Transcriptome analysis identifies key miRNAs involved in the regulation of rabbit intramuscular fat deposition

Abstract

Background

Intramuscular fat (IMF) is a critical determinant of rabbit meat quality, influencing tenderness, juiciness, and flavor. miRNAs are crucial for IMF synthesis by regulating genes involved in lipid deposition, but the mechanisms by which miRNAs regulate IMF deposition in rabbits remain poorly understood. This study aimed to investigate the role of microRNAs (miRNAs) in IMF deposition in Hycole rabbits (HR) and Rex rabbits (RR) at three developmental stages (35, 75, and 150 days of age) using small RNA sequencing.

Results

A total of 195 differentially expressed (DE) miRNAs were identified in HR across the three stages, including 49 co-expressed DE miRNAs (e.g., ocu-miR-29c-5p, ocu-miR-382-3p, ocu-miR-29b-3p). In RR, 222 DE miRNAs were detected, with 45 co-expressed DE miRNAs (e.g., ocu-miR-182-5p, ocu-miR-486-5p, ocu-miR-370-5p). Comparative analysis between breeds identified 128 DE miRNAs, but none were co-expressed. GO and KEGG enrichment analyses showed DE miRNAs and their target genes were significantly enriched in pathways like PI3K-Akt, MAPK, Hippo, and HIF-1, which were closely associated with lipogenesis, metabolism, and obesity.

Conclusions

This study elucidates the transcriptional regulatory mechanisms of IMF deposition in rabbits and identifies key miRNAs influencing IMF accumulation. Our findings provide new insights into the molecular mechanisms underlying rabbit IMF deposition and offer important resources to support future genetic improvement and management strategies in rabbit production.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12864-026-12816-6.

Background

Rabbit meat is a high-protein, low-fat food, making it an excellent source of quality protein for individuals seeking weight control or those with metabolic syndrome [1]. However, the relatively poor flavor of rabbit meat limits consumer acceptance, and consequently constrains production and market demand. IMF is a critical determinant of rabbit meat quality and economic value because its distribution and accumulation are directly related to meat flavor, texture, and nutritional value [2]. IMF primarily consists of triglycerides and phospholipids, among which phospholipids play a key role in determining meat flavor, juiciness, and tenderness [3]. Lecithin, a major phospholipid component, also serves as a precursor of flavor compounds in muscle. IMF influences juiciness and tenderness and is closely correlated with the overall flavor, thereby shaping the sensory attributes of rabbit meat [4].

MicroRNAs (miRNAs) are endogenous non-coding RNAs, typically 18–24 nucleotides in length, that play crucial roles in post-transcriptional gene regulation [5]. Growing evidence indicates that miRNAs play central roles in regulating IMF deposition. For instance, in poultry, miR-15a targets downstream genes of the peroxisome proliferator-activated receptor (PPAR), participating in the β-oxidation of fatty acids and thus influencing fat deposition [6]. Additionally, the expression of miR-451 is negatively correlated with IMF, as it regulates lipid accumulation and fatty acid composition by targeting acetyl-CoA carboxylase alpha (ACACA) [7]. Individual miRNAs can directly or indirectly regulate multiple targets. For example, miR-34a inhibits lipogenesis by targeting PDGFRα [8], and also regulates IMF deposition in porcine adipocytes by targeting long-chain acyl-CoA synthetase 4 (ACSL4) [9]. Likewise, miR-181a promotes lipogenesis in swine by targeting TNF-α [10], whereas miR-146a-5p inhibits TNF-α-induced lipogenesis by targeting the insulin receptor [11]. Furthermore, a comprehensive miRNA expression profile in Jiaxing black pigs identified 741 miRNAs, including 155 DE miRNAs [12]. These findings underscore the pivotal role of miRNAs in fat deposition. However, the studies on miRNA expression profiles and functions during IMF deposition in domestic rabbits remain relatively scarce.

Hycole rabbits are widely used for commercial meat production and are frequently studied for IMF content, with their primary economic value derived from meat yield [13]. In contrast, Rex rabbits are internationally recognized for both meat and fur production [14, 15]. IMF plays a critical role in determining meat quality by influencing flavor, tenderness, and juiciness. Our previous transcriptomic study demonstrated significant differences in IMF content between rabbit breeds and across developmental stages, and identified key genes and lncRNAs associated with IMF deposition [16]. However, the role of miRNAs in regulating IMF accumulation in rabbits remains largely unexplored. To address this gap, the present study integrated previously reported IMF phenotypic data with newly generated miRNA expression profiles to elucidate the potential regulatory functions of miRNAs in IMF deposition.

