Gestational Feed Restriction Induces Differential Expression of Novel Long Non‐Coding RNAs in Offspring Skeletal Muscle Authors: Maria M. M. Muniz, Thaís C. Costa, Lívia S. Rodrigues, Luiza V. Kladt, Luiz Jardel M. Motta, Mariana M. Lopes, Nicola V. L. Serão, Mateus P. Gionbelli, Marcio Duarte Published: 2026 ABSTRACT ABSTRACTThis study aimed to investigate how maternal feed restriction at different gestational stages alters the expression of novel long non‐coding RNAs (lncRNAs) in the skeletal muscle of goat offspring. LncRNAs are emerging as important regulators of biological processes, including myogenesis, yet their role in fetal programming remains largely unexplored. Using RNA sequencing (RNA‐Seq), we analyzed lncRNA expression in the Longissimus thoracis muscle of kids born to goats submitted to two feeding diets: early gestation restriction (50% of maintenance from Day 8–84 of gestation, then 100% until parturition; RM, n = 6) and late gestation restriction (100% from Day 8–84, then 50% until parturition; MR, n = 8). RNA‐Seq data underwent quality control, alignment, transcript assembly, and differential expression analysis using Cufflinks and FEELnc pipelines. Three differentially expressed (DE; FDR ≤ 0.05) lncRNAs were identified and classified based on genomic position and predicted regulatory potential. The upregulated lncRNA_10996.1 in the RM group was linked to acyl‐CoA dehydrogenase family member 10 and GTP‐binding nuclear protein Ran‐like genes. The downregulated lncRNAs, lncRNA_1180.1 and lncRNA_7992.1, were associated with ZNF546, PSMC4, SNORA65, ZNF79, LRSAM1, and RPL12 genes. Overall, these results suggest that maternal nutrition restriction is associated with alterations in lncRNA expression profile and may influence muscle development through potential regulatory interactions with protein‐coding genes. These findings identified candidate lncRNAs and provide exploratory insights into the molecular mechanisms potentially involved in fetal programming in livestock. Introduction The efficiency and sustainability of livestock production are crucial in the face of global population growth (Bilotto et al. 2024). The rising demand for high quality protein and the need to reduce environmental impact are great challenges for the meat industry, necessitating research and innovative strategies to meet this global demand. One promising approach is fetal programming through maternal nutrition during gestation, aimed at optimizing skeletal muscle development, a key factor in meat quality and production efficiency (Du et al. 2010; Gionbelli et al. 2018; Costa et al. 2019, 2021). Skeletal muscle development begins in the intrauterine stage and is significantly influenced by maternal nutrition (Yan et al. 2013; Santos et al. 2022). The availability of nutrients during gestation affects myogenesis, process of muscle fiber formation, which is essential for determining the quantity and quality of meat produced. Nutrient deficiencies during critical gestation periods can interfere in muscle fiber formation and growth, underscoring the importance of optimizing maternal nutrition management strategies (Santos et al. 2022; Nascimento et al. 2024). Additionally, the maternal diet also affects adipogenesis, impacting intramuscular fat deposition and marbling, which are crucial factors in meat quality and the fetal development (Martins et al. 2015; Dean et al. 2024). Studies in various livestock species have shown that nutrient restriction can alter muscle fiber number and transcriptional regulation of myogenic genes in offspring (Diniz et al. 2021; Gundersen and Anas 2025; Zhao et al. 2025). Costa et al. (2021, 2022) analyzed the same dataset used in this study to investigate gene expression (e.g., coding protein) at both the transcriptomic and proteomic levels. Their findings identified genes associated with satellite cells, impaired insulin sensitivity, and alterations in intramuscular fat composition and oxidative metabolism. In their proteomic analysis, Costa et al. (2022) further reported that feed restriction during early gestation affects energy metabolism in the skeletal muscle of offspring by modifying glycolytic pathways and influencing the use of glycogen as a carbon source. Although these findings contribute to a better understanding of the molecular basis underlying fetal programming in response to maternal nutrition, the regulatory mechanisms remain incompletely understood. Therefore, the present study aims to further explore the underlying gene regulatory networks, with a focus on identifying regulatory elements such as lncRNAs, contributing to a more precise understanding of the genetic architecture of muscle development modulated by nutritional restriction. Long non‐coding RNAs (lncRNAs) have emerged as significant regulatory molecules in various biological processes, including skeletal muscle development (Muniz et al. 2022; Han et al. 2022). Unlike protein‐coding RNAs, lncRNAs exceed 200 nucleotides in length and do not translate into proteins (Mattick et al. 2023). This type of transcript has been found functioning through interactions with DNA, RNA, and proteins to modulate gene expression at transcriptional, post‐transcriptional levels, and epigenetic modification (Iwakiri et al. 2017). Recent studies have demonstrated that lncRNAs play crucial roles in muscle