Abstract Background Megalobrama amblycephala presents unsynchronized growth, which affects its productivity and profitability. The liver is essential for substance exchange and energy metabolism, significantly influencing the growth of fish. Results To investigate the differential metabolites and genes governing growth, and understand the mechanism underlying their unsynchronized growth, we conducted comprehensive transcriptomic and metabolomic analyses of liver from fast-growing (FG) and slow-growing (SG) M. amblycephala individuals. A total of 2,097 differentially expressed genes (DEGs) were identified between FG and SG, with 830 genes exhibiting significantly higher expression level in FG. KEGG and GO enrichment analysis indicated that the DEGs with higher expression level were significantly correlated with insulin signaling pathway, steroid hormone and lipid metabolism related pathway (PPAR signaling pathway and fatty acid degradation). In the metabolomic analysis, 224 differentially expressed metabolites (DEMs) were detected, of which 128 were significantly more abundant in FG. These more abundant DEMs were prominently enriched in pathways associated with cell proliferation and energy metabolism (Oxidative phosphorylation, mTOR signaling pathway and FoxO signaling pathway). In addition, DEGs and DEMs in adenosine diphosphate (ATP) hydrolysis activity and associate with fatty acid metabolism, glucose metabolism, and amino acid metabolism pathways were both found in the transcriptomic and metabolomic integrated data. These findings suggest that the large amounts of energy generated by fatty acid, glucose metabolism and other energy metabolism pathway promote the rapid growth of FG. Conclusions This research is the first to integrate metabolomic and transcriptomic analyses of liver to identify key genes, metabolites, and pathways to uncover the molecular and metabolic mechanisms of unsynchronized growth in M. amblycephala. The identified metabolic and genes can be potential targets for selective breeding programs to improve growth performance in aquaculture. Supplementary Information The online version contains supplementary material available at 10.1186/s12864-025-11208-6. Keywords: Transcriptome, Metabolome, Liver, Growth difference, Megalobrama amblycephala Background Growth is critical for aquaculture production efficiency and sustainability. Variation in growth rates among individuals raised up in the similar environmental condition present a major challenge in aquaculture [[36]1–[37]3]. Fish growth involves various biological processes, including hormonal regulation, energy metabolism, tissue growth and development, nutrient absorption, and cellular proliferation [[38]4–[39]6]. The liver is an important digestive and metabolic organ in fish growth and development [[40]7]. It has various functions in fish, such as regulating storage of glycogen, protein synthesis, lipid metabolism, and bile secretion, providing substances needed by other tissues [[41]8, [42]9]. Additionally, the liver produces various hormones to regulate cell proliferation and differentiation, and stimulating protein synthesis [[43]10]. Therefore, the liver’s metabolic activity is intimately associated with the growth of organisms, and alterations in the expression profiles of crucial genes within the liver can impact growth rates [[44]11]. Furthermore, as the principal organ of energy balance in the body, the liver provides energy through fatty acid oxidation and glycogen metabolism [[45]12]. Therefore, understanding the differences in genes and metabolites in the liver of fast-growing (FG) and slow-growing (SG) individuals can help explore their differences in energy metabolism and the unsynchronized growth mechanisms in fish. The transcriptome allows for the swift acquisition of the sequences and expression levels of nearly all transcripts within a tissue under particular conditions [[46]13]. Liver transcriptome of unsynchronized growth grass carp (Ctenopharyngodon idella) showed growth-related differentially expressed genes (DEGs) such as ig1, ghr, and igf1r were identified and DEGs were significantly enriched in growth-related pathway (mTOR signaling pathways) [[47]14]. Metabolomics aims to study changes in metabolites