mechanisms governing hepatic metabolism and immune defense in chickens.

CRediT authorship contribution statement

Bingjie Xu: Writing – original draft, Methodology, Formal analysis, Investigation. Hui Wang: Formal analysis, Investigation. Xin Shu: Formal analysis. Li Liu: Formal analysis. Yutong Wang: Investigation. Yimeng Li: Investigation. Ziqiu Niu: Formal analysis. Ziwei Chen: Investigation. Xiaotong Zheng: Conceptualization. Jianfei Chen: Writing – original draft, Conceptualization, Supervision.

Disclosures

The authors declare no conflict of interest.

Acknowledgments

This work was supported by the Natural Science Foundation of China (32302725), the Natural Science Foundation of Jiangsu Province (BK20220648), the Natural Science Research of Jiangsu Higher Education Institutions of China (24KJB230001), and the Postgraduate Research & Practice Innovation Program of Jiangsu Province (SJCX25_2617, SJCX24_2575).

Footnotes

Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.psj.2026.107377.

Appendix. Supplementary materials

mmc1.docx (3.6MB, docx)
mmc2.xlsx (2.1MB, xlsx)

Data availability

The RNA-seq data from this study have been submitted in the Genome Sequence Archive in the National Genomics Data Center under accession number CRA040380.

References

  1. Abdalla M.M.I., Bhatnagar P., Zulkaflee M.H.B., Mohammed Irfan A.M.S., Eid N. Intermittent fasting enhances cancer therapyvia autophagy-dependent and independent mechanisms: evidence and implications. World J. Clin. Oncol. 2026;17(2) doi: 10.5306/wjco.v17.i2.115289. CrossRef [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Ament I.H., DeBruyne N.., Wang F., Lin L. Long-read RNA sequencing: a transformative technology for exploring transcriptome complexity in human diseases. Mol. Ther.: J. Am. Soc. Gene Ther. 2025;33(3):883–894. doi: 10.1016/j.ymthe.2024.11.025. CrossRef [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Anderson V.E., Hammes G.G. Distribution of reaction intermediates on chicken liver fatty acid synthase. Biochem.-Us. 1985;24(9):2147–2154. doi: 10.1021/bi00330a007. CrossRef [DOI] [PubMed] [Google Scholar]
  4. Ayansola H., Luo Y., Wan Y., Yu X., Lei J., Yu K., Liao C., Guo Y., Zhang B., Wang B. Restricted feeding regimens improve white striping associated muscular defects in broiler chickens. Anim. Nutr. 2022;12:128–137. doi: 10.1016/j.aninu.2022.09.009. CrossRef [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Cao M., Ding Z., Wang X., Guo S., Kang Y., Hu L., Zhang B., Pei J., Ma Y., Guo X. Full-length transcriptome sequencing of the longissimus dorsi muscle of yak and cattle-yak using nanopore technology. Int. J. Biol. Macromol. 2025;284 doi: 10.1016/j.ijbiomac.2024.138071. CrossRef [DOI] [PubMed] [Google Scholar]
  6. Chen J., Hua G., Han D., Zheng X., Dong X., Wang S., Long J., Zheng Z., Wang A., Wang J., Wang X., Deng X. An EAV-HP insertion in the promoter region of SLCO1B3 has pleiotropic effects on chicken liver metabolism based on the transcriptome and proteome analysis. Sci. Rep.-Uk. 2021;11(1):7571. doi: 10.1038/s41598-021-87054-9. CrossRef [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Chen J., Chen Z., Zhang J., Zhuang W., Zheng X. Screening of reliable reference genes for the normalization of RT-qPCR in chicken gastrointestinal tract. Poult. Sci. 2023;102(12) doi: 10.1016/j.psj.2023.103169. CrossRef [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Chen Z., Hua G., Shu X., Zhuang W., Zhang J., Zhu R., Zheng X., Chen J. Screening of reliable reference genes for the normalization of RT-qPCR in chicken liver tissues and LMH cells. Sci. Rep.-Uk. 2024;14(1) doi: 10.1038/s41598-024-68752-6. CrossRef [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Chen Z., Liu L., Shu X., Wang H., Xu B., Zhang J., Wang M., Shen M., Zheng X., Chen J. Changes of chicken liver-enriched antimicrobial peptide 2 across feeding states and body weight and its regulatory role in feed intake. Brit Poult. Sci. 2026;67(1):151–158. doi: 10.1080/00071668.2025.2527227. CrossRef [DOI] [PubMed] [Google Scholar]