differentiation, growth, and regeneration by influencing myogenic regulatory factors (MRFs) and signaling pathways (Zhou et al. 2015; Muniz et al. 2022; Lv et al. 2022; Li et al. 2023; Chen et al. 2024). Additionally, lncRNAs has been found to be tissue‐specific, species‐specific, and condition‐specific (Park and Kim 2023), suggesting that lncRNAs are potential biomarkers and target regions for genetic interventions aimed at improving meat traits and muscle development. In goats, some studies have identified lncRNAs related to skeletal muscle growth and development, cell proliferation, transcriptional regulation, embryonic morphogenesis, intramuscular fat deposition, and meat tenderness (Zhan et al. 2016; Ling et al. 2019; Xu et al. 2021; Han et al. 2022; Shen et al. 2023; Tang et al. 2024). In addition, these studies have shown that lncRNA expression is less conserved than mRNA in skeletal muscle, with low conservation of lncRNAs across different species, such as cattle, mice, and humans, despite the relatively higher similarity observed between goat and human lncRNAs (Ling et al. 2019). These findings suggest the need to identify regulatory elements such as lncRNAs in a species‐specific manner and support the use of goats as a valuable animal model for human studies. Understanding the role of lncRNAs in muscle tissue provides insights into the molecular mechanisms underlying muscle development and may contribute to build the way for innovative strategies to improve livestock production and meat quality. The primary objective of this study was to perform an exploratory transcriptomic analysis to identify differentially expressed lncRNA between two maternal dietary treatments restriction‐maintenance (RM), in which pregnant dams were fed at 50% of their maintenance requirements from Days 8 to 84 of gestation, followed by 100% of the maintenance requirements from Day 85 to parturition; and maintenance‐restriction (MR), in which dams received 100% of maintenance requirements from Days 8 to 84, followed by 50% from Day 85 to parturition. Additionally, this study explored the impact of maternal feed restriction at different gestational stages on the profile of differentially expressed lncRNAs identified in the skeletal muscle of goat kids. By investigating how maternal nutrition influences lncRNA expression, this research seeks to contribute to a better understanding of the gene regulatory mechanisms underpinning fetal programming. Materials and Methods Animal Care All animal procedures were approved by the Ethical Committee on Animal Use of the Department of Animal Science at Federal University of Viçosa, Minas Gerais, Brazil (protocol number 09/2017). Database Information This study utilized phenotypic and RNA sequencing (RNA‐Seq) data generated by Costa et al. (2019) to investigate how feed restriction during different stages of gestation affects the gene expression profiles in the muscle tissue of offspring. The experimental management practices are fully detailed in Costa et al. (2019). Briefly, 60 nulliparous Saanen dairy goats underwent estrous synchronization using prostaglandin administered at a 7‐day interval, followed by artificial insemination with semen from a single male (Costa et al. 2019). The day of the insemination was considered Day 0 of the pregnancy and the beginning of the experiment. All the animals were kept under the same diet and management conditions for an adaptation period of 7 days. Pregnancy diagnosis was performed at Day 30 of gestation to exclude non‐pregnant goats and those carrying female fetuses. The remaining 14 pregnant goats were then kept to one of two feeding regimens: Restriction‐Maintenance (RM) group (n = 6): fed 50% of maintenance requirements from Days 8–84 of gestation, followed by 100% from Day 85 to parturition. Maintenance‐Restriction (MR) group (n = 8): fed 100% of maintenance requirements from Days 8–84, followed by 50% from Day 85 to parturition. Both groups received the same diet composition; only the timing of the restriction varied according to gestational stage. Full details on the dietary composition and management are available in Costa et al. (2019). At parturition, male kids were separated from the does prior to colostrum intake, weighed, and euthanized immediately after birth. In the case of twin births, the heavier male was selected. Despite the feeding regimen differences, there were no significant statistical differences in birth weight (p = 0.46) between kids born to RM does and those born to MR does, as reported by Costa et al. (2019). Samples of Longissimus thoracis muscle were collected from all newborn goats for total RNA extraction and sequencing. Total RNA was extracted using RNeasy Fibrous Tissue Mini Kit (Qiagen Inc., Germantown, MD, USA) following the manufacturer's recommendation. RNA quality and integrity were assessed using Agilent 2100 Bioanalyzer (Agilent Technologies Inc., Santa Clara, CA, USA). All samples had RNA integrity number (RIN) greater than 8. RNA‐Seq libraries were constructed using poly(A) selection and sequenced on the Illumina HiSeq 4000 platform using a strand‐specific protocol, generating 150 bp paired‐end reads with an average of approximately 23 million raw reads per sample. The data discussed in this publication have been deposited in NCBI's Gene Expression Omnibus and are accessible