within biological systems, which can directly reflect the overall state and function of an organism, as well as the dynamic processes of its changes [[48]15]. Thus, it was widely applied in the study of fish growth. Metabolomic analysis of 3-month-old Chinese sturgeon (Acipenser dabryanus) with different body types found that the growth advantage of Chinese sturgeon is intimately connected to antioxidant functions, protein synthesis and glycolysis [[49]16]. Then the integration of metabolomics and transcriptomics is an efficient approach for clarifying the mechanism of growth variation. In FG Dorper sheep, PHGDH, LPL and PPARGC1A genes are highly expressed. These genes leading to the upregulation of pathways involved in energy metabolism, while purine and pyrimidine are also more copious in FG [[50]17]. In addition, Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis of co-enriched differentially expressed metabolite (DEMs) and DEGs in the rapidly growing loach (Paramisgurnus dabryanus) revealed significantly enriched in glycerophospholipid metabolism, glycolysis/gluconeogenesis, and arginine and proline metabolism [[51]18]. The blunt snout bream (Megalobrama amblycephala) is highly valued for its delicious taste, rapid growth, substantial nutritional content, and significant economic value, making it as one of the most important freshwater fish species in China [[52]19]. Whole transcriptome analysis of M. amblycephala with different growth rates discovered novel_circ_0001608 and novel_circ_0002886 could potentially serve significant regulatory roles in FG muscle growth [[53]1]. Additionally, Zou et al. (2024) found that the DEGs with higher expression level in FG were mainly involved in pathways associated with cell proliferation, leading to muscle hyperplasia at 4 months, while the DEGs with higher expression level in FG were primarily involved in pathways associated with both cell proliferation and protein synthesis, resulting in muscle hypertrophy and hyperplasia at 10 months in M. amblycephala [[54]3]. However, the liver, as one of the vital organs for growth, has been rarely studied in relation to growth differences in M. amblycephala. Therefore, we performed transcriptomic and metabolomic analyses of FG and SG livers and identify candidate genes, metabolite and regulatory pathways underlying growth variations in M. amblycephala. Our findings will offer a foundation to uncover growth difference and enhance the improvements productivity of M. amblycephala. Methods Animals and samples preparation All experimental animals in this study were offspring of M. amblycephala “Huahai No. 1”, which were bred simultaneously at the Ezhou breeding facility of Huazhong Agricultural University and reared under identical conditions in the same pool. At four months, total length (TL), body length (BL) and body weight (BW) of the fish were measured. Subsequently, 12 fish were sampled based on the BW, including 6 fish with the maximum body weight (designated as FG), and 6 fish with the minimum body weight (designated as SG). The fish were anesthesia with tricaine methanesulfonate (MS-222) (100 mg/L) (Sigma, Saint Louis, MO, USA). We collected liver samples from each fish, which were subsequently preserved at −80 °C for future utilization. For RNA-seq and real-time quantitative reverse transcription PCR (RT-qPCR), three individuals from each of the FG and SG group were used, while six individuals from each group were used for metabolome analysis. RNA extraction, library construction, and sequencing To extract total RNA from liver, RNAiso Plus (TaKaRa, Dalian, China) was used. After removing genomic DNA and assessing the quality and integrity of RNA, we utilized high-quality mRNA from FG and SG to construct library. Magnetic beads were used to enrich mRNA in samples and fragment it into short fragments. The first strand of cDNA was synthesized using a random hexamer primer and M-MuLV Reverse Transcriptase (RNase H-), while the second strand of cDNA was synthesized using DNA Polymerase I and RNase H. The 3’ end was repaired, the poly(A) tail was ligated to the sequencing linker, the fragment size was selected, the double-stranded cDNA was purified and PCR amplified. AMPUREXP system (Beckman Coulter, Beverly, MA, USA) was