  10. Chen Z., Zheng X., Shu X., Hua G., Zhu R., Sun L., Chen J. Supplemental L-arginine promotes hepatocyte proliferation and alters liver fatty acid metabolism in the late embryonic phase: an RNA-seq analysis. Poult. Sci. 2024;103(11) doi: 10.1016/j.psj.2024.104175. CrossRef [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Clifton K.K., Ma C..X., Fontana L., Peterson L.L. Intermittent fasting in the prevention and treatment of cancer. CA: Cancer J. Clin. 2021;71(6):527–546. doi: 10.3322/caac.21694. [DOI] [PubMed] [Google Scholar]
  12. Crhanova M., Hradecka H., Faldynova M., Matulova M., Havlickova H., Sisak F., Rychlik I. Immune response of chicken gut to natural colonization by gut microflora and to Salmonella enterica serovar enteritidis infection. Infect. Immun. 2011;79(7):2755–2763. doi: 10.1128/IAI.01375-10. CrossRef [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Dong J., Li M., Peng R., Zhang Y., Qiao Z., Sun N. ACACA reduces lipid accumulation through dual regulation of lipid metabolism and mitochondrial function via AMPK- pparα- CPT1A axis. J. Transl. Med. 2024;22(1):196. doi: 10.1186/s12967-024-04942-0. CrossRef [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Dukewich M., Yuan L., Terrault N.A. At the crossroads of health and disease: consequences of fat in the liver. Annu. Rev. Physiol. 2025;87(1):325–352. doi: 10.1146/annurev-physiol-022724-105515. CrossRef [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Edwards P.A., Muroya H.., Gould R.G. In vivo demonstration of the circadian rhythm of cholesterol biosynthesis in the liver and intestine of the rat. J. Lipid Res. 1972;13(3):396–401. CrossRef [PubMed] [Google Scholar]
  16. Gil N., Ulitsky I. Regulation of gene expression by cis-acting long non-coding RNAs. Nature reviews. Genetics. 2020;21(2):102–117. doi: 10.1038/s41576-019-0184-5. CrossRef [DOI] [PubMed] [Google Scholar]
  17. Grammatikakis I., Lal A. Significance of lncRNA abundance to function. Mamm. Genome. 2022;33(2):271–280. doi: 10.1007/s00335-021-09901-4. CrossRef [DOI] [PubMed] [Google Scholar]
  18. Guan D., Bai Z., Zhu X., Zhong C., Hou Y., Zhu D., Li H., Lan F., Diao S., Yao Y., Zhao B., Li X., Pan Z., Gao Y., Wang Y., Zou D., Wang R., Xu T., Sun C., Yin H., Teng J., Xu Z., Lin Q., Shi S., Shao D., Degalez F., Lagarrigue S., Wang Y., Wang M., Peng M., Rocha D., Charles M., Smith J., Watson K., Buitenhuis A.J., Sahana G., Lund M.S., Warren W., Frantz L., Larson G., Lamont S.J., Si W., Zhao X., Li B., Zhang H., Luo C., Shu D., Qu H., Luo W., Li Z., Nie Q., Zhang X., Xiang R., Liu S., Zhang Z., Zhang Z., Liu G.E., Cheng H., Yang N., Hu X., Zhou H., Fang L. Genetic regulation of gene expression across multiple tissues in chickens. Nat. Genet. 2025;57(5):1298–1308. doi: 10.1038/s41588-025-02155-9. CrossRef [DOI] [PubMed] [Google Scholar]
  19. Jia X., Yuan S., Wang Y., Fu Y., Ge Y., Ge Y., Lan X., Feng Y., Qiu F., Li P., Chen S., Xu A. The role of alternative polyadenylation in the antiviral innate immune response. Nat. Commun. 2017;8 doi: 10.1038/ncomms14605. CrossRef [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Jiang C., Li P., Cao H. Unveiling tissue-specific RNA landscapes in mouse organs during fasting and feeding using nanopore direct RNA sequencing. Adv. Sci. 2025;12(5) doi: 10.1002/advs.202408054. CrossRef [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Kawashima K., Andreata F., Beccaria C.G., Iannacone M. Priming and maintenance of adaptive immunity in the liver. Annu. Rev. Immunol. 2024;42(1):375–399. doi: 10.1146/annurev-immunol-090122-041354. CrossRef [DOI] [PubMed] [Google Scholar]
  22. Lambert J.E., Parks E.J. Postprandial metabolism of meal triglyceride in humans. Biochim. Biophys. Acta (BBA) - Mol. Cell Biol. Lipids. 2012;1821(5):721–726. doi: 10.1016/j.bbalip.2012.01.006. CrossRef [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Lee S.I., Jang H..J., Jeon M., Lee M.O., Kim J.S., Jeon I., Byun S.J. Transcriptional regulation of cathelicidin genes in chicken bone marrow cells. Poult. Sci. 2016;95(4):912–919. doi: 10.3382/ps/pev361. CrossRef [DOI] [PubMed] [Google Scholar]
  24. Leveille G.A., O'Hea E..K., Chakbabarty K. In vivo lipogenesis in the domestic chicken. Proc. Soc. Exp. Biol. Med., Soc. Exp. Biol. Med. (N. Y. N.Y.) 1968;128(2):398–401. doi: 10.3181/00379727-128-33022. CrossRef [DOI] [PubMed] [Google Scholar]
  25. Li D., Zhong C., Sun Y., Kang L., Jiang Y. Identification of genes involved in chicken follicle selection by ONT sequencing on granulosa cells. Front. Genet. 2023;13 doi: 10.3389/fgene.2022.1090603. CrossRef [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Li T., Huang N., Chen H., Yang Y., Zhang J., Xu W., Gong H., Gong C., Yang M., Zhao T., Wang F., Xiao H. Daytime-restricted feeding alleviates D-galactose-induced aging in mice and regulates the AMPK and mTORC1 activities. J. Cell Physiol. 2025;240(3) doi: 10.1002/jcp.70020. CrossRef [DOI] [PubMed] [Google Scholar]
  27. Lin J., Guan L., Ge L., Liu G., Bai Y., Liu X. Nanopore-based full-length transcriptome sequencing of Muscovy duck (Cairina moschata) ovary. Poult. Sci. 2021;100(8) doi: 10.1016/j.psj.2021.101246. CrossRef [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Lindholm C., Batakis P., Altimiras J., Lees J. Intermittent fasting induces chronic changes in the hepatic gene expression of Red Jungle Fowl (Gallus gallus) Bmc Genom. 2022;23(1):304. doi: 10.1186/s12864-022-08533-5. CrossRef [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Liu P., Lu Z., Liu L., Li R., Liang Z., Shen M., Xu H., Ren D., Ji M., Yuan S., Shang D., Zhang Y., Liu H., Tu Z. NOD-like receptor signaling in inflammation-associated cancers: from functions to targeted therapies. Phytomed.: Int. J. Phytother. Phytopharm. 2019;64 doi: 10.1016/j.phymed.2019.152925. CrossRef [DOI] [PubMed] [Google Scholar]
  30. Liu Y., Wang X., Li G., Chen S., Jia H., Dai J., He D. Investigating the impact of fasting and refeeding on blood biochemical indicators and transcriptional profiles in the hypothalamus and subcutaneous adipose tissue in geese. Animals. 2024;14(18):2746. doi: 10.3390/ani14182746. CrossRef [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Lu X., Shi X., Hu A., Wang J., Ding Y., Jiang W., Sun M., Zhao X., Luo J., Qi W., Song B. Feeding induces cholesterol biosynthesis via the mTORC1-USP20–HMGCR axis. Nature. 2020;588(7838):479–484. doi: 10.1038/s41586-020-2928-y. CrossRef [DOI] [PubMed] [Google Scholar]
  32. Luo J., Yang H., Song B. Mechanisms and regulation of cholesterol homeostasis. Nature reviews. Mol. Cell Biol. 2020;21(4):225–245. doi: 10.1038/s41580-019-0190-7. CrossRef [DOI] [PubMed] [Google Scholar]
  33. Lv M., Liu C., Jiang X., Ge M., Wang H., Yu W. Molting of laying hens can activate AMPK- lipophagy - lipid metabolism pathway and improve intestinal digestion and absorption. Poult. Sci. 2025;104(1) doi: 10.1016/j.psj.2024.104641. CrossRef [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Niu Y., Yang X., Wang Y., Zhu L., Hu Z., Wang X. Dynamic alternative polyadenylation in ovaries of Muscovy ducks across different laying stages based on full-length transcriptomics. Poult. Sci. 2025;104(12) doi: 10.1016/j.psj.2025.106069. CrossRef [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Ozkan S., Akba Y., Altan O., Altan A., Ayhan V., Ozkan K. The effect of short-term fasting on performance traits and rectal temperature of broilers during the summer season. Brit Poult. Sci. 2003;44(1):88–95. doi: 10.1080/0007166031000085292. CrossRef [DOI] [PubMed] [Google Scholar]