Materials and methods

Animal sample collection

In this study, we used Hycole and Rex rabbit breeds at three ages (35, 75, and 150 days). A total of 54 healthy male rabbits were selected, and divided into six groups with nine biological replicates per group (HR35, RR35, HR75, RR75, HR150, RR150; 18 rabbits per age stage). All experimental rabbits were selected from three different litters (one rabbit per litter) to avoid litter effects. As previously described in our earlier study [16], all rabbits were raised in the same feeding batch with uniform housing conditions (temperature: 22 ± 2 °C, photoperiod: 12 h light/12 h dark) at the Livestock Experimental Base of Northwest A&F University, China, with consistent commercial feed and free access to water and food. Prior to sample collection, all rabbits underwent a 24-hour fasting period without water. Subsequently, they were euthanized via intravenous injection of sodium pentobarbital (a dosage of 100 mg/kg) administered through the ear vein. Following euthanasia, the longissimus dorsi (LD) muscle was harvested, immediately frozen in liquid nitrogen, and stored at -80 °C for RNA extraction.

IMF content measurement

IMF content measurement methods were consistent with our previous research [16]. Briefly, IMF content in the LD muscle of rabbits at three key developmental stages (35, 75, and 150 days) was determined using the Soxhlet petroleum ether extraction method. Approximately 30 g of LD muscle was dehydrated to constant weight at 105 °C and ground. The weights of the sample before extraction (w), after drying (w1), and after extraction (w2) were recorded. Each sample was assayed in triplicate, and IMF content was calculated as follows: IMF content (%) = (w1 − w2) / w × 100%.

Total RNA extraction

Total RNA was extracted from the LD muscle tissues of Hycole and Rex rabbits (n = 18; three per group) using Trizol reagent (Invitrogen, Carlsbad, CA, USA) according to the manufacturer’s protocol. RNA purity was assessed by agarose gel electrophoresis. RNA concentration and integrity were assessed using a NanoDrop (Thermo Fisher Scientific, Waltham, MA, USA) and an Agilent 5400 Bioanalyzer (Santa Clara, CA, USA). Only samples with an RNA Integrity Number (RIN) exceeding 7.5 were subjected to subsequent sequencing analysis.

Library construction and sequencing

The miRNA cDNA library was constructed using NEB E7300L according to the manufacturer’s instructions. Briefly, total RNA was used as the starting material, and adapters were ligated to both ends of the small RNA, followed by reverse transcription to cDNA. The cDNA was then amplified via PCR, and target DNA fragments were separated using PAGE gel electrophoresis, followed by gel extraction to obtain the cDNA library. The quality of all libraries was evaluated using the Agilent 5400 Bioanalyzer (Agilent, USA), and quantification was performed via qPCR. Qualified small RNA cDNA libraries were sequenced on the Illumina platform (Beijing Novogene Technology Co., Ltd.) using the SE50 strategy.

Sequencing data processing

To ensure the quality of subsequent data analysis and the reliability of sequencing data, all raw data obtained from sequencing were filtered using fastp software [17]. Low-quality reads containing sequencing adapters were removed, resulting in clean data (Q20 ≥ 98% and Q30 ≥ 96%). MiRNAs within the length range of 18–40 nt were selected for downstream analysis. The filtered miRNAs were mapped to the rabbit reference genome (OryCun2.0) using Bowtie [18], and the distribution of miRNAs on the reference sequence was analyzed. New miRNAs were predicted based on the characteristic hairpin structure of miRNA precursors using miREvo [19] and mirDeep2 [20]. Finally, the expression levels of the identified miRNAs were quantified and normalized using transcripts per million (TPM).

Differential expression analysis

DE miRNAs between comparison groups were identified using DESeq2 [21]. A negative binomial distribution-based model was employed for statistical calculations of P values, and multiple testing correction was performed using the Benjamini-Hochberg method to calculate the false discovery rate (FDR), with the adjusted P value representing the FDR-corrected P value. The fold change was determined by the ratio of the average miRNA expression levels between the groups. DE miRNAs were selected based on the criteria of an adjusted P value < 0.05 and |log₂ (fold change)| > 1. Additionally, the target genes of the significantly differentially expressed miRNAs were predicted using TargetScan (score ≥ 50) and miRanda (energy < -10) [22].