used to purify PCR products for sequencing. Six cDNA libraries from FG and SG were sequenced using the Illumina sequencing platform (Illumina NovaSeq 6000) (Illumina, San Diego, CA, USA) by BioNovoGene Co. (Suzhou, China). Transcriptome analysis Fastp software (version 0.19.7) is used for initial processing of fastq format raw data (raw reads). Hisat2 v2.0.5 was used to construct an index of the reference genome, and the same version of software was used to align paired-end clear reads to M. amblycephala reference genome (ASM1881202V). Fragments per kilobase of transcript per million mapped fragments (FPKM) method was used to quantitative analysis gene expression levels [[55]20]. The DESEQ 2R package with readcount (version 1.20.0) was used to analyze the differential expression between FG and SG. Benjamini and Hochberg’s method was used to obtain p-adjust. Significance thresholds of p-adjust < 0.05 and |log2foldchange| > 1 were applied to identify significant differences in transcriptional expression. DEGs were analyzed using GO enrichment and KEGG pathway enrichment with a significance threshold set at p < 0.05, and heat maps were plotted by bioinformatics ([56]https://www.bioinformatics.com.cn) (accessed on 12 April 2024). Metabolites extracted and data analysis FG and SG each had 6 biological replicates and were subjected to metabolomic analyzed by BioNovoGene Co. (Suzhou, China). For each liver placement, the following methods were performed for metabolite extraction. Weight 100 ± 1 mg sample into a 2 mL centrifuge tube, adding 1000 µL of tissue extraction solution [75% (9:1 methanol: chloroform): 25% H[2]O] (−20 °C storage), and add 3 steel balls. Grinded tissue in grinder for 60 s at 50 Hz, and repeat the above steps 2 times. Ultrasonicate at room temperature for 30 min, and place on ice for 30 min. Centrifuge at 12,000 rpm at 4 °C for 10 min, transfer all the supernatant to a centrifuge tube, and concentrate and dry. After add 200 µL 50% acetonitrile solution is prepared with 2-chloro-l-phenylalanine (4ppm) (stored at 4℃) to redissolve the sample, filter the supernatant by 0.22 μm membrane and transfer into the detection bottle for LC-MS detection [[57]21]. MSConvert in the ProteoWizard package (v3.0.8789) was used to convert raw data to MZXML format [[58]22] and feature extraction, retention time correction, and alignment were performed using XCMS [[59]23]. Data normalization is based on robust LOESS signal correction (QC-RLSC). Multivariate data analysis and modeling were done by Ropls software [[60]24]. After scaling data processing, the model was established by principal component analysis (PCA), orthogonal partial least square discriminant analysis (PLS-DA) and partial least square discriminant analysis (OPLS-DA). OPLS-DA allowed the determination of discriminating metabolites using the variable importance on projection (VIP). Finally, VIP values > 1 and p < 0.05 were considered to be statistically significant metabolites. Metaboanalyst is used for pathway analysis of DEMs [[61]25]. The detected metabolites in metabolomics were subsequently mapped to the KEGG pathway to interpret higher-level systemic functions biologically. Correlation analysis of transcriptomic and metabolomic The joint analysis of DEGs and DEMs was conducted using the cor command in R (4.0.3) to construct a transcriptome-metabolome network based on Pearson correlation coefficient > 0.8 [[62]26]. Pathway analysis was performed on DEGs and DEMs, and their common pathway information was mapped to KEGG [[63]27]. A bar graph was plotted based on the significant pathways (p < 0.05) for the enrichment of DEMs and DEMs into the transcriptome and metabolome. RT-qPCR analysis To validate our RNA sequencing results, RT-qPCR analysis was performed on selected 9 DEGs associated with energy metabolism. The QuantStudio 6 Flex Real-Time PCR System (Life Technologies, Carlsbad, CA, USA) with HieffTM qPCR SYBR^® Green Master Mix (Yeasen, Shanghai, China) were used to perform RT-qPCR. Total RNA was extracted from FG and SG samples, and using the HiScript™ Q RT SuperMix (Vazyme, Nanjing, China) to synthesize cDNA. PCR primers used for RT-qPCR amplification are displayed in Table [64]1. Following the amplifications, melting curve analyses