through GEO Series accession number GSE307250 (https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE307250). Data Analysis Workflow RNA‐Sequence Data Quality Control, Alignment, and Assembly of Reads The FastQC (v.0.11.4) software was used to assess the quality of RNA‐Seq reads focusing on parameters such as quality scores, N and GC contents, length distribution, overrepresented sequences, duplication levels, and K‐mer content (Andrews 2010). Based on Quality Control (QC) reports, reads trimming was performed using Trimmomatic (v0.39) (Bolger et al. 2014). The trimming parameters included removing adapter sequences, leading low‐quality or N bases with a quality score below 20, trailing low‐quality or N bases with a quality score below 15, and scanning the reads with a 4‐base wide sliding window to cut when the average quality per base dropped below 15. Additionally, reads were retained only if they had a Phred score above 30 and reads shorter than 50 bases were discarded. The trimmed reads were aligned to Goat genome reference ( Capra hircus . ARS1.112; https://ftp.ensembl.org/pub/current_fasta/capra_hircus/dna/) using the HISAT2 (v.2.0.5) software (Kim et al. 2015) with the parameters “‐dta” for the de novo discovery of splice sites and “‐known‐splicesite‐infile” provided with a list of known splice sites to facilitate aligning reads with small anchors. The aligned reads were subsequently assembled using Cufflinks (v2.2.1) (Trapnell et al. 2012) with the following parameters: ‐small‐anchor‐fraction 0.08, ‐min‐isoform‐fraction 0.20, ‐max‐intron‐length 300 000, ‐overlap‐radius 25, ‐max‐bundle‐frags 2 000 000, ‐trim‐3‐avgcov‐thresh 0.05, ‐junc‐alpha 0.001, and pre‐mrna‐fraction 0.15 (Salatta et al. 2025). RNA‐Sequence Quantification The individual transcript assemblies were merged into a master transcriptome using Cuffmerge to serve as a reference for quantifying and generating normalized expression counts in Fragments per Kilobase Million (FPKM) for each individual sample through the Cuffquant and Cuffnorm software's (Cuffmerge/Cuffquant/Cuffnorm pipeline; http://cole‐trapnell‐lab.github.io/cufflinks/manual/; Trapnell et al. 2012). Default settings were employed throughout the process (Salatta et al. 2025). A normalized counts matrix was constructed, and to ensure the reliability of downstream analyses, transcripts with no detectable expression (FPKM value = “0” in all samples) and with lengths shorter than 200 bp were excluded (Li et al. 2019). Differential Expression Transcript Analysis The LncDIFF is a package within R environment, particularly designed for differential expression analysis to detect low‐abundance non‐coding RNA transcripts such as lncRNAs (Li et al. 2019; https://CRAN.R‐project.org/package=lncDIFF). A generalized linear model with a zero‐inflated exponential quasi‐likelihood was used to estimate the effects of maternal nutrition groups on normalized counts, while accounting for overdispersion and zero inflation, followed by a likelihood ratio test to identify differentially expressed genes. Samples from the RM group (n = 6) were compared with the MR group (n = 8) using a generalized linear model with a log‐link function. No batch effects or other technical sources of variation (e.g., lane or technical replicate effects) were found that would justify inclusion as covariates in the model. Transcripts were classified as differentially expressed based on a log2 fold change (|log2 FC| > 2) and a false discovery rate (FDR) < 0.05. Identification and Classification of Non‐Coding Long RNAs The Feelnc pipeline (v.0.2.1) was used to identify potential lncRNAs from a total of 60 344 transcripts mapped. Firstly, the Feelnc‐filter pipeline was used to filter out transcripts with a length shorter than 200 bp, biotype‐coding protein, single‐exon transcripts, and biexonic transcripts having one exon size shorter than 25 bp were discarded (Etebari et al. 2015; Wucher et al. 2017; Muniz et al. 2022). Secondly, the Feelnc‐codpot was used to compute the coding potential score for each candidate transcript and classifying them into putative lncRNAs and protein‐coding RNAs. Lastly, the FEElnc‐classifier pipeline was used for possible function prediction of differentially expressed lncRNAs based on their nearest‐neighbor protein‐coding genes. The classification of lncRNAs was done based on type of interactions, which can be Genic or Intergenic (lincRNA). It will be genic when the DE lncRNA overlaps an RNA gene annotated in the Goat genome reference ( Capra hircus ARS1.112) and intergenic otherwise. The genic lnRNA can be further classified in subcategories according to the orientation of the interactions and localization of the transcription of proximal RNA transcripts; more details can be found in the FEElnc (https://github.com/tderrien/FEELnc; Wucher et al. 2017). To assess potential associations between differentially expressed lncRNAs and their predicted co‐located target transcripts, co‐expression analyses were performed using Spearman's rank correlation coefficient. Correlations were calculated between normalized expression values of each DE lncRNA and its corresponding positional target genes across all samples (n = 14). Spearman's method was selected due to its robustness to non‐normal data distributions and its suitability for RNA‐seq expression data (Looney and Hagