Skip to main content
NCBI home page
BMC Genomics logoLink to BMC Genomics
. 2026 Jun 6;27:667. doi: 10.1186/s12864-026-13033-x

A circRNA-miRNA-mRNA network potentially regulates skeletal muscle adaptation to hypoxia in Tibetan sheep

Pengfei Zhao 1, Yaqin Zhang 1, Jing Ma 1, Nanjie Dangzi 1, Wenlu Ma 1, Xiong Ma 1,✉
PMCID: PMC13463523  PMID: 42251254

Abstract

Background

The extreme environmental conditions of high-altitude regions, particularly hypoxia, pose significant challenges to animal survival. Tibetan sheep serve as an exemplary model for studying mammalian hypoxic adaptation. Skeletal muscle, crucial for movement and energy metabolism, undergoes adaptive restructuring, yet the underlying molecular mechanisms, particularly the role of the competitive endogenous RNA (ceRNA) network, remain poorly understood. This study aimed to investigate the morphological changes in skeletal muscle across different altitudes and elucidate the potential ceRNA-mediated regulatory mechanisms.

Results

Our results revealed that Tibetan sheep from Qinghai-Tibet Plateau high-altitude (QH) regions exhibited significantly greater muscle fiber diameter and area in the biceps femoris compared to those from lower altitudes. To explore the underlying molecular mechanisms, a circRNA-miRNA-mRNA ceRNA network was constructed, identifying several potential key regulatory axes. These included novel_circ_003862/miR-2440-z/LPL, novel_circ_015124/miR-20-y/LGR4, and novel_circ_012468/miR-2440-z/E2F6. Notably, the expression of LPL and E2F6 was upregulated, while LGR4 was downregulated in the QH group, suggesting a potential coordinated shift towards glycolytic metabolism to maintain energy supply under hypoxic stress.

Conclusions

These findings suggest that hypoxic environments may promote the adaptive remodeling of skeletal muscle fibers by modulating a specific ceRNA network. The altered expression of key genes like LPL, LGR4, and E2F6 potentially enhances glycolytic metabolism to cope with hypoxic stress. This study provides new insights into the molecular genetic basis of high-altitude adaptation in mammals, though experimental validation is required.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12864-026-13033-x.

Keywords: Tibetan Sheep, Hypoxia Adaptation, Biceps Femoris, ceRNA Network

Introduction

The extreme environmental conditions prevalent in high-altitude regions, characterized primarily by hypoxia, low temperatures, and intense ultraviolet radiation, present significant challenges to animal survival. Therefore, these regions serve as ideal natural laboratories for the study of animal adaptive evolution [1–3]. The Qinghai-Tibet Plateau, often termed the “Roof of the World”, is home to a diverse array of unique species that have evolved through prolonged natural selection and domestication. Notably, Tibetan sheep represent not only a vital resource for the local animal husbandry economy but also serve as exemplary models for studying mammalian adaptation to hypoxic environments [4]. To survive and thrive in such challenging conditions, Tibetan sheep have evolved a complex array of adaptive mechanisms across various physiological, biochemical, and genetic dimensions [5–8]. As skeletal muscle constitutes approximately 40% of mammalian body mass and is fundamental for locomotion and energy metabolism [9], its adaptive restructuring is crucial for the resilience of Tibetan sheep against the environmental stresses of the plateau.

The fundamental unit of skeletal muscle is the muscle fiber [10], and adaptive changes in skeletal muscle are primarily characterized by transformations in fiber types. In ruminant species, muscle fibers are commonly classified into types I (slow-twitch oxidative), IIA (fast-twitch oxidative), and IIX (fast-twitch oxidative-glycolytic) based on distinct myosin heavy chain isoforms [11, 12]. Type I fibers are the thinnest in diameter, rich in mitochondria, and entirely dependent on aerobic metabolism, exhibiting exceptional fatigue resistance. Type IIA fibers possess both oxidative and glycolytic capacities, with intermediate contraction speed and fatigue resistance between types I and IIX. Type IIX fibers are the thickest in diameter, primarily depend on glycolysis for energy, contracting rapidly but are prone to fatigue [13]. Consequently, we hypothesize that the proportion of type IIX muscle fibers-which rely heavily on glycolysis for energy-will increase with altitude to compensate for reduced oxygen availability and maintain motor function [14]. Consequently, we hypothesize that the proportion of type IIX muscle fibers, which depend on anaerobic glycolysis, will increase to adapt to reduced oxygen levels and maintain necessary motor functions. The current study chose the biceps femoris, a predominant muscle in the hind limb essential for supporting and facilitating movement in animals, as the subject of investigation. Given the significant variations in diameter among different muscle fiber types, we can deduce alterations in the composition ratio of the biceps femoris muscle fibers through the measurement of the average diameter and cross-sectional area of the fibers [13]. Though direct molecular typing using specific markers (e.g., Myosin heavy chain isoforms) is required for definitive validation.

The conversion of muscle fiber types is a sophisticated biological process, fundamentally characterized by modifications in gene expression profiles. Apart from messenger RNAs (mRNAs) that serve as the executors of cellular functions, a substantial array of non-coding RNAs are pivotal in the post-transcriptional regulation of gene expression [15–17]. Notably, circular RNAs (circRNAs) exert their influence by sequestering specific microRNAs (miRNAs), thereby indirectly enhancing the expression of other target mRNAs associated with these miRNAs, culminating in the formation of a circRNA-miRNA-mRNA competitive endogenous RNA (ceRNA) regulatory network [18]. Mounting evidence indicates that the ceRNA network is central to the regulation of muscle development and the transformation of muscle fiber types [19–21]. Consequently, we hypothesize that the ceRNA network is a fundamental molecular mechanism mediating the conversion of muscle fiber types within the biceps femoris of Tibetan sheep.

To date, inquiries into the hypoxia adaptability of Tibetan sheep have predominantly targeted physiological systems such as blood [22, 23], cardiovascular [24], and pulmonary [6, 25]. In contrast, there is a notable lack of research dedicated to skeletal muscle, particularly investigations that synthesize the analysis of muscle fiber morphological alterations with the ceRNA regulatory network. This study bridges this gap by employing a cohesive approach that amalgamates morphological assessment with high-throughput transcriptome sequencing, thereby elucidating the adaptive changes in the biceps femoris of Tibetan sheep across varying altitude gradients and their underlying molecular mechanisms. The findings of this research contribute to the discovery of novel mechanisms of muscle regulation in Tibetan sheep’s adaptation to high-altitudes, enhance our comprehension of the molecular genetic underpinnings of environmental adaptability in mammals, and offer potential molecular markers and a theoretical framework for the genetic enhancement of sheep breeds in high-altitude regions.