Weighted Gene Coexpression Network Analysis (WGCNA)

In this study, a weighted miRNA co-expression network was constructed using the WGCNA package in R software. Following quality control of the expression profile matrix, the soft threshold was set at β = 7 to evaluate the scale-free network topology fit and average connectivity. The dynamic tree cut algorithm was applied to partition co-expression modules, after which module eigengenes (MEs) of each module were calculated. Pearson correlation analysis was performed to assess the correlations between these MEs and phenotypic traits of rabbits, including IMF content, breed, and age. The module showing the most significant correlation with IMF content was defined as the core module, and its hub miRNAs were extracted as key targets for subsequent functional analysis. Target genes of these hub miRNAs were predicted using miRanda software. The predicted target gene set was then subjected to functional intersection screening against a lipid metabolism-related gene set. Finally, the screened interaction pairs between core miRNAs and key target genes were imported into Cytoscape software to generate a visual regulatory network map of miRNA-mRNA targeting.

Functional enrichment analysis

Functional enrichment analysis of the DE miRNA target genes was performed using the clusterProfiler package in R software [23], based on two mainstream databases: Gene Ontology (GO) and the Kyoto Encyclopedia of Genes and Genomes (KEGG). For the GO enrichment analysis, the target genes of the DE miRNAs were mapped to various entries in the GO database, and hypergeometric tests were conducted to identify significantly enriched GO terms. The top 10 significantly enriched terms from the three GO categories—Biological Process (BP), Cellular Component (CC), and Molecular Function (MF)—were presented using bar plots. In the KEGG enrichment analysis, the target genes of the DE miRNAs were mapped to various pathway entries in the KEGG database, and the number of genes mapped to each pathway was counted; significant pathways were identified using hypergeometric tests. The results of the KEGG enrichment analysis were processed using the ggplot2 package in R software and presented in the form of bubble charts.

Real-time quantitative PCR validation

To validate the accuracy and reliability of the transcriptome sequencing data, RT-qPCR was conducted. Four DE miRNAs were randomly selected, with U6 serving as the internal control gene. Primers were designed using the NCBI Primer-BLAST online tool (Table S1). The amplification reactions were performed using 2×ChamQ SYBR qPCR Master Mix as the fluorescent dye, with a total reaction volume of 10 µL. Finally, the quantitative results were analyzed using the 2^−ΔΔCt method and compared with the transcriptome sequencing results.

Statistical analysis

All data in this study were statistically analyzed using SPSS 26.0 software, and visualized using GraphPad Prism (v8.0.2). Statistical analysis was conducted using the t-test, with data presented as means ± standard deviations. A P value of < 0.05 was considered statistically significant, and < 0.01 was considered highly significant.

Results

Differential IMF deposition in Hycole and Rex rabbits across developmental stages

To investigate the role of miRNAs in IMF deposition in rabbits, we collected LD muscle samples from both Hycole and Rex rabbits at three developmental ages: the weaned period (35 days), mid-growth period (75 days), and late-growth period (150 days). Through the measurement of IMF content, we revealed that the IMF levels in both breeds significantly increased with age (Fig. 1A and B, Table S2). At the same age, the IMF content in Rex rabbits was significantly higher than that in Hycole rabbits (Fig. 1C, Table S3).

Fig. 1.

Differential IMF deposition between Hycole and Rex rabbits across developmental stages. A Comparison of IMF content in Hycole rabbits at different ages. Different letters (a, b, c) indicate significant differences among groups (P < 0.05). B Comparison of IMF content in Rex rabbits at different ages. Different letters (a, b, c) indicate significant differences among groups (P < 0.05). C Comparison of IMF content between Hycole rabbits and Rex rabbits at the same age. “*” indicates a significant difference between groups (P < 0.05)

Small RNA sequencing of rabbit LD muscle at different ages

We constructed a total of 18 miRNA libraries from Hycole (n = 3) and Rex rabbits (n = 3) at the three different ages. As shown in Table S2, each library generated between 10,827,225 and 13,672,980 raw reads. After filtering out low-quality reads and sequences containing ambiguous bases (Q20 > 98.78%, Q30 > 96.39%), the clean reads accounted for 95.53% to 99.57% of the total reads per sample, yielding a high-quality sequence alignment percentage ranging from 91.14% to 97.07%. An analysis of the length distribution of miRNAs revealed that the lengths predominantly ranged from 21 to 24 nucleotides, with the majority being 22 nucleotides, which is the typical length of miRNAs generated by the Dicer enzyme (Figure S1A and S1B). Through miRNA alignment, identification, and predictive analysis, we identified a total of 631 miRNAs in the LD muscles of rabbits at different ages. We then performed TPM density distribution and correlation analysis based on the expression levels of miRNAs in the samples. The TPM density distribution results indicated minimal differences in expression between biological replicates for each sample, demonstrating sufficient reproducibility (Figure S2). Correlation analysis revealed high correlation coefficients among samples, indicating good clustering of expression levels (Figure S3).