were performed. The relative gene expression was calculated using the 2^−ΔΔCt method. The amplification efficiency was assessed through the gradient dilution method. The initial CT value was employed to calculate the primer amplification efficiency and R^2 using the formula (E = (10^−1/k − 1) × 100% ), with the slope obtained from the linear regression equation in Excel (version 2016) (Table [65]S1). Table 1. Primers used for gene expression analysis via RT-qPCR Gene Accession number Location Primer dpyda.1 [66]XM_048175396.1 Span the 22 and 23 exons, primarily in the 23 exons F: TGCTGGATACCAGGCCATTG R: CTTTGGCACGTAAGGCGTTG gfpt2 [67]XM_048166049.1 Span the 1 and 2 exons, primarily in the 1 exon F: GTCGTGGCGGCTGTAAGTAA R: AGGCGAAAATTCCGCACATC irs2a [68]XM_048197081.1 Span the 1 and 2 exons, primarily in the second exon F: GCAACCACTGTACAAGATTGAGC R: AGGTGCAAAGGTCACAGTTCG cremb [69]XM_048174087.1 Span the 5 and 6 exons, primarily in the 5 exons F: CTGCATCTGGCCTTTCCCAG R: CGGGCAGCTTCCCTGTTTTT gsk3ba [70]XM_048193076.1 Span the 8 and 9 exons, primarily in the 8 exons F: ATCAAGGTGCTTGGCACTCC R: TGGCCGGAACACCTGTTTAC irs2b [71]XM_048193118.1 Span the 1 and 2 exons, primarily in the 2 exons F: ACGGCAACAACAGTGAAAGACTAA R: CGTGGTCCATCTGTCCACAC tecra [72]XM_048171274.1 Span the 3 and 4 exons, primarily in the 4 exons F: GCCTGGACCCAAAAGGGAAA R: GCACTCGACCAGGAACACAG cdkn1cb [73]XM_048158265.1 Span the 1 and 2 exons, primarily in the 2 exons F: GGACATGCGCATTACAGACTTTT R: CTTTTCCCGCCTCCGTGAC arg2 [74]XM_048205354.1 Span the 6 and 7 exons, primarily in the 6 exons F: TGTGGATCCAGGCGAGCATA R: CGGCCTTTGTTTCCTTGCCA β-actin [75]XM_048197769.1 Span the 2 and 3 exons, primarily in the 3 exons F: CGTGCTGTTTTCCCTTCCATT R: CAATACCGTGCTCAAAGGATACTT [76]Open in a new tab Statistical analysis Statistical analysis and mapping were performed using GraphPad Prism 8.0, and all data were expressed as mean ± standard deviation. The data was subjected to multiple comparisons within groups using one-way analysis of variance (ANOVA), and the evaluation of differences between the means of individual groups using the t-test. A significance level of p < 0.05 was used as the threshold for determining statistical significance (*p < 0.05, **p < 0.01, ***p < 0.001). Result Growth difference of FG and SG The growth performance data of M. amblycephala were obtained from our previous experiment [[77]3]. We collected FG and SG samples at 4 months for growth performance analysis. The average TL, BL and BW of FG and SG had significant difference (p < 0.01). The TL of FG and SG were 18.04 ± 0.64 cm and 8.62 ± 0.42 cm, respectively. Additionally, the BL were 14.86 ± 0.52 cm and 7.03 ± 0.42 cm and the BW were 67.26 ± 8.01 g and 6.87 ± 0.35 g, respectively (Fig. [78]S1). Remarkably, FG individuals weighed about ten times more than SG. Transcriptome analysis of liver from FG and SG Table [79]2 presents the raw reads, raw bases, clean reads, clean bases, and Q30 values for 6 libraries. The Q30 values for each sample exceeded 93.76%. High-quality sequencing data (Q30 > 93%) from all libraries will be utilized for further analyses. A total of 2,097 DEGs were identified between the FG and SG of 4-month-old blunt snout bream liver. Of these DEGs in FG, 830 exhibited increased expression, while 1,267 showed decreased expression. (Fig. [80]1A). For GO analysis of DEGs with higher expression level in FG, lipid metabolism (“lipid oxidation” (20 DEGs) and “fatty acid oxidation” (20 DEGs)) were significantly enriched (Fig. [81]1B, Table [82]S2). For KEGG analysis of DEGs with higher expression level in FG, pathways related to metabolic were mainly, including pathways associated with lipid metabolism (“PPAR signaling pathway” (8 DEGs) and “fatty acid degradation” (8 DEGs)), as well as pathways associated with glucose metabolism (“insulin signaling pathway” (12 DEGs)). Furthermore, “steroid hormone biosynthesis” (8 DEGs), “metabolism of xenobiotics by cytochrome P450” (9 DEGs) and “pantothenate and CoA biosynthesis” (8 DEGs) were also significantly enriched (Fig. [83]1C, Table [84]S3). The DEGs enriched in pathways associated with growth differences were compared and incorporated into the heatmap. DEGs with higher expression level were mostly enriched in lipid metabolism and glucose metabolism (Fig. [85]1D). Table 2. Transcriptome sequencing data of liver from FG and SG Sample Raw reads Raw base Clean reads Clean bases Q30 FG_L1 43,975,364 6.6G 41,422,224 6.22G 93.76% FG_L2 45,723,792 6.86G 43,043,472 6.46G 94.83% FG_L3 45,880,192 6.88G 43,273,988 6.49G 94.09% SG_L1 43,692,146 6.55G 41,187,244 6.18G 94.11% SG_L2 44,519,954 6.68G 41,934,828 6.29G 94.29% SG_L3 50,164,746 7.52G 47,213,492 7.08G 94.93% [86]Open in a new tab Fig. 1. [87]Fig. 1 [88]Open in a new tab Transcriptome analysis of liver from FG and SG. A Volcano plot for DEGs. B GO enrichment analysis of DEGs with higher expression level in FG. C KEGG enrichment analysis of DEGs with higher expression level in FG. D The heatmap of 8 DEGs related to PPAR signaling pathway, 9 DEGs related to fatty acid degradation and 11 DEGs related to insulin signaling pathway Metabolome analysis of liver from FG and SG To gain better understanding of the metabolic distinctions between the liver of FG and SG, this study further investigated the liver of M. amblycephala using a metabolomics approach. PCA of the samples revealed a complete separation in the composition of liver metabolites between the FG and SG, with distinct grouping trends observed between the different groups (Fig. [89]2A). The method used in this study to screen for DEMs involved combining t-test p-values, and VIP values and the criteria for screening were p-values < 0.05 and VIP > 1.0. A clear distinction between the SG and FG comparisons was shown by clustering analysis according to the DEMs (Fig. [90]2B). The results indicated that 128 DEMs exhibited higher abundance in FG, whereas 96 DEMs exhibited lower abundance (Fig. [91]2C). This indicates a significant difference in liver metabolites, with carboxylic acids and their derivatives showing the largest difference (17 DEMs), followed by fatty acyls (14 DEMs). To gain better understanding of the different metabolites functions and explore the potential metabolic pathways influencing the differential growth of FG and SG, KEGG enrichment analysis was conducted on significantly DEMs with more abundance in FG. The result presented that the higher DEMs were mainly enriched in the “Neuroactive ligand-receptor interaction” pathway (7 DEMs) (Fig. [92]2D, Table [93]S4), and pathways related to energy metabolism including the “mTOR signaling pathway” (2 DEMs), “Oxidative phosphorylation” (3 DEMs) and “FoxO signaling pathway” (2 DEMs). Fig. 2. [94]Fig. 2 [95]Open in a new tab Metabolome analysis of liver from FG and SG. A PCA score plot of the metabolome. B Heatmap of the differential metabolites. C Volcano plot for differential metabolites expression; D KEGG enrichment analysis of DEMs with higher abundance in FG Integrated analysis of metabolome and transcriptome of liver from FG and SG To further elucidate the differences in the liver of FG and SG, we conducted a combined analysis of transcriptomics and metabolomics of the liver. We conducted correlation analysis on DEMs and DEGs associated with energy metabolism based on the KEGG pathway enrichment analysis of transcriptome and metabolome data. Using Pearson correlation analysis between 14 DEMs and 74 DEGs (Fig. [96]3A; Table [97]S5). The results showed the association of each transcript with a specific metabolite. ATP, adenosine diphosphate (ADP), adenosine monophosphate (AMP), nicotinamide adenine dinucleotide (NAD) and reduced nicotinamide adenine dinucleotide (NADH) along with gene abcg8 (ATP binding cassette subfamily G member 8) and pdk2a (pyruvate dehydrogenase kinase 2a) were more abundant in FG. Energy metabolism has an important effect on growth difference. Furthermore, to understand the important pathways in energy metabolism and elucidate the relationships between DEGs and DEMs within them. We also focused on the DEGs and DEMs involved in fatty acid metabolism, glucose metabolism, and amino acid metabolism pathways to elucidate their significant roles in FG (Fig. [98]3B). The DEGs acyl-CoA synthetase bubblegum family member 2 (acsbg2), carnitine palmitoyltransferase 1 (cpt1), and acetyl-CoA carboxylase beta (acacb) in fatty acid metabolism, ATP citrate lyase (acly) and isocitrate dehydrogenase (NADP (+)) 1 (idh1) in the tricarboxylic acid cycle, and