Materials and methods

Experimental design and sample acquisition

This study was designed to investigate the effects of different altitudes on the muscle characteristics of Tibetan sheep. A total of 15 three-year-old castrated male Tibetan sheep were selected from three distinct altitudinal gradients on the Qinghai-Tibet Plateau: high-altitude (QH, 4500 m, Maduo County, Guoluo Tibetan Autonomous Prefecture, Qinghai Province, China); and middle (QM, 3500 m, Maqu County) and low-altitude (QL, 2500 m, Xiahe County), both in Gannan Tibetan Autonomous Prefecture, Gansu Province, China. Five clinically healthy sheep of similar body weight were randomly chosen from each altitude. All selected sheep were reared under traditional year‑round free‑range grazing conditions without artificial supplementation, and the three sites shared a similar alpine steppe ecosystem with common dominant forages [26, 27]. Sampling in July, when the temperature difference between altitudes is smaller compared to other seasons, collectively reduced the potential confounding effects of diet and temperature on muscle phenotypes. The detailed environmental parameters of the sampling sites are presented in Table 1.

Table 1.

Environmental parameters of the three sampling sites for Tibetan sheep

Group Sampling Site Altitude (m) Mean July Temperature (°C) O2 Partial Pressure (kPa / % of Sea Level) Grassland Type Dominant Forage Species
QL Quao Town, Xiahe County ~ 2500 9.0–22.0 ~ 15.7 / ~74% Alpine steppe Kobresia, Stipa
QM Oula Town, Maqu County ~ 3500 6.0–17.0 ~ 13.8 / ~65% Alpine steppe / Alpine meadow Kobresia, Stipa
QH Huashixia Town, Maduo County ~ 4500 4.0–16.0 ~ 12.1 / ~57% Alpine steppe Kobresia, Stipa
Open in a new tab

Data on altitude and temperature were obtained from local government official websites. Partial pressure of oxygen was derived from reference [14]. Information on grassland types and dominant forage species was cited from references [26, 27]

To ensure animal welfare, all sheep were slaughtered in strict accordance with industry standards. The procedure involved electrical stunning followed by immediate exsanguination to minimize animal suffering. Immediately after exsanguination, samples of the biceps femoris muscle were rapidly collected from the left side of each carcass. Each muscle sample was divided into two portions: one was fixed in 4% paraformaldehyde for subsequent histological analysis of muscle fiber diameter and area, while the other was snap-frozen in liquid nitrogen and stored at − 80 °C for transcriptome sequencing.

Histological analysis of biceps femoris muscle fibers

To evaluate the impact of altitude on the morphological characteristics of muscle fibers in Tibetan sheep, a histological analysis of the biceps femoris was conducted. The muscle tissue samples, previously fixed in 4% paraformaldehyde, were rinsed under running water and then dehydrated through a graded series of ethanol solutions (75%, 85%, 95%, and 100%). Following dehydration, the tissues were cleared with xylene and embedded in paraffin to form paraffin blocks. The paraffin blocks were sectioned into 5 μm thick slices using a Leica 2016 rotary microtome (Leica, Wetzlar, Germany). The sections were then mounted on glass slides and subjected to routine Hematoxylin and Eosin staining.

The stained sections were examined and imaged using a BA210 digital microscope (Motic, Xiamen, China) at 400× magnification. For quantitative analysis, the diameter and cross-sectional area of 10 intact and clearly demarcated muscle fibers were randomly measured from per sample using Image-Pro Plus 6.0 software. All statistical analyses were performed using SPSS 25.0 software (SPSS, Chicago, IL, USA). A one-way analysis of variance was conducted to compare the differences in muscle fiber diameter and cross-sectional area among the three altitude groups. The results are presented as mean ± standard deviation, and a p-value of less than 0.05 was considered statistically significant.

Transcriptome sequencing

To analyze the gene expression profiles, total RNA was isolated from the biceps femoris of five sheep per altitude group using TRIzol reagent (Invitrogen, Carlsbad, CA, USA). The integrity of the extracted RNA was visualized on a 1% agarose gel, while its purity and concentration were quantified with a Nanodrop 2000 (Thermo Fisher Scientific, Waltham, MA, USA) and a Qubit® 2.0 Fluorometer (Life Technologies, Carlsbad, CA, USA). RNA Integrity Number (RIN) was assessed using an Agilent 2100 Bioanalyzer (Agilent Technologies, Palo Alto, CA, USA). Only high-quality samples (concentration > 80 ng/µL, A260/A280 ratio between 1.8 and 2.1, and RIN > 7) were used. For circRNA analysis, ribosomal RNA was depleted using the Ribo-Zero Gold rRNA Removal Kit (Illumina, San Diego, CA, USA), and cDNA libraries were constructed with the NEBNext Ultra RNA Library Prep Kit (New England Biolabs, Ipswich, MA, USA). For small RNA profiling, 18–30 nt fragments were isolated via polyacrylamide gel electrophoresis, ligated with 3’ and 5’ adaptors, and reverse-transcribed to generate cDNA libraries using the NEBNext® Multiplex Small RNA Library Prep Set for Illumina (New England Biolabs, Ipswich, MA, USA). Library quality was validated using the Agilent 2100 Bioanalyzer, and all libraries were sequenced by Gene Denovo Biotechnology Co., Ltd. (Guangzhou, China).

Identification of circRNA

To ensure the integrity of downstream analyses, the raw sequencing data were initially processed using fastp v.0.18.0 to generate high-quality clean reads. This step involved discarding reads with adapter sequences, those containing more than 10% ambiguous nucleotides, and low-quality reads where over 50% of bases had a Q-value ≤ 20 [28]. Following this, to eliminate any residual rRNA, the clean reads were aligned to an rRNA database using Bowtie2 v.2.2.8, and all mapped reads were excluded [29]. The remaining, rRNA-depleted reads were then aligned to the reference genome (ARS-UI_Ramb_v3.0, https://www.ncbi.nlm.nih.gov/datasets/genome/GCF_016772045.2/) with HISAT2 v.2.1.1 [30]. Crucially, for circRNA discovery, we collected the reads that failed to align to the genome, as these are candidates for circular transcripts. For the identification of circRNAs, 20-mer anchor sequences were extracted from both ends of these unmapped reads and aligned back to the reference genome. Pairs of anchors aligning in a reverse, head-to-tail orientation, indicative of back-splicing, were processed with the CIRIquant tool [31]. A circRNA was considered valid if it was supported by at least two unique back-spliced reads in at least one sample.

Identification of miRNA

The analysis of small RNAs began with a stringent filtering of the raw reads to obtain clean tags. This process removed reads with low-quality bases (Q-value ≤ 20) or ambiguous nucleotides, those lacking a 3’ adapter, and any sequences containing a 5’ adapter without an intervening small RNA fragment. Additional exclusions were made for reads containing poly(A) tails and those shorter than 18 nucleotides in length [28]. To isolate miRNAs, the resulting clean tags were systematically compared against the GeneBank (Release 209.0) and Rfam (Release 11.0) databases. Any tags matching known non-miRNA small RNAs, such as rRNA, scRNA, snoRNA, snRNA, and tRNA, were discarded [32]. Furthermore, tags were aligned to the reference genome (ARS-UI_Ramb_v3.0) to eliminate those originating from mRNA degradation (mapping to exons or introns) or from repetitive regions [30]. The remaining unannotated tags were then used for miRNA identification. Known miRNAs for the studied species were identified by searching the miRBase database (Release 22) [33]. For species not represented in miRBase, a cross-species alignment approach was employed to find conserved miRNAs. Finally, novel miRNA candidates were predicted from the remaining tags by mapping them to the genome and assessing their potential to form characteristic hairpin structures using the mirdeep2 software [34]. All tags were then assigned a final annotation based on a hierarchical priority system, and those that could not be classified were labeled as unannotated.

Quantification and differential expression analysis

To facilitate quantitative comparisons, normalization methods were applied to correct for variations in sequencing depth. The abundance of circRNAs was assessed by expressing back-spliced junction reads as Reads Per Million mapped reads. Similarly, expression levels for the total miRNA pool, encompassing existing, known, and novel miRNAs, were calculated and normalized to Transcripts Per Million. Subsequently, differential expression analysis for both circRNAs and miRNAs was performed using the edgeR package (v3.12.1) [35]. A circRNA or miRNA was considered significantly differentially expressed if it demonstrated a fold change of at least 2 (|log2FC| ≥ 1) with a p-value less than 0.05 in a given comparison [36].

Functional enrichment and expression pattern analysis

To investigate the functions of differentially expressed circRNAs (DEcircRNAs) and miRNAs (DEmiRNAs), we first predicted the target genes for all identified miRNAs (including existing, known, and novel) using a consensus approach with Miranda v.3.3a (http://www.microrna.org) and TargetScan v.7.0 (https://www.targetscan.org/vert_70/). To ensure high-confidence, only the overlapping predictions from both tools were retained as candidate targets. Subsequently, Gene Ontology (GO, http://www.geneontology.org/) and Kyoto Encyclopedia of Genes and Genomes (KEGG, http://www.genome.jp/kegg/) enrichment analyses were performed on two gene sets: the source genes of DEcircRNAs and the predicted targets of DEmiRNAs, using the cluster-Profiler R package v.4.16.0 [37–39]. Significantly enriched GO terms and KEGG pathways were identified by applying Fisher’s exact test, with a significance threshold of q < 0.05 [40].

To elucidate the dynamic expression patterns of DEcircRNAs and DEmiRNAs, their expression levels were first calculated as log2FC relative to the L group. These normalized values were then subjected to clustering analysis using the Short Time-series Expression Miner software [41]. Model profiles with a p-value ≤ 0.05 were deemed statistically significant.

Construction of the circRNA-miRNA-mRNA ceRNA network

To investigate the regulatory mechanisms governing muscle fiber type composition and transformation in Tibetan sheep across different altitudes, a ceRNA network was constructed. This analysis integrated the expression profiles of DEcircRNAs, DEmiRNAs, and our existing DEmRNAs data (the detailed expression changes of the DEmRNAs, including log2FC and p-values, are listed in Table S1) from the biceps femoris tissue. The fundamental principle was to identify circRNAs that act as ‘sponges’ for miRNAs, thereby modulating mRNA expression. First, potential circRNA-miRNA and miRNA-mRNA regulatory pairs were screened based on a significant negative correlation in their expression levels, using Spearman’s rank correlation coefficient (SCC < − 0.5) with a significance threshold of p < 0.05. Subsequently, to establish the co-regulatory relationship between the circRNA and its target mRNA, a strong positive correlation (Pearson’s r > 0.7, p < 0.05) was required. By integrating these criteria, where a circRNA and an mRNA share a negatively correlated miRNA and are themselves positively correlated, the final circRNA-miRNA-mRNA ceRNA axes were assembled. The resulting network was visualized using a Sankey diagram, and key hub nodes within this network were identified as candidate regulators potentially involved in the high-altitude adaptation of skeletal muscle fibers.

Results

Morphological differences in the biceps femoris muscle fibers of Tibetan sheep at different altitudes

No significant differences were found in the biceps femoris muscle fiber diameter and area between the QL and QM groups. However, both muscle fiber diameter and area in the QH group were significantly larger than those in the QL and QM groups (p < 0.05) (Fig. 1).

Fig. 1.

Fig. 1

Open in a new tab

Effects of altitude on biceps femoris muscle fiber diameter (A) and area (B) in Tibetan sheep from Qinghai-Tibet Plateau low (QL), middle (QM), and high (QH) altitude groups. Biceps femoris muscle fibers under an optical microscope are shown (QL, QM, and QH), asterisks (*) indicate significant differences (p < 0.05), and ‘ns’ indicates no significant difference (p > 0.05)

Characteristics of circRNAs

Analysis of biceps femoris tissues from Tibetan sheep at three altitudes led to the identification of a total of 69,026 circRNAs (Table S2, S3). Specifically, the QL, QM, and QH groups yielded 13,587, 15,216, and 13,739 unique circRNAs, respectively, with 14,730 circRNAs being common to all three groups (Fig. 2A). Chromosomal mapping revealed that these circRNAs were broadly distributed across the Tibetan sheep genome, with a significant enrichment observed on chromosomes 1–3 (Fig. 2B). Furthermore, the majority of the identified circRNAs were less than 1000 bp in length (Fig. 2C). In terms of sequence classification, circRNAs derived from annot-exons were the most prevalent, constituting 76.99% of the total, followed by exon-intron (12.86%) and intronic (3.58%) types (Fig. 2D).

Fig. 2.

Fig. 2

Open in a new tab

Characteristics of circRNAs in the biceps femoris of Tibetan sheep across three altitudes (QL, QM, and QH). A Venn diagram of identified circRNAs. B Chromosomal distribution of circRNAs. C Length distribution of circRNAs. D Classification of circRNA types

Characteristics of miRNAs

Sequencing of the biceps femoris samples from the QL, QM, and QH groups yielded an average of 11,581,767, 11,185,965, and 12,731,379 raw reads, respectively (Table S4, S5). Following the removal of reads containing adapters, low-quality bases, poly(A) tails, or unknown nucleotides, an average of 10,917,712, 10,606,383, and 11,793,089 clean reads were obtained. Subsequently, 81.51%, 76.93%, and 75.68% of these clean reads were successfully mapped to the sheep reference genome for the QL, QM, and QH libraries, respectively.

Analysis of the length distribution revealed that the majority of small RNAs were between 21 and 23 nucleotides, with 22 nucleotides being the most abundant class, constituting 35.81%, 35.41%, and 35.93% of the total reads in the QL, QM, and QH groups, respectively (Fig. 3A). Principal component analysis demonstrated that samples from the same altitude tended to cluster together, suggesting that altitude is a key factor influencing the overall miRNA expression landscape (Fig. 3B). Furthermore, existing and known miRNAs formed the predominant component of the small RNA libraries for all three groups (Fig. 3C). In total, 638 miRNAs were identified in the biceps femoris tissues of Tibetan sheep across the three altitudes. Among these, 553 were co-expressed across all groups, while 9, 15, and 24 miRNAs were unique to the QL, QM, and QH groups, respectively (Fig. 3D).

Fig. 3.