Identification of differential miRNAs in Hycole and Rex rabbits at different growth stages

To investigate the role of miRNAs in IMF deposition in Hycole rabbits, we identified DE miRNAs across three distinct growth stages. The results are illustrated in Fig. 2. Compared to the HR35 group, the HR75 group exhibited 26 upregulated and 64 downregulated DE miRNAs (Fig. 2A), while the HR150 group displayed 71 upregulated and 110 downregulated DE miRNAs (Fig. 2B). In the comparison between the HR150 and HR75 groups, we observed 29 upregulated and 70 downregulated DE miRNAs (Fig. 2C). Hierarchical clustering of the DE miRNAs revealed two major clusters (Fig. 2D), with 89 miRNAs displaying a continuous upregulation trend from day 35 to 150, and 106 miRNAs displaying rapid downregulation between day 35 and 75 and subsequently remaining at low expression levels from day 75 to 150. Venn diagram analysis identified a total of 195 DE miRNAs across the three comparison groups, including 49 co-expressed DE miRNAs (i.e., miRNAs that were differentially expressed in all groups). Notably, this included miRNAs associated with IMF deposition, such as ocu-miR-29c-5p, ocu-miR-382-3p, and ocu-miR-29b-3p (Fig. 2E).

Fig. 2.

Differential expression analysis of miRNAs in the longissimus dorsi muscle of Hycole Rabbits at different growth stages. A Volcano plot of DE miRNAs in the comparison group HR75 vs. HR35. B Volcano plot of DE miRNAs in the comparison group HR150 vs. HR75. C Volcano plot of DE miRNAs in the comparison group HR150 vs. HR35. The x-axis represents the log₂ fold change in miRNA expression between different samples or comparison groups, while the y-axis indicates the significance level of the expression differences. Upregulated miRNAs are represented by blue dots, downregulated miRNAs by orange dots, and gray dots represent miRNAs that did not show significant changes. D Clustering heatmap of DE miRNAs at different time points in Hycole rabbits. Blue indicates high expression of miRNAs, while orange indicates low expression. E Venn diagram of DE miRNAs

To further explore the role of miRNAs in IMF deposition in Rex rabbits, we identified DE miRNAs across three distinct growth stages. In the comparison between the RR75 and RR35 groups, we identified a total of 102 DE miRNAs, consisting of 34 upregulated and 68 downregulated miRNAs (Fig. 3A). In the RR150 versus RR75 comparison, we found 130 DE miRNAs, with 64 upregulated and 66 downregulated (Fig. 3B). The RR150 versus RR35 comparison yielded 190 DE miRNAs, comprising 83 upregulated and 107 downregulated (Fig. 3C). Hierarchical clustering of the DE miRNAs identified two major clusters with conserved expression patterns in Hycole rabbits (Fig. 3D), with 98 miRNAs showing consistent upregulation from day 35 to 150 and 124 miRNAs exhibiting rapid downregulation during this period. According to the Venn diagram results, a total of 222 differential miRNAs were identified across the three comparison groups, with 45 miRNAs exhibiting differential expression in all groups. Notably, this included miRNAs associated with IMF deposition, such as ocu-miR-182-5p, ocu-miR-486-5p, and ocu-miR-370-5p (Fig. 3E).

Fig. 3.

Differential expression analysis of miRNAs in the longissimus dorsi muscle of Rex Rabbits at different growth stages. A Volcano plot of DE miRNAs in the comparison group RR75 vs. RR35. B Volcano plot of DE miRNAs in the comparison group RR150 vs. RR75. C Volcano plot of DE miRNAs in the comparison group RR150 vs. RR35. The x-axis represents the log₂ fold change in miRNA expression between different samples or comparison groups, while the y-axis indicates the significance level of the expression differences. Upregulated miRNAs are represented by blue dots, downregulated miRNAs by orange dots, and gray dots represent miRNAs that did not show significant changes. D Clustering heatmap of DE miRNAs at different time points in Rex rabbits. blue indicates high expression of miRNAs, while orange indicates low expression. E Venn diagram of DE miRNAs

Identification of differential miRNAs between different breeds at the same growth stage

To investigate whether miRNAs influence IMF deposition between different breeds, we compared miRNAs across the same growth stages of these two rabbit breeds, applying thresholds of adjusted P value < 0.05 and |log₂ (fold change)| > 1 to identify DE miRNAs. The results are shown in Fig. 4. In the comparisons of RR35 versus HR35, RR75 versus HR75, and RR150 versus HR150, we identified 12, 13, and 110 DE miRNAs (ocu-miR-143-3p, ocu-miR-27b-5p, ocu-let-7b-5p, ocu-miR-34a-5p, ocu-miR-486-5p) (Fig. 4A-C). According to the Venn diagram results (Fig. 4D), a total of 128 differential miRNAs were identified across the three comparison groups, with no overlapping DE miRNAs.

Fig. 4.

Differential expression analysis of miRNAs in the longissimus dorsi muscle of different rabbit breeds. A Volcano plot of DE miRNAs in the comparison group RR35 vs. HR35. B Volcano plot of DE miRNAs in the comparison group RR75 vs. HR75. C Volcano plot of DE miRNAs in the comparison group RR150 vs. HR150. The x-axis represents the log₂ fold change in miRNA expression between different samples or comparison groups, while the y-axis indicates the significance level of the expression differences. Upregulated miRNAs are represented by blue dots, downregulated miRNAs by orange dots, and gray dots represent miRNAs that did not show significant changes. D Venn diagram of DE miRNAs

Functional enrichment analysis of target genes for DE miRNAs in Hycole and Rex rabbits at different growth stages

To investigate the role of miRNAs in IMF deposition in Hycole and Rex rabbits, we utilized miRanda (miRanda-3.3a) for target gene prediction of the DE miRNAs, followed by GO and KEGG enrichment analyses of the predicted target genes. A total of 195 DE miRNAs were identified across the three comparison groups in Hycole rabbits at different growth stages, predicting 12,400 target genes (Table S5). GO enrichment analysis revealed 2,147 GO terms, which included 1,294 biological processes, 510 molecular functions, and 343 cellular components. Notably, the target genes of the DE miRNAs were significantly enriched in categories related to the cytoskeleton and other terms (P < 0.05) (Fig. 5A). KEGG pathway analysis indicated that the target genes of the DE miRNAs were enriched in 339 pathways, with five of these showing significant enrichment (P < 0.05). Among the top 20 pathways analyzed, we identified several linked to IMF deposition, including the PI3K-Akt signaling pathway, AMPK signaling pathway, JAK-STAT signaling pathway, and fructose and mannose metabolism (Fig. 5B).

Fig. 5.

Functional enrichment analysis of target genes for DE miRNAs in Hycole and Rex rabbits at different growth stages. A GO enrichment bar chart. The x-axis represents the GO terms, while the y-axis indicates the significance level of GO term enrichment, expressed as -log₁₀(P). Different colors represent various functional categories. B KEGG enrichment bubble plot. The x-axis represents the pathways, and the y-axis indicates the Rich Factor value, where a higher Rich Factor signifies greater enrichment in the KEGG pathways. C GO enrichment bar chart. The x-axis represents the GO terms, while the y-axis indicates the significance level of GO term enrichment, expressed as -log₁₀(P). Different colors represent various functional categories. D KEGG enrichment bubble plot. The x-axis represents the pathways, and the y-axis indicates the Rich Factor value, where a higher Rich Factor signifies greater enrichment in the KEGG pathways

To investigate the role of miRNAs in IMF deposition in Rex rabbits, we identified a total of 222 DE miRNAs across three comparison groups at different growth stages, predicting 14,432 target genes (Table S6). GO enrichment analysis (Fig. 5C) revealed a total of 2,147 GO terms, including 1,294 biological processes, 510 molecular functions, and 343 cellular components. Notably, the target genes of the DE miRNAs were significantly enriched in categories related to the kinetochore and condensed chromosome kinetochore (P < 0.05). KEGG pathway analysis indicated that the target genes of the DE miRNAs were enriched in 339 pathways. Among the top 20 pathways analyzed, several linked to IMF deposition were identified, including the PI3K-Akt signaling pathway, MAPK signaling pathway, p53 signaling pathway, and Hippo signaling pathway (Fig. 5D).

Functional enrichment analysis of target genes for DE miRNAs across different rabbit breeds

In the same growth stage, a total of 128 DE miRNAs were identified in two rabbit breeds, predicting 11,834 target genes. GO enrichment analysis (Fig. 6A) resulted in a total of 2,147 GO terms, including 1,294 biological processes, 510 molecular functions, and 343 cellular components. However, no significantly enriched terms were observed. KEGG pathway analysis (Fig. 6B) indicated that the target genes of the DE miRNAs were enriched in 339 pathways. Among the top 20 pathways analyzed, several linked to IMF deposition were identified, including the PI3K-Akt signaling pathway, Hedgehog signaling pathway, mTOR signaling pathway, and AMPK signaling pathway.

Fig. 6.

Functional enrichment analysis of Target Genes for DE miRNAs in different rabbit breeds. A GO enrichment bar chart. The x-axis represents the GO terms, while the y-axis indicates the significance level of GO term enrichment, expressed as -log₁₀(P). Different colors represent various functional categories. B KEGG enrichment bubble plot. The x-axis represents the pathways, and the y-axis indicates the Rich Factor value, where a higher Rich Factor signifies greater enrichment in the KEGG pathways

Identification of IMF-related core miRNA modules and hub miRNAs by WGCNA

To investigate the molecular regulatory network underlying intramuscular fat (IMF) deposition in rabbits from a systems biology perspective, we employed weighted gene co-expression network analysis (WGCNA) on miRNA transcriptomic data. For network construction, a soft-thresholding power of β = 7 was selected, as this value resulted in a network that closely approximated a scale-free topology while maintaining high mean connectivity (Fig. 7A). Using this threshold and the dynamic tree cut algorithm, miRNAs with similar expression profiles were clustered into co-expression modules, and highly similar modules (e.g., the yellow and green modules) were subsequently merged. This process ultimately identified five distinct co-expression modules, designated as blue, red, turquoise, green, and brown (Fig. 7B).

Fig. 7.

WGCNA identifies IMF-related core miRNA modules and hub miRNAs. A Selection of soft-thresholding power for the weighted miRNA co-expression network. (Left) Scale-free topology model fit at different soft-thresholding powers. (Right) Mean connectivity at different soft-thresholding powers. B Clustering dendrogram of miRNA co-expression modules. C Heatmap of module-trait relationships. Rows represent module eigengenes (MEs), and columns represent three phenotypic traits: IMF content, Breed, and Age. Each cell contains the correlation coefficient with the corresponding P value in parentheses. The color gradient ranges from blue (negative correlation) to red (positive correlation). D Regulatory network of key miRNAs and their potential target genes related to lipid metabolism. The yellow oval nodes represent the four hub miRNAs identified by WGCNA. The blue oval nodes indicate the key target genes predicted by miRanda. E KEGG enrichment bubble plot. The x-axis indicates the Rich Factor value, and the y-axis represents the pathways

To pinpoint the key modules associated with IMF deposition, we performed Pearson correlation analyses between the module eigengenes (MEs) of these five modules and key phenotypic traits, including IMF content, breed, and age (Fig. 7C). The results revealed that the brown module exhibited an extremely significant positive correlation with IMF content (r = 0.76, P = 3e-04), establishing it as the core regulatory module most closely associated with IMF deposition in rabbits. In addition, the red module was significantly positively correlated with breed (r = 0.74), whereas the turquoise module showed a strong negative correlation with age (r = -0.90). Given its direct and robust association with the IMF phenotype, miRNAs within the brown module were designated as hub miRNAs and served as the primary focus for subsequent target gene prediction and network construction.

To further investigate the potential biological functions of these hub miRNAs in IMF deposition, their target genes were predicted using miRanda software. Focusing on lipid metabolism-related genes, we constructed a regulatory network comprising four key miRNAs (ocu-miR-423-5p, ocu-miR-1343-3p, ocu-miR-150-5p, ocu-miR-30b-3p) and their 20 target genes (Fig. 7D). Subsequent KEGG pathway enrichment analysis indicated that these target genes were significantly enriched in pathways related to lipid metabolism and fatty acid degradation (Fig. 7E).

Validation of DE miRNAs by RT-qPCR

To verify the accuracy and reliability of transcriptome sequencing data, we conducted RT-qPCR validation on four DE miRNAs from the RNA-seq results. The qPCR results demonstrate that the expression trends of the DE miRNAs in Hycole rabbits (Fig. 8A-D) and Rex rabbits (Fig. 8E-H) closely align with the RNA-Seq findings. This indicates that the RNA-Seq results from this study are accurate.

Fig. 8.

RT-qPCR Validation of DE miRNAs. A-D Comparison of RT-qPCR data and RNA-Seq data for DE miRNAs in Hycole rabbit samples. Blue represents RT-qPCR data, while red represents RNA-Seq data. E-H Comparison of RT-qPCR data and RNA-Seq data for DE miRNAs in Rex rabbit samples. Blue represents RT-qPCR data, while red represents RNA-Seq data

Discussion

IMF deposition is a crucial factor affecting meat quality and economic efficiency, particularly in meat-producing animals, where the distribution and accumulation of IMF directly influence the flavor, texture, and nutritional value of the meat [24]. As an important meat source, the quality traits of rabbit meat are significantly impacted by IMF deposition patterns. miRNAs play essential roles in regulating gene expression and are widely involved in various physiological processes, such as cell proliferation [25], differentiation [26], and apoptosis [27]. Recent studies have indicated that miRNAs can impact fat metabolism and related physiological phenomena by targeting specific genes. However, the expression characteristics of miRNAs in rabbit longissimus muscle and their functions in IMF deposition processes remain unclear. In this study, we utilized miRNA sequencing to characterize changes in miRNA expression during IMF deposition at different growth stages in rabbits.

Using RNA sequencing technology, we compared miRNA expression profiles across different growth stages. We identified 195 DE miRNAs in Hycole rabbits at various developmental periods, with a total of 49 co-expressed DE miRNAs, including IMF deposition-related miRNAs such as ocu-miR-29c-5p, ocu-miR-382-3p, ocu-miR-29b-3p, ocu-miR-370-3p, ocu-miR-136-5p, ocu-miR-495-3p, and ocu-miR-655-3p. Ocu-miR-29c-5p was identified as the key downregulated miRNA in HR. Target gene prediction revealed that it specifically targets the 3’-UTR of PPARγ, a critical adipogenic transcription factor, with a conserved seed region (5’-UAGCACCA-3’) and a binding free energy of -14.2 kcal/mol (miRanda). Notably, miR-29c is closely associated with IMF in chickens [28], and our target gene prediction showed that ocu-miR-29b-3p targets PPARγ, consistent with its role in upregulating PPARγ to promote adipocyte differentiation [29]. Additionally, miR-29b-3p promotes the proliferation and differentiation of preadipocytes by targeting IGF1 [30], while also inhibiting the proliferation and inflammation of vascular smooth muscle in atherosclerosis [31]. MiR-370-3p has been found to regulate pre-adipocyte growth by targeting Mknk1 [32], and is closely related to fat deposition in pigs, mediating IMF deposition through the regulation of CPT1 expression, thereby improving growth traits in pork [33]. miR-655 has also been shown to target and inhibit the expression of the SIRT2 gene, promoting lipogenesis in bovine mammary epithelial cell lines [34].

In different periods of Rex rabbits, a total of 45 DE miRNAs were identified as co-expressed, including miR-182-5p, miR-486-5p, and miR-370-5p, which are associated with IMF deposition. For ocu-miR-182-5p, the core miRNA in RR, our study found that it targets rabbit C/EBPα and FASN. Notably, studies have shown that miR-182 negatively regulates adipogenesis by targeting C/EBPα [35]. miR-182-5p promotes the expression and secretion of FGF21 in adipocytes by inhibiting the nuclear receptor subfamily 1 group D member 1 (Nr1d1), thus enhancing thermogenesis in beige adipose tissue [36]. Additionally, miR-182-5p also alleviates high-fat diet-induced nonalcoholic steatohepatitis in mice [37]. These findings indicate that miR-182-5p has evolved a species-specific pathway preference in rabbits, which is the molecular basis for the higher IMF deposition rate in RR. Furthermore, the conserved miRNA ocu-miR-486-5p, which is associated with fat deposition in multiple species, exhibits breed-specific expression patterns in rabbits. MiR-486 has been shown to suppress fat generation in the femoral head through TBX2 [38], and it influences subcutaneous fat deposition in pigs via the PPAR signaling pathway [39]. Importantly, this study identified co-expressed miRNAs related to fat deposition, such as miR-432-5p, miR-299-3p, miR-127-5p, and miR-495-3p, which were found in both the co-expressed genes of different growth stages of the Hycole rabbit and the Rex rabbit. MiR-432 has been implicated in IMF deposition and differentiation in both pigs and sheep [40, 41]. In a study on the miRNA transcriptome of local pig breeds, miR-432, miR-29, and miR-127 were found to be significantly downregulated in the skeletal muscle of adult pigs, and experimental evidence has demonstrated that miR-127 can inhibit both muscle development and fat deposition in the porcine LD muscle [42]. Additionally, miR-432 has also been associated with IMF deposition and differentiation in sheep [41]. Furthermore, the number of DE miRNAs among different breeds is significantly lower than that across different growth stages, indicating that age is a more critical factor influencing IMF deposition, which is consistent with previous research findings [16].

WGCNA analysis further confirmed the direct correlation between key miRNAs and IMF deposition. ocu-miR-423-5p, ocu-miR-1343-3p, ocu-miR-150-5p, and ocu-miR-30b-3p were identified as hub miRNAs in the brown module (r = 0.76). The overall expression level of miRNAs in this module was significantly upregulated with increasing IMF content, suggesting that these hub miRNAs may serve as critical drivers regulating IMF deposition in rabbits. Previous studies reported that ocu-miR-423-5p promotes adipogenic differentiation of 3T3-L1 cells by binding to DVL3 and inhibiting its expression. This miRNA has also been associated with lipid metabolism and adipogenesis in pigs and dairy cows [43, 44].During porcine backfat development, ocu-miR-1343-3p directly targets TCF3 and FASN at different stages, thereby regulating adipocyte hyperplasia and hypertrophy [45].Meanwhile, miR-1343-3p modulates lipid metabolism by targeting ACAA1, thus affecting androgenetic alopecia [46].miR-150-5p targets and suppresses GDF11 expression, thereby alleviating bone marrow fat accumulation in diabetes [47].