  36. Reddy B.L., Reddy V..S., Saier Jr M.H. Health benefits of intermittent fasting. Microb. Physiol. 2024;34(1):142–152. doi: 10.1159/000540068. CrossRef [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Repa J.J., Mangelsdorf D.J. The role of orphan nuclear receptors in the regulation of cholesterol homeostasis. Annu. Rev. Cell Dev. Bi. 2000;16:459–481. doi: 10.1146/annurev.cellbio.16.1.459. CrossRef [DOI] [PubMed] [Google Scholar]
  38. Sandberg R., Neilson J.R., Sarma A., Sharp P.A., Burge C.B. Proliferating cells express mRNAs with shortened 3′ untranslated regions and fewer microRNA target sites. Science. 2008;320(5883):1643–1647. doi: 10.1126/science.1155390. CrossRef [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Song Q., Li J., Li S., Cao H., Jin X., Zeng Y., Chen W. Full-length transcriptome analysis of skeletal muscle of Jiangquan black pig at different developmental stages. Int. J. Mol. Sci. 2024;25(11):6095. doi: 10.3390/ijms25116095. CrossRef [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Sunkara L.T., Achanta M.., Schreiber N.B., Bommineni Y.R., Dai G., Jiang W., Lamont S., Lillehoj H.S., Beker A., Teeter R.G., Zhang G. Butyrate enhances disease resistance of chickens by inducing antimicrobial host defense peptide gene expression. Plos One. 2011;6(11) doi: 10.1371/journal.pone.0027225. CrossRef [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Sutton E.F., Beyl R.., Early K.S., Cefalu W.T., Ravussin E., Peterson C.M. Early time-restricted feeding improves insulin sensitivity, blood pressure, and oxidative stress even without weight loss in men with prediabetes. Cell Metab. 2018;27(6):1212–1221. doi: 10.1016/j.cmet.2018.04.010. CrossRef [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Tang M., Liu Y., Zhang H., Sun L., Lü P., Chen K. Comprehensive transcriptome sequencing of silkworm midguts: uncovering extensive isoform diversity and alternative splicing in BmNPV-sensitive and BmNPV-resistant strains. J. Invertebr. Pathol. 2024;204 doi: 10.1016/j.jip.2024.108104. CrossRef [DOI] [PubMed] [Google Scholar]
  43. Tian K., Zhang C., Gao C., Shi J., Xu C., Xie W., Yan S., Xiao C., Jia X., Tian Y., Sun G., Kang X., Wang K., Li W. Full-length transcriptome sequencing of seven tissues of GuShi chickens. Poult. Sci. 2025;104(2) doi: 10.1016/j.psj.2024.104697. CrossRef [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Trefts E., Gannon M., Wasserman D.H. The liver. Curr. Biol. 2017;27(21):R1147–R1151. doi: 10.1016/j.cub.2017.09.019. CrossRef [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Turner R.E., Pattison A..D., Beilharz T.H. Alternative polyadenylation in the regulation and dysregulation of gene expression. Semin. Cell Dev. Biol. 2018;75:61–69. doi: 10.1016/j.semcdb.2017.08.056. CrossRef [DOI] [PubMed] [Google Scholar]
  46. Wang P., Gong Y., Li D., Zhao X., Zhang Y., Zhang J., Geng X., Zhang X., Tian Y., Li W., Sun G., Han R., Kang X., Li Z., Jiang R. Effect of induced molting on ovarian function remodeling in laying hens. Poult. Sci. 2023;102(8) doi: 10.1016/j.psj.2023.102820. CrossRef [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. Wang R., Zheng D., Yehia G., Tian B. A compendium of conserved cleavage and polyadenylation events in mammalian genes. Genome Res. 2018;28(10):1427–1441. doi: 10.1101/gr.237826.118. CrossRef [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Wang T., Ji Z., Xiao X., Zhu D., Li H., Li X. Identification of reproduction-related genes in the hypothalamus of sheep (Ovis aries) using the nanopore full-length transcriptome sequencing technology. Sci. Rep.-Uk. 2024;14(1) doi: 10.1038/s41598-024-79140-5. CrossRef [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Wang X., Guo S., Xiong L., Wu X., Bao P., Kang Y., Cao M., Ding Z., Liang C., Pei J., Guo X. Complete characterization of the yak testicular development using accurate full-length transcriptome sequencing. Int. J. Biol. Macromol. 2024;271(Pt 1) doi: 10.1016/j.ijbiomac.2024.132400. CrossRef [DOI] [PubMed] [Google Scholar]