Fig. 3

Open in a new tab

Characteristics of miRNAs in the biceps femoris of Tibetan sheep from the QL, QM, and QH groups. A Length distribution of small RNAs. B Principal component analysis of samples based on miRNA expression. C Composition of small RNA types. D Venn diagram of identified miRNAs

Differential expression of circRNA and miRNA

To investigate the impact of altitude on non-coding RNA expression, we compared the expression profiles of circRNAs and miRNAs in the biceps femoris of Tibetan sheep. The analysis of circRNAs revealed a total of 555, 525, and 501 DEcircRNAs in the QL-vs-QM, QM-vs-QH, and QL-vs-QH comparisons, respectively (Table S6). This included 315, 261, and 288 upregulated DEcircRNAs, alongside 240, 264, and 213 downregulated DEcircRNAs (Fig. 4A-C). Similarly, the comparison of miRNA expression identified 107, 107, and 46 DEmiRNAs for the QL-vs-QM, QM-vs-QH, and QL-vs-QH groups, respectively (Table S7). Of these, 38, 76, and 36 were significantly upregulated, while 69, 31, and 10 were downregulated (Fig. 4D-F).

Fig. 4.

Fig. 4

Open in a new tab

Differential expression analysis of circRNAs and miRNAs in the biceps femoris of Tibetan sheep from the QL, QM, and QH groups. A, D Bar graphs quantify the specific numbers of upregulated and downregulated DEcircRNA and DEmiRNA in the QL-vs-QM, QM-vs-QH, and QL-vs-QH comparisons. B, E Scatter plots illustrate the expression-level changes of individual RNAs, where the spatial distribution of the dots represents the magnitude of expression differences. C, F Upset plots depict the distribution of DEcircRNA and DEmiRNA datasets across different altitudinal comparisons, highlighting specific RNAs that are shared or unique between consecutive altitude transitions

GO and KEGG enrichment analyses of DEcircRNAs source genes

To explore the potential functions of the identified DEcircRNAs, GO and KEGG enrichment analyses were performed on their source genes (Table S8). The GO analysis revealed significant enrichment across all three altitudinal comparisons. In the QL-vs-QM comparison, source genes were enriched in 208 cellular components, 364 molecular functions, and 1387 biological processes, with key terms associated with skeletal muscle adaptation to hypoxia, such as myofibril (GO:0030016), sarcomere (GO:0030017), ATP binding (GO:0005524), and cell differentiation (GO:0030154) (Fig. 5A). For the QM-vs-QH comparison, 189 cellular components, 235 molecular functions, and 1017 biological processes terms were enriched. Including key GO terms such as myofibril (GO:0030016), ATP binding (GO:0005524), muscle system process (GO:0003012), and regulation of TOR signaling (GO:0032006) (Fig. 5B). The QL-vs-QH comparison enriched 235 cellular components, 221 molecular functions, and 991 biological processes terms, with a focus on muscle fiber transformation, including myofibril (GO:0030016), cellular catabolic process (GO:0044248), muscle tissue development (GO:0060537), and cell differentiation (GO:0030154) (Fig. 5C).

Fig. 5.

Fig. 5

Open in a new tab

GO and KEGG enrichment analyses of source genes for DEcircRNAs in Tibetan sheep biceps femoris. A-C The top 20 enriched GO terms in the QL-vs-QM, QM-vs-QH, and QL-vs-QH comparisons, respectively. D-F The top 20 enriched KEGG pathways for the same comparisons

KEGG pathway analysis further elucidated the underlying biological pathways. The source genes from the QL-vs-QM comparison were significantly enriched in 56 pathways, prominently featuring the mTOR signaling pathway (ko04150), and HIF-1 signaling pathway (ko04066) (Fig. 5D). For the QM-vs-QH comparison, 66 pathways were enriched, with the MAPK signaling pathway (ko04010), VEGF signaling pathway (ko04370), and mTOR signaling pathway (ko04150) (Fig. 5E). In addition, the QL-vs-QH comparison identified 27 enriched pathways, primarily related to Glucagon signaling pathway (ko04922), and Growth hormone synthesis, secretion and action (ko04935) (Fig. 5F).

GO and KEGG enrichment analyses of DEmiRNA target genes

GO term annotation analysis revealed that the target genes of DEmiRNAs between the QL and QM groups were significantly enriched in 649 cellular component, 882 molecular function, and 4739 biological process terms. Notably, several terms closely associated with muscle development and energy metabolism were identified, such as sarcolemma (GO:0042383), ATP binding (GO:0005524), cell differentiation (GO:0030154), and regulation of metabolic process (GO:0019222) (Fig. 6A). Similarly, in the QM-vs-QH comparison, GO analysis showed significant enrichment of DEmiRNA target genes in 607 cellular components, 880 molecular functions, and 4689 biological processes. Terms strongly related to skeletal muscle development were highlighted, including myofibril (GO:0030016), cell differentiation (GO:0030154), regulation of metabolic process (GO:0019222), and muscle system process (GO:0003012) (Fig. 6B). For the QL-vs-QH group, GO analysis identified significant enrichment in 672 cellular components, 983 molecular functions, and 5275 biological process terms. Among these, key terms associated with muscle development and differentiation were discerned, such as sarcolemma (GO:0042383), MAP kinase activity (GO:0008349), cell differentiation (GO:0030154), and response to oxygen-containing compound (GO:1901700) (Fig. 6C) (Table S9).

Fig. 6.

Fig. 6

Open in a new tab

GO and KEGG enrichment analyses of target genes of DEmiRNAs in the biceps femoris of Tibetan sheep. A-C The top 20 enriched GO terms for the QL-vs-QM, QM-vs-QH, and QL-vs-QH comparisons, respectively. D-F The top 20 enriched KEGG pathways for the same comparisons

KEGG pathway analysis further elucidated the underlying biological pathways (Table S9). The target genes from the QL-vs-QM comparison were significantly enriched in 176 pathways, prominently featuring the MAPK signaling pathway (ko04010), Notch signaling pathway (ko04330), mTOR signaling pathway (ko04150), and VEGF signaling pathway (ko04370) (Fig. 6D). For the QM-vs-QH comparison, 170 pathways were enriched, including the MAPK signaling pathway (ko04010), Notch signaling pathway (ko04330), mTOR signaling pathway (ko04150), Carbohydrate digestion and absorption (ko04973), and VEGF signaling pathway (ko04370) (Fig. 6E). Furthermore, the QL-vs-QH comparison identified 164 enriched pathways, mainly involving the MAPK signaling pathway (ko04010), Notch signaling pathway (ko04330), mTOR signaling pathway (ko04150), Carbohydrate digestion and absorption (ko04973), HIF-1 signaling pathway (ko04066), Wnt signaling pathway (ko04310), and Fatty acid metabolism (ko01212) (Fig. 6F).

Expression Pattern Analysis of DEcircRNAs and DEmiRNAs

To investigate the dynamic expression patterns of DEcircRNAs and DEmiRNAs across altitudes, we performed a trend analysis. This analysis classified 1304 DEcircRNAs into eight distinct profiles, three of which (Profiles 5, 6, and 7) showed significant enrichment as altitude increased (p < 0.05) (Fig. 7A). Specifically, DEcircRNAs in Profile 5 were initially upregulated and then downregulated, those in Profile 6 were upregulated before stabilizing, and DEcircRNAs in Profile 7 displayed a consistent upregulation with increasing altitude. A parallel analysis of 181 DEmiRNAs also identified eight expression profiles, with Profile 2 being significantly enriched (p < 0.05). DEmiRNAs within this profile exhibited a pattern of initial downregulation followed by a significant upregulation as altitude increased (Fig. 7B).

Fig. 7.

Fig. 7

Open in a new tab

Expression profiles of DEcircRNAs and DEmiRNAs in Tibetan sheep biceps femoris. A Eight expression profiles for 1304 DEcircRNAs. B Eight expression profiles for 181 DEmiRNAs. The profiles were clustered based on expression trends across the QL, QM, and QH groups. Each profile box displays the p-value for enrichment, the number of included RNAs, and the trend line. Representative mRNAs for relevant profile also listed

Construction of the circRNA-miRNA-mRNA ceRNA Network

To elucidate the regulatory mechanisms involving circRNAs, we constructed a ceRNA network. First, we identified the target mRNAs and circRNAs for 181 DEmiRNAs across the three altitudinal groups. This analysis yielded 2044 miRNA-mRNA pairs involving 920 DEmRNAs and 1268 circRNA-miRNA pairs involving 1,282 DEcircRNAs, all with a SCC less than − 0.5 and a p-value < 0.05. By integrating these interaction datasets, we then filtered for high-confidence ceRNA pairs, retaining those with a Pearson’s r greater than 0.70 and a p-value < 0.05. This process resulted in a final network comprising 133 significant ceRNA pairs, which are visualized as a Sankey diagram in Fig. 8.

Fig. 8.

Fig. 8

Open in a new tab

The circRNA-miRNA-mRNA ceRNA network. The Sankey diagram illustrates the 133 significant ceRNA pairs identified between DEcircRNAs, DEmiRNAs, and DEmRNAs. The thickness of the ribbons represents the number of interactions, and the network was constructed based on a positive correlation analysis (Pearson’s r > 0.70, p < 0.05)

Discussion

This study utilized transcriptome sequencing to investigate the effects of altitude on the expression profiles of circRNAs, miRNAs, and mRNAs in the skeletal muscle of Tibetan sheep, aiming to correlate these molecular shifts with alterations in muscle fiber phenotype. Our results revealed that Tibetan sheep from QH regions exhibited significantly greater muscle fiber diameter and area (Fig. 1), Morphologically, type IIX muscle fibers are characterized by having the largest diameter among fiber types [13]. Therefore, the observed increase in fiber diameter and area is consistent with a shift in fiber composition towards type IIX fibers. This inference aligns with prior research indicating that hypoxic exposure induces significant remodeling of skeletal muscle, often characterized by a marked increase in the proportion of Type IIx fibers [42]. Possessing the largest diameter, these fibers primarily generate energy through glycolysis [13]. This metabolic strategy, reducing fatty acid and carbohydrate oxidation while enhancing glycolysis, appears to conserve limited oxygen under hypoxic conditions [43, 44]. Consequently, when glycolytic activity is heightened, ATP production can proceed with a corresponding decrease in the organism’s oxygen demand.

circRNAs are recognized as key regulators in the processes of skeletal muscle fiber transformation and energy metabolism [45–48]. Similarly, miRNAs exert their influence by targeting and suppressing key mRNAs crucial for myogenesis and muscle development, ultimately dictating muscle fiber type composition and growth [49–51]. Our transcriptome sequencing data revealed that high-altitude environments significantly influence the expression profiles of both circRNAs and miRNAs within the biceps femoris muscle of Tibetan sheep (Fig. 4). Functional analysis, including GO term annotation and KEGG pathway enrichment, highlighted that the DEcircRNAs were mainly implicated in muscle fiber transformation and energy metabolism. Notably, their source genes were significantly enriched in the mTOR, HIF-1, MAPK, and VEGF signaling pathways (Fig. 5). In parallel, the DEmiRNAs were found to primarily regulate muscle fiber type and growth, with their target genes enriched in a network of pathways including MAPK, Notch, mTOR, VEGF, HIF-1 and Wnt signaling pathways (Fig. 6). The MAPK signaling pathway actively facilitates muscle fiber transformation [52, 53], while the Notch signaling pathway critically regulates muscle fiber regeneration [54]. Acting as a central regulator, the mTOR pathway governs essential processes including muscle growth and energy metabolism [55–57]. In response to hypoxia, an interplay between the VEGF and HIF-1 signaling pathways promotes angiogenesis, thereby sustaining the energy and oxygen supply [58, 59]. Likewise, within these hypoxic environments, the HIF-1 pathway engages in crosstalk with the Wnt pathway, where hypoxia-induced HIF-1α modulates Wnt signaling to govern myogenesis [60, 61]. These findings suggest that the circRNAs and miRNAs identified in this study act as key players in the molecular machinery governing muscle fiber transformation and energy metabolism in response to hypoxic stress, likely through modulation of the pathways discussed.

By acting as a molecular sponge, circRNA can competitively bind with miRNA, thereby mitigating the miRNA’s inhibitory impact on its target mRNA [18]. Consequently, this research delineated a circRNA-miRNA-mRNA interaction network. The data predicted that novel_circ_003862 and novel_circ_004421 might attenuate the inhibition of LPL by interacting with miR-2440-z, miR-9985-z, and novel-m0122-5p. Similarly, novel_circ_015124, novel_circ_027921, novel_circ_043788, and novel_circ_051007 may alleviated the inhibition of LGR4 through binding with miR-20-y. Additionally, novel_circ_003862 and novel_circ_012468 may lessened the inhibition of E2F6 by interacting with miR-2440-z (Fig. 8). In this study, the expression levels of the LPL and E2F6 genes were significantly elevated in the biceps femoris of Tibetan sheep from the QH group, while the expression of the LGR4 gene was significantly reduced (Fig. 7). This expression pattern is highly significant when considering the distinct functions of these genes. Research has found that mice with overexpressed LPL in skeletal muscle accumulate triglycerides and develop insulin resistance, thereby maintaining hyperglycemia [62]. Conversely, the absence of LPL in skeletal muscle reduces lipid storage and enhances insulin signaling, leading to lower blood glucose [63]. Mouse models with LGR4 overexpression exhibit enhanced gluconeogenesis, which results in reduced blood glucose [64]. In contrast, LGR4 deficiency enhances the flexibility of skeletal muscle in response to glucose depletion, enabling it to switch its metabolic substrate from glucose to fatty acids. This suggests that LGR4 may play an adaptive regulatory role in the balance between glucose and lipid metabolism in skeletal muscle [65]. Maintaining high blood glucose, as facilitated by high LPL and low LGR4, helps to provide an ample substrate for the glycolytic reaction in the skeletal muscle of QH group sheep, ensuring a sufficient energy supply. Furthermore, E2F6 functions to regulate E2F activity and protects myocytes from apoptosis, which contributes to maintaining cellular homeostasis under extreme environmental conditions [66]. In summary, LPL, LGR4, and E2F6 may collectively maintain the glycolytic level of muscle fibers in the biceps femoris of the QH group Tibetan sheep. The aforementioned circRNA-miRNA-mRNA interaction network in which they participate may therefore play a crucial role in the molecular mechanism of hypoxic adaptation in the skeletal muscle of Tibetan sheep.

Despite these insights, several limitations must be acknowledged to objectively define the scope of our findings. First, the shift toward type IIX fibers was inferred from morphological data and literature, lacking direct molecular validation using specific markers (e.g., Myosin heavy chain immunofluorescence). Second, the constructed ceRNA network and the authenticity of novel circRNAs are based on bioinformatic predictions and sequencing data without experimental validation, such as luciferase reporter assays or Sanger sequencing. Third, this study focused exclusively on the biceps femoris, which may not fully represent the systemic adaptive mechanisms across different muscle types. Future studies employing in vitro and in vivo functional assays are essential to verify these specific regulatory axes and to comprehensively elucidate the multi-tissue ceRNA networks underlying high-altitude adaptation.

Conclusion

In conclusion, this study demonstrates that high-altitude induces significant phenotypic adaptation in the skeletal muscle of Tibetan sheep, characterized by increased muscle fiber diameter and cross-sectional area in the biceps femoris. This adaptive morphological change is potentially associated with a complex ceRNA regulatory network. Key circRNAs, such as novel_circ_003862, novel_circ_015124, and novel_circ_012468, are predicted to function as molecular sponges for specific miRNAs (e.g., miR-2440-z, miR-20-y), thereby modulating the expression of critical target genes LPL, LGR4, and E2F6. These genes may collaboratively sustain a glycolytic metabolic profile and cellular homeostasis in muscle fibers under hypoxic stress. While this research provides novel insights into the molecular underpinnings of hypoxic adaptation, subsequent functional validation is imperative to confirm these regulatory axes and to assess their utility as molecular markers for breeding livestock resilient to high-altitude stress.

Supplementary Information

Supplementary Material 1. (1.3MB, xlsx)
Supplementary Material 2. (18.1KB, xlsx)
Supplementary Material 3. (16.4MB, xlsx)
Supplementary Material 4. (15.2KB, xlsx)
Supplementary Material 5. (295.1KB, xlsx)
Supplementary Material 6. (1.6MB, xlsx)
Supplementary Material 7. (67.4KB, xlsx)
Supplementary Material 8. (359.5KB, xlsx)
Supplementary Material 9. (60.4MB, xlsx)

Acknowledgements

We are grateful to our Tibetan friends Jiayangdaji and Daojirenqing for their enthusiastic help during sample collection.

Authors’ contributions

**P. Z.:** Resources, Supervision, Writing-original draft. **Y. Z.:** Validation, Data curation. **J. M.:** Formal analysis, Visualization. **N. D.:** Investigation, project administration. **W. M.:** Conceptualization. **X. M.:** Writing–review & editing, Funding acquisition. All authors reviewed the manuscript.

Funding

This research was funded by the 2024 Gansu Province Universities Industrial Support Plan Projects (2024CYZC-59); the 2024 Gansu Province Youth Science and Technology Fund (24JRRP006); and the 2025 Gansu Province Universities Young Doctor Support Projects (2025QB-098).

Data availability

Sequence data that support the findings of this study have been deposited in the NCBI Sequence Read Archive with the primary accession code PRJNA1289012 and PRJNA1284595.

Declarations

Ethics approval and consent to participate

All experimental animals used in this study were treated strictly in accordance with the Animal Ethics Procedures and Guidelines of the People’s Republic of China. The experimental procedures were reviewed and approved by the Animal Administration and Ethics Committee of Gansu Minzu Normal University (approval number: GMNU-EC-2024-022).

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.

Footnotes

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

References

  • 1.Storz JF, Cheviron ZA. Physiological Genomics of Adaptation to High-Altitude Hypoxia. Annu Rev Anim Biosci. 2021;9:149–71. 10.1146/annurev-animal-072820-102736. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Storz JF. High-Altitude Adaptation: Mechanistic Insights from Integrated Genomics and Physiology. Mol Biol Evol. 2021;38(7):2677–91. 10.1093/molbev/msab064. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Berger MM, Luks AM. High Altitude. Semin Respir Crit Care Med. 2023;44(5):681–95. 10.1055/s-0043-1770063. [DOI] [PubMed] [Google Scholar]
  • 4.Hu X, Yang J, Xie X, et al. The Genome Landscape of Tibetan Sheep Reveals Adaptive Introgression from Argali and the History of Early Human Settlements on the Qinghai-Tibetan Plateau. Mol Biol Evol. 2019;36(2):283–303. 10.1093/molbev/msy208. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Wang G, He Y, Luo Y. Expression of OPA1 and Mic60 genes and their association with mitochondrial cristae morphology in Tibetan sheep. Cell Tissue Res. 2019;376(2):273–9. 10.1007/s00441-018-2975-y. [DOI] [PubMed] [Google Scholar]
  • 6.Zhao P, Zhao F, Hu J, et al. Physiology and Transcriptomics Analysis Reveal the Contribution of Lungs on High-Altitude Hypoxia Adaptation in Tibetan Sheep. Front Physiol. 2022;13:885444. 10.3389/fphys.2022.885444. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Han B, Tian D, Li X, et al. Multiomics Analyses Provide New Insight into Genetic Variation of Reproductive Adaptability in Tibetan Sheep. Mol Biol Evol. 2024;41(3):msae058. 10.1093/molbev/msae058. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Liang X, Duan Q, Li B, et al. Genomic structural variation contributes to evolved changes in gene expression in high-altitude Tibetan sheep. Proc Natl Acad Sci USA. 2024;121(27):e2322291121. 10.1073/pnas.2322291121. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Bogomolova AP, Katrukha IA. Troponins and Skeletal Muscle Pathologies. Biochem (Mosc). 2024;89(12):2083–106. 10.1134/S0006297924120010. [DOI] [PubMed] [Google Scholar]
  • 10.Frontera WR, Ochala J. Skeletal muscle: a brief review of structure and function. Calcif Tissue Int. 2015;96(3):183–95. 10.1007/s00223-014-9915-y. [DOI] [PubMed] [Google Scholar]
  • 11.Chaosap C, Sitthigripong R, Sivapirunthep P, et al. Myosin heavy chain isoforms expression, calpain system and quality characteristics of different muscles in goats. Food Chem. 2020;321:126677. 10.1016/j.foodchem.2020.126677. [DOI] [PubMed] [Google Scholar]
  • 12.Song S, Ahn CH, Kim GD. Muscle Fiber Typing in Bovine and Porcine Skeletal Muscles Using Immunofluorescence with Monoclonal Antibodies Specific to Myosin Heavy Chain Isoforms. Food Sci Anim Resour. 2020;40(1):132–44. 10.5851/kosfa.2019.e97. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Lefaucheur L. A second look into fibre typing–relation to meat quality. Meat Sci. 2010;84(2):257–70. 10.1016/j.meatsci.2009.05.004. [DOI] [PubMed] [Google Scholar]
  • 14.Beall CM. Adaptation to High Altitude: Phenotypes and Genotypes. Annu Rev Anthropol. 2014;43:251–72. 10.1146/annurev-anthro-102313-030000. [Google Scholar]
  • 15.Guttman M, Rinn JL. Modular regulatory principles of large non-coding RNAs. Nature. 2012;482(7385):339–46. 10.1038/nature10887. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Kaikkonen MU, Lam MT, Glass CK. Non-coding RNAs as regulators of gene expression and epigenetics. Cardiovasc Res. 2011;90(3):430–40. 10.1093/cvr/cvr097. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.An S, Song J. The coded functions of noncoding RNAs for gene regulation. Mol Cells. 2011;31(6):491–6. 10.1007/s10059-011-1004-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Tay Y, Rinn J, Pandolfi PP. The multilayered complexity of ceRNA crosstalk and competition. Nature. 2014;505(7483):344–52. 10.1038/nature12986. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Zhang Y, Yan H, Zhou P, et al. MicroRNA-152 Promotes Slow-Twitch Myofiber Formation via Targeting Uncoupling Protein-3 Gene. Animals. 2019;9(9):669. 10.3390/ani9090669. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Li M, Zhang N, Zhang W, et al. Comprehensive analysis of differentially expressed circRNAs and ceRNA regulatory network in porcine skeletal muscle. BMC Genomics. 2021;22(1):320. 10.1186/s12864-021-07645-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Yun J, Huang X, Liu C, et al. Genome-wide analysis of circular RNA-mediated ceRNA regulation in porcine skeletal muscle development. BMC Genomics. 2023;24(1):196. 10.1186/s12864-023-09284-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Zhao P, He Z, Xi Q, et al. Variations in HIF-1α Contributed to High Altitude Hypoxia Adaptation via Affected Oxygen Metabolism in Tibetan Sheep. Animals. 2021;12(1):58. 10.3390/ani12010058. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Zhao P, Ma X, Ren J, et al. Variations in HBA gene contribute to high-altitude hypoxia adaptation via affected O2 transfer in Tibetan sheep. Front Zool. 2024;21(1):30. 10.1186/s12983-024-00551-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Wen Y, Li S, Zhao F, et al. Changes in the Mitochondrial Dynamics and Functions Together with the mRNA/miRNA Network in the Heart Tissue Contribute to Hypoxia Adaptation in Tibetan Sheep. Animals. 2022;12(5):583. 10.3390/ani12050583. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Zhao P, Li S, He Z, et al. Physiology and Proteomic Basis of Lung Adaptation to High-Altitude Hypoxia in Tibetan Sheep. Animals. 2022;12(16):2134. 10.3390/ani12162134. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Miehe G, Bach K, Miehe S, et al. Alpine steppe plant communities of the Tibetan highlands. Appl Veg Sci. 2011;14:547–60. 10.1111/j.1654-109X.2011.01147.x. [Google Scholar]
  • 27.Yuan H, Matthew C, He X, et al. Seasonal Variation in Soil and Herbage CO2 Efflux for a Sheep-Grazed Alpine Meadow on the North-East Qinghai-Tibetan Plateau and Estimated Net Annual CO2 Exchange. Front Plant Sci. 2022;13:860739. 10.3389/fpls.2022.860739. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Chen S, Zhou Y, Chen Y, et al. fastp: an ultra-fast all-in-one FASTQ preprocessor. Bioinformatics. 2018;34(17):i884–90. 10.1093/bioinformatics/bty560. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Langmead B, Salzberg SL. Fast gapped-read alignment with Bowtie 2. Nat Methods. 2012;9(4):357–9. 10.1038/nmeth.1923. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Kim D, Langmead B, Salzberg SL. HISAT: a fast spliced aligner with low memory requirements. Nat Methods. 2015;12(4):357–60. 10.1038/nmeth.3317. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Zhang J, Chen S, Yang J, et al. Accurate quantification of circular RNAs identifies extensive circular isoform switching events. Nat Commun. 2020;11(1):90. 10.1038/s41467-019-13840-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32.Kalvari I, Nawrocki EP, Argasinska J, et al. Non-Coding RNA Analysis Using the Rfam Database. Curr Protoc Bioinf. 2018;62(1):e51. 10.1002/cpbi.51. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Kozomara A, Griffiths-Jones S. miRBase: annotating high confidence microRNAs using deep sequencing data. Nucleic Acids Res. 2014;42:D68–73. 10.1093/nar/gkt1181. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34.Mackowiak SD. Identification of novel and known miRNAs in deep-sequencing data with miRDeep2. Curr Protoc Bioinf. 2011;Chap1212101–121015. 10.1002/0471250953.bi1210s36. [DOI] [PubMed]
  • 35.Robinson MD, McCarthy DJ, Smyth GK. edgeR: a Bioconductor package for differential expression analysis of digital gene expression data. Bioinformatics. 2010;26(1):139–40. 10.1093/bioinformatics/btp616. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36.Anders S, Huber W. Differential expression analysis for sequence count data. Genome Biol. 2010;11(10):R106. 10.1186/gb-2010-11-10-r106. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Yu G, Wang L, Han Y, et al. clusterProfiler: an R package for comparing biological themes among gene clusters. OMICS. 2012;16(5):284–7. 10.1089/omi.2011.0118. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38.Ashburner M, Ball CA, Blake JA, et al. Gene ontology: tool for the unification of biology. The Gene Ontology Consortium. Nat Genet. 2000;25(1):25–9. 10.1038/75556. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39.Ogata H, Goto S, Sato K, et al. KEGG: Kyoto Encyclopedia of Genes and Genomes. Nucleic Acids Res. 1999;27(1):29–34. 10.1093/nar/27.1.29. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40.Huang da W, Sherman BT, Zheng X, et al. Extracting biological meaning from large gene lists with DAVID. Curr Protoc Bioinf. 2009. 10.1002/0471250953.bi1311s27. Chap. 13: Unit 13.11. [DOI] [PubMed] [Google Scholar]
  • 41.Ernst J, Bar-Joseph Z. STEM: a tool for the analysis of short time series gene expression data. BMC Bioinformatics. 2006;7:191. 10.1186/1471-2105-7-191. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 42.Chaillou T. Skeletal Muscle Fiber Type in Hypoxia: Adaptation to High-Altitude Exposure and Under Conditions of Pathological Hypoxia. Front Physiol. 2018;9:1450. 10.3389/fphys.2018.01450. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 43.Horscroft JA, Kotwica AO, Laner V, et al. Metabolic basis to Sherpa altitude adaptation. Proc Natl Acad Sci USA. 2017;114(24):6382–7. 10.1073/pnas.1700527114. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 44.An Z, Wei D, Wei L, et al. The expression of Ldh-c in the skeletal muscle of plateau pika (Ochotona curzoniae) enhances adaptation to a hypoxic environment. Biol Open. 2017;6(9):1336–41. 10.1242/bio.024943. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 45.Cao H, Liu J, Du T, et al. Circular RNA screening identifies circMYLK4 as a regulator of fast/slow myofibers in porcine skeletal muscles. Mol Genet Genomics. 2022;297(1):87–99. 10.1007/s00438-021-01835-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 46.Shen L, Gan M, Tang Q, et al. Comprehensive Analysis of lncRNAs and circRNAs Reveals the Metabolic Specialization in Oxidative and Glycolytic Skeletal Muscles. Int J Mol Sci. 2019;20(12):2855. 10.3390/ijms20122855. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 47.Huang K, Chen M, Zhong D, et al. Circular RNA Profiling Reveals an Abundant circEch1 That Promotes Myogenesis and Differentiation of Bovine Skeletal Muscle. J Agric Food Chem. 2021;69(1):592–601. 10.1021/acs.jafc.0c06400. [DOI] [PubMed] [Google Scholar]
  • 48.Wu P, Zhou K, Zhang J, et al. Identification of crucial circRNAs in skeletal muscle during chicken embryonic development. BMC Genomics. 2022;23(1):330. 10.1186/s12864-022-08588-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 49.Silva WJ, Graça FA, Cruz A, et al. miR-29c improves skeletal muscle mass and function throughout myocyte proliferation and differentiation and by repressing atrophy-related genes. Acta Physiol (Oxf). 2019;226(4):e13278. 10.1111/apha.13278. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 50.Wang Y, Zhang D, Yu S, et al. miR-196-5p regulates myogenesis and induces slow-switch fibers formation by targeting PBX1. Int J Biol Macromol. 2025;305(Pt 1):141137. 10.1016/j.ijbiomac.2025.141137. [DOI] [PubMed] [Google Scholar]
  • 51.Wang X, Chen X, Huang Z, et al. MicroRNA-499-5p regulates porcine myofiber specification by controlling Sox6 expression. Animal. 2017;11(12):2268–74. 10.1017/S1751731117001008. [DOI] [PubMed] [Google Scholar]
  • 52.Luo X, Zhang H, Cao X, et al. Endurance Exercise-Induced Fgf21 Promotes Skeletal Muscle Fiber Conversion through TGF-β1 and p38 MAPK Signaling Pathway. Int J Mol Sci. 2023;24(14):11401. 10.3390/ijms241411401. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 53.Wang T, Sun X, Zhang Y, et al. Baicalin Promotes Skeletal Muscle Fiber Remodeling by Activating the p38MAPK/PGC-1α Signaling Pathway. J Agric Food Chem. 2025;73(11):6878–89. 10.1021/acs.jafc.5c00300. [DOI] [PubMed] [Google Scholar]
  • 54.Hindi SM, Paul PK, Dahiya S, et al. Reciprocal interaction between TRAF6 and notch signaling regulates adult myofiber regeneration upon injury. Mol Cell Biol. 2012;32(23):4833–45. 10.1128/MCB.00717-12. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 55.Rion N, Castets P, Lin S, et al. mTOR controls embryonic and adult myogenesis via mTORC1. Development. 2019;146(7):dev172460. 10.1242/dev.172460. [DOI] [PubMed] [Google Scholar]
  • 56.Zhang P, Liang X, Shan T, et al. mTOR is necessary for proper satellite cell activity and skeletal muscle regeneration. Biochem Biophys Res Commun. 2015;463(1–2):102–8. 10.1016/j.bbrc.2015.05.032. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 57.Saxton RA, Sabatini DM. mTOR Signaling in Growth, Metabolism, and Disease. Cell. 2017;169(2):361–71. 10.1016/j.cell.2017.03.035. [DOI] [PubMed] [Google Scholar]
  • 58.Shin J, Nunomiya A, Kitajima Y, et al. Prolyl hydroxylase domain 2 deficiency promotes skeletal muscle fiber-type transition via a calcineurin/NFATc1-dependent pathway. Skelet Muscle. 2016;6:5. 10.1186/s13395-016-0079-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 59.Breen E, Tang K, Olfert M, et al. Skeletal muscle capillarity during hypoxia: VEGF and its activation. High Alt Med Biol. 2008;9(2):158–66. 10.1089/ham.2008.1010. [DOI] [PubMed] [Google Scholar]
  • 60.Majmundar AJ, Lee DS, Skuli N, et al. HIF modulation of Wnt signaling regulates skeletal myogenesis in vivo. Development. 2015;142(14):2405–12. 10.1242/dev.123026. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 61.Cirillo F, Resmini G, Ghiroldi A, et al. Activation of the hypoxia-inducible factor 1α promotes myogenesis through the noncanonical Wnt pathway, leading to hypertrophic myotubes. FASEB J. 2017;31(5):2146–56. 10.1096/fj.201600878R. [DOI] [PubMed] [Google Scholar]
  • 62.Wang H, Eckel RH. Lipoprotein lipase: from gene to obesity. Am J Physiol Endocrinol Metab. 2009;297(2):E271–88. 10.1152/ajpendo.90920.2008. [DOI] [PubMed] [Google Scholar]
  • 63.Wang H, Knaub LA, Jensen DR, et al. Skeletal muscle-specific deletion of lipoprotein lipase enhances insulin signaling in skeletal muscle but causes insulin resistance in liver and other tissues. Diabetes. 2009;58(1):116–24. 10.2337/db07-1839. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 64.Fang Q, Ye L, Han L, et al. LGR4 is a key regulator of hepatic gluconeogenesis. Free Radic Biol Med. 2025;229:183–94. 10.1016/j.freeradbiomed.2025.01.025. [DOI] [PubMed] [Google Scholar]
  • 65.Sun Y, Hong J, Chen M, et al. Ablation of Lgr4 enhances energy adaptation in skeletal muscle via activation of Ampk/Sirt1/Pgc1α pathway. Biochem Biophys Res Commun. 2015;464(2):396–400. 10.1016/j.bbrc.2015.06.066. [DOI] [PubMed] [Google Scholar]
  • 66.Major JL, Salih M, Tuana BS. E2F6 protein levels modulate drug induced apoptosis in cardiomyocytes. Cell Signal. 2017;40:230–8. 10.1016/j.cellsig.2017.09.022. [DOI] [PubMed] [Google Scholar]

Associated Data

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

Supplementary Materials

Supplementary Material 1. (1.3MB, xlsx)
Supplementary Material 2. (18.1KB, xlsx)
Supplementary Material 3. (16.4MB, xlsx)
Supplementary Material 4. (15.2KB, xlsx)
Supplementary Material 5. (295.1KB, xlsx)
Supplementary Material 6. (1.6MB, xlsx)
Supplementary Material 7. (67.4KB, xlsx)
Supplementary Material 8. (359.5KB, xlsx)
Supplementary Material 9. (60.4MB, xlsx)

Data Availability Statement

Sequence data that support the findings of this study have been deposited in the NCBI Sequence Read Archive with the primary accession code PRJNA1289012 and PRJNA1284595.

Articles from BMC Genomics are provided here courtesy of BMC

RESOURCES