proline dehydrogenase 2 (prodh2) and aldehyde dehydrogenase (aldh) in glucose metabolism are all significantly enriched in FG. In amino acid metabolism, the DEMs N-Acetyl-L-citrulline, L-Arginosuccinic acid, L-Arginine, and L-Proline are significantly enriched in FG. Fig. 3. [99]Fig. 3 [100]Open in a new tab Comprehensive analysis of metabolome and transcriptome of liver from FG and SG. A Correlation analysis of DEGs and DEMs of liver from FG and SG. Positive correlations were marked in red and negative correlations in blue between the transcriptomic and metabolomic analyses. * p < 0.05. B Important regulatory networks of DEGs and DEMs. High expression in FG is shown in red font, whereas high expression in SG is shown in green font, the italics indicate genes, and the arrow points metabolites RT-qPCR validation Through gene differential expression analysis and screening of growth and development-related DEGs, 9 energy metabolism-related DEGs with differential expression were selected from the liver transcriptome, including dihydropyrimidine dehydrogenase a, tandem duplicate 1 (dpyda.1), glutamine-fructose-6-phosphate transaminase 2 (gfpt2), insulin receptor substrate 2a (irs2a), cyclin-dependent kinase inhibitor 1Cb (cdkn1cb), insulin receptor substrate 2b (irs2b), cAMP responsive element modulator b (cremb), trans-2,3-enoyl-CoA reductase a (tecra), glycogen synthase kinase 3 beta, genome duplicate a (gsk3ba) and arginase 2 (arg2) were validated by RT-qPCR, and the expression patterns of these 8 DEGs demonstrated consistency with the sequencing results with p < 0.05, while the p value of tecra expression level in qPCR is 0.0537 with a consistent pattern of sequencing data (Fig. [101]4). Fig. 4. [102]Fig. 4 [103]Open in a new tab RT-qPCR verification of 9 DEGs associated with energy metabolism. RNA-seq expression levels were displayed on the right vertical axis, whereas RT-qPCR relative expression levels were displayed on the left vertical axis different lowercase letters denote significant differences (p < 0.05) Discussion The liver, functioning as a critical organ for substance exchange and energy metabolism [[104]28], is significantly correlated with the growth of fish [[105]29]. Therefore, exploring the DEGs and DEMs in the livers of FG and SG M. amblycephala can help us identify the mechanisms that significantly influence their growth rate. In our transcriptome results, DEGs such as igf2b, pik3rl, which are related to insulin, exhibit higher expression in the FG, and the “insulin signaling pathway” was significantly enriched in FG. Insulin-like growth factors (IGF), including IGF1 and IGF2, play a crucial role in regulating growth in fish [[106]30]. IGF2 was considered very important in postnatal growth. Knockout of igf2a or igf2b genes in zebrafish resulted in growth retardation, and igf2b was found to have a greater impact on zebrafish growth [[107]31]. Furthermore, evidence from zebrafish study indicated that GH directs muscle tissue growth via the production of endocrine IGF2 (from the liver) and autocrine/paracrine actions of muscle igf2b [[108]32]. In juvenile grass carp, higher GH levels result in elevated expression of igf2b in the liver, indicating its important role in regulating grass carp growth [[109]33]. In our study, igf2b and its downstream gene pik3r1 were highly expressed in FG. PIK3R1, as a regulatory subunit of PI3K, has a considerable impact on regulating insulin signaling and metabolic homeostasis [[110]34, [111]35]. In addition, steroid hormone biosynthesis and metabolism of xenobiotics by cytochrome P450 are significantly enriched in KEGG. Steroid hormones have important impacts on the growth of fish, including sex steroids (estrogen and androgen), and cytochrome P450 is one of the important enzymes for synthesizing steroid hormones [[112]36]. These results suggest that hormone regulation may be one of the important factors influencing growth differences in M. amblycephala. Furthermore, many DEGs (acsbg2, carnitine palmitoyltransferase 1B (cpt1b), and acacb) and pathways correlate with lipid metabolism were enriched in KEGG, such as the “PPAR signaling pathway” and “fatty acid degradation”. In our study, acsbg2, cpt1b, and carnitine palmitoyltransferase 1Ab (cpt1ab) with higher