  50. Wu J., Liu S., Jiang D., Zhou Y.N., Jiang H., Xiao X., Zha B., Fang Y., Huang J., Hu X., Mao H., Liu S., Chen B. Exploring gene expression and alternative splicing in duck embryonic myoblasts via full-length transcriptome sequencing. Vet. Sci. 2024;11(12):601. doi: 10.3390/vetsci11120601. CrossRef [DOI] [PMC free article] [PubMed] [Google Scholar]
  51. Xie Y., Kang R., Klionsky D.J., Tang D. GPX4 in cell death, autophagy, and disease. Autophagy. 2023;19(10):2621–2638. doi: 10.1080/15548627.2023.2218764. CrossRef [DOI] [PMC free article] [PubMed] [Google Scholar]
  52. Xu X., Jian Y., Huang L., Luo W., Wu B., Feng S., Zhou C., Zhang L. Characterization of avian β-defensin genes in Galliformes reveals widespread evolutionary diversification and distinct evolutionary relationships with infection risk. Bmc Genom. 2025;26(1):211. doi: 10.1186/s12864-025-11390-7. CrossRef [DOI] [PMC free article] [PubMed] [Google Scholar]
  53. Yang J., Chen S., Tang Y., Wang X., Wang L., Liu H., Ma F., Zhao Q., Xing K., Yu Y., Wang C. Integrated analysis of GWAS and molQTLs reveals cell-specific genetic variants in the porcine immune system. Commun. Biol. 2026;9(1):408. doi: 10.1038/s42003-026-09605-y. CrossRef [DOI] [PMC free article] [PubMed] [Google Scholar]
  54. Yang Q., Fong L., Lyu W., Sunkara L.T., Xiao K., Zhang G. Synergistic induction of chicken antimicrobial host defense peptide gene expression by butyrate and sugars. Front. Microbiol. 2021;12 doi: 10.3389/fmicb.2021.781649. CrossRef [DOI] [PMC free article] [PubMed] [Google Scholar]
  55. Zhang H., Jia M., Su Y., Zhu W. Feeding frequency affects glucose and lipid metabolism through SIRT1/AMPK pathway in growing pigs with the same amount of daily feed. J. Nutr. Biochem. 2022;100 doi: 10.1016/j.jnutbio.2021.108919. CrossRef [DOI] [PubMed] [Google Scholar]
  56. Zhang J., Luo Z., Zhang M., Gong Y., Tian Y., Han R., Li D., Guo Y., Kang X., Yang Y., Li Z., Jiang R. The promotion of liver vitamin metabolism is of great significance for laying hens during fasting. Bmc Genom. 2025;26(1):603. doi: 10.1186/s12864-025-11730-7. CrossRef [DOI] [PMC free article] [PubMed] [Google Scholar]
  57. Zhang L., An Z., Li W., Liu X., Zhou E., Cui P., Zhang Y., Guo Y., Li G., Li Z., Tian Y., Huang S., Jiang R., Kang X., Han R. Effect of induced molting on bone remodeling in laying hens. Bmc Genom. 2025;26(1):1113–1128. doi: 10.1186/s12864-025-12335-w. CrossRef [DOI] [PMC free article] [PubMed] [Google Scholar]
  58. Zheng X., Chen Z., Zhuang W., Zhang J., He J., Xie Y., Chen J. Chicken LEAP2 level substantially changes with feed intake and may be regulated by CDX4 in small intestine. Animals (Basel) 2022;12(24):3496. doi: 10.3390/ani12243496. CrossRef [DOI] [PMC free article] [PubMed] [Google Scholar]
  59. Zhong C., Liu Z., Li D., Kang L., Jiang Y. Long-read sequencing reveals the effect of follicle-stimulating hormone on the mRNA profile of chicken granulosa cells from prehierarchical follicles. Poult. Sci. 2023;102(5) doi: 10.1016/j.psj.2023.102600. CrossRef [DOI] [PMC free article] [PubMed] [Google Scholar]
  60. Zhu T., Liang W., He Y., Zhang B., Liu C., Wang D., Deng L., Li D., Li W., Yan F., Tian Y., Han R., Kang X., Li Z., Jiang R., Sun G. Transcriptomic analysis of mechanism underlying the effect of induced molting on semen quality and reproductive performance in aged Houdan roosters. Poult. Sci. 2023;102(10) doi: 10.1016/j.psj.2023.102935. CrossRef [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

mmc1.docx (3.6MB, docx)
mmc2.xlsx (2.1MB, xlsx)

Data Availability Statement

The RNA-seq data from this study have been submitted in the Genome Sequence Archive in the National Genomics Data Center under accession number CRA040380.

Articles from Poultry Science are provided here courtesy of Elsevier

RESOURCES