Through an in-depth analysis of DE miRNAs and their target genes, we explored the roles of these miRNAs in fat metabolism-related signaling pathways. Our study demonstrated that the DE genes were significantly enriched in several key pathways, including the PI3K-Akt pathway, MAPK pathway, mTOR pathway, Hippo pathway, and HIF-1 signaling pathway. These enriched pathways are closely related to lipogenesis, metabolism, and obesity [48–50]. The PI3K-Akt signaling pathway regulates glucose production and lipid metabolism by modulating various downstream pathways, such as mTORC1, glycogen synthase kinase, FoxO transcription factors, and the activation of lipogenic factors (Fasn, ACC, Scd1, and Elovl6) [51]. The MAPK pathway plays a crucial role in energy homeostasis and normal blood glucose levels, mediating reductions in circulating thyroid hormone levels, thyroid hormone-dependent gene expression, and oxidative metabolic rates, which collectively decrease energy expenditure and promote obesity [52]. However, this study has several inherent limitations that should be acknowledged. First, the small sample size (three biological replicates per group, n = 3) may limit the statistical power, particularly for detecting subtle differences in molecular expression or phenotypic traits. Second, the regulatory mechanisms of the identified key miRNAs and their target genes were not experimentally validated in vitro. To address these limitations, future work will focus on the following directions: (1) increasing the sample size to 5–6 rabbits per group to improve statistical robustness; and (2) conducting in vitro functional validation using rabbit primary intramuscular preadipocytes. This will include dual-luciferase reporter assays to verify direct miRNA–target gene interactions, as well as functional assessments of lipid droplet accumulation, adipocyte differentiation, and intramuscular fat deposition following modulation of key molecules.

Hence, the findings of this study contribute to a better understanding of the role of miRNAs in IMF deposition in rabbits. The identification of specific DE miRNAs and their target genes provide a foundation for further functional studies aimed at elucidating the molecular mechanisms underlying fat metabolism. This knowledge could ultimately enhance meat quality and production efficiency in rabbit breeding programs. Future research should focus on validating the functional roles of these miRNAs and exploring their potential as biomarkers or targets for genetic selection in meat-producing rabbits.

Conclusion

This study enhances our understanding of IMF deposition in rabbits by identifying key DE miRNAs across growth stages in Hycole and Rex breeds. Notable miRNAs, such as ocu-miR-29c-5p, ocu-miR-182-5p, and ocu-miR-370-5p, are linked to fat metabolism and involve critical signaling pathways like PI3K-Akt and MAPK. The varying miRNA expression profiles highlight the influence of age on IMF accumulation, suggesting that targeted genetic selection could enhance meat quality traits in rabbits. This research lays the groundwork for further studies on these miRNAs and their potential to improve rabbit meat production.

Abbreviations

miRNA

microRNA

sRNA

Small RNA

IMF

Intramuscular fat

HR

Hycole rabbit

RR

Rex rabbit

LD

Longissimus dorsi

DE

Differentially expressed

GO

Gene Ontology

KEGG

Kyoto Encyclopedia of Genes and Genomes

PPAR

Peroxisome proliferator-activated receptor

ACACA

Acetyl-CoA carboxylase alpha

ACSL4

Acyl-CoA synthetase 4

TPM

Transcripts per million

BP

Biological Process

CC

Cellular Component

MF

Molecular Function

RT-qPCR

Real-Time Quantitative PCR

Authors’ contributions

XD conceived and designed the project. GS, HW and ZJ performed the experiments and analyzed the data. GS and ZJ performed the visualization. GH, HW, QD, and IA collected samples. SW and RZ recruited animal resources. GS, HW, and XD wrote and revised the manuscript. All authors read and approved the final draft.

Funding

This work was supported by the Chinese Universities Scientific Fund (no. 2452020187) and Xi’an Science and Technology Plan Project (no. 25NJSYB00008).

Data availability

The raw sequence datasets from this study have been deposited in the NCBI SRA database under BioProject ID PRJNA1272316.

Declarations

Ethics approval and consent to participate

All animal procedures in this study were performed in accordance with the guidelines of the Chinese Association for Laboratory Animal Sciences and were approved by the Institutional Animal Care and Use Committee of Northwest A&F University (Protocol No. 202205A29).

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.