expression level were enriched in the fatty acid metabolism pathway. ACSBG2, a key enzyme in fatty acid activation, showed significantly higher expression levels in rapidly growing chicken liver compared to SG ones, suggesting its potential role in promoting growth by increasing intracellular fatty acid synthesis and lipid accumulation [[113]37, [114]38]. acacb is a crucial rate-limiting enzyme in fatty acid metabolism. Previous studies indicating its potential regulatory role in fat deposition and fatty acid metabolism in obese pigs [[115]39]. Carnitine palmitoyltransferase Iα (CPT1α) is a key factor in regulating fatty acid β-oxidation, and high carbohydrate intake resulted in down-regulation of cpt1a gene expression and then induced significant body weight loss in Micropterus Salmoides [[116]40]. The high expression of cpt1ab in this study suggests its potential role in promoting growth of FG through the regulation of lipid absorption and utilization. On the contrary, in SG, genes that positively regulate fatty acid synthesis—acyl-CoA synthetase long chain family member 4b (acsl4b), glycerol-3-phosphate acyltransferase 4 (gpat4), and lpin1a (lpin1a)—are significantly high expressed. acsl4b is involved in the activation of long-chain fatty acids, converting free fatty acids into acyl-CoA, which is the initial step in fatty acid synthesis [[117]41]. And gpat4 limits the oxidation of exogenous fatty acids in brown adipocytes [[118]42]. Lipin-1 is a phosphatidate phosphatase enzyme that converts phosphatidic acid into diacylglycerol, a key step in triglyceride synthesis [[119]43]. The regulation of these genes may inhibit the involvement of fatty acids in metabolism in SG individuals, reducing the energy supply for growth and development, thereby slowing down the growth rate. Our metabolomic results revealed enrichment of many energies metabolism-related pathways in FG, including “oxidative phosphorylation”, “mTOR signaling pathway” and “FoxO signaling pathway”. Oxidative phosphorylation is a crucial pathway for generating ATP [[120]44], often referred to as the “energy currency” of the cell, playing a crucial role in many important physiological processes [[121]45]. Regarding compounds involved in purine metabolism, this study uncovered four metabolites, including inosine monophosphate (IMP), ADP, ATP, and xanthine. These metabolites are involved in the degradation of adenosine ribonucleotides, closely linked to cellular energy metabolism. Due to the relatively low energy efficiency of glycolysis, which cannot meet the energy demands of cellular metabolism, ATP needs to be metabolized into ADP and further broken down into IMP, xanthine, and other products to provide more energy to meet the growth requirements of fish [[122]46, [123]47]. The “mTOR signaling pathway” and “FoxO signaling pathway” are considered to be important pathways in regulating energy metabolism [[124]48, [125]49]. And AMP, which was more abundant in FG, was identified in these two pathways. The result indicates that the two pathways have important effects on energy metabolism of the liver during the growth of M. amblycephala. Furthermore, metabolomic analysis revealed a significant enrichment of DEMs related to lipid metabolism, including linoleic acid, eicosapentaenoic acid (EPA), and L-palmitoylcarnitine. Studies have shown that linoleic acid constitutes an important fatty acid for coho salmon (Oncorhynchus kisutch), playing a significant role in growth and lipid utilization [[126]50]. In fish, EPA is closely associated with hepatic fatty acid oxidation and is vital for regulating growth and lipid metabolism [[127]51]. Additionally, the ester derivative of carnitine, L-palmitoylcarnitine, contributes to fatty acid metabolism and its abundance during liver lipid storage [[128]52]. Through the comprehensive analysis of transcriptomic and metabolomic from liver, we identified significant differences in lipid and amino acid metabolism between FG and SG. In the KEGG pathway analysis, the “pantothenate and CoA biosynthesis” pathways showed significant enrichment in the transcriptome and metabolome. Coenzyme A (CoA) is an indispensable cofactor participated in various metabolic reactions, including fatty acid synthesis and degradation, and the operation of the tricarboxylic acid cycle [[129]53]. Fatty acids serve as the primary energy source for animal growth, with animals primarily using dietary glucose and fatty acids to meet their energy need [[130]54]. As mentioned earlier, DEGs and DEMs associated with lipid metabolism were identified in transcriptome and metabolome data, respectively. These genes collectively promote fatty acid metabolism in M. amblycephala FG individuals, leading to an increase in acetyl-CoA entering the tricarboxylic acid cycle, providing more energy for growth and development. This suggests the crucial role of liver-regulated lipid metabolism in influencing growth differences in M. amblycephala. Furthermore, amino acids have a direct promoting effect on the growth of fish, increasing the rate of protein synthesis while reducing protein degradation [[131]55, [132]56]. Previous studies have shown that amino acids can serve as controllers of protein turnover in fish muscles [[133]57]. In this study, we also found that lysine, arginine, proline, tryptophan, and phenylalanine, essential amino acids for fish growth, were highly expressed in FG [[134]58]. Arginine deficiency led to growth inhibition, reduced feed conversion rate, and protein deposition rate in turbot [[135]59]. Additionally, increased activity of arginase could serve as a significant constraint on the maximum growth of Micropterus salmoides [[136]60]. Proline supplementation in the diet of abalone (Haliotis midae) acts as a substrate for amino acid metabolism, promoting energy production in the tricarboxylic acid cycle [[137]61]. The enriched amino acid metabolites in the FG discovered in this study contribute to improving the growth of fish, thereby enhancing growth rate. The growth differences of the M. amblycephala are influenced by various factors, such as genetic factors, growth environment, and nutritional feed. Further research is necessary to thoroughly explore the causes of the growth variations in M. amblycephala. Conclusions In this study, we utilized transcriptomic and metabolomic analyses to investigate the transcriptional and metabolic differences in liver of M. amblycephala individuals with growth differences. The exploration using transcriptomic-metabolomic analysis revealed the metabolites and associated genes that influence the rapid growth of M. amblycephala. Within the transcriptome, it was observed that energy metabolism, affecting the growth and development of FG individuals, is mainly influenced in liver through the “PPAR signaling pathway”, “fatty acid degradation”, and “insulin signaling pathway”. In the metabolome, DEMs with higher abundance such as a large number of fatty acyls in FG were enriched, closely associated with fatty acid metabolism. Additionally, it was found that purine metabolism pathways related to energy supply were significantly enriched, along with compounds associated with purine metabolism. Furthermore, the correlation analysis between transcriptome and metabolome revealed significant differences in lipid and amino acid metabolism. In conclusion, these results unveil the differences in liver metabolism between M. amblycephala FG and SG, providing crucial insights into the regulatory mechanisms of growth in M. amblycephala for further understanding. Supplementary Information [138]12864_2025_11208_MOESM1_ESM.pdf^ (90.6KB, pdf) Additional file1: Fig. S1. Growth difference between FG and SG. (A) Total length. (B) Body length. (C) Body weight. *p < 0.05, **p <0.01, ***p < 0.001. [139]12864_2025_11208_MOESM2_ESM.xlsx^ (10.4KB, xlsx) Additional file2: Table S1. Analysis of qPCR amplification characteristics. [140]12864_2025_11208_MOESM3_ESM.xlsx^ (33.2KB, xlsx) Additional file3: Table S2. The GO enrichment results of DEGs with higher expression level in FG. [141]12864_2025_11208_MOESM4_ESM.xlsx^ (24.7KB, xlsx) Additional file4: Table S3. The KEGG enrichment results of DEGs with higher expression level in FG. [142]12864_2025_11208_MOESM5_ESM.xlsx^ (18.5KB, xlsx) Additional file5: Table S4. The KEGG enrichment results of DEMs with higher abundance in FG. [143]12864_2025_11208_MOESM6_ESM.xlsx^ (30KB, xlsx) Additional file6: Table S5. DEGs and DEMs related to energy metabolism between FG and SG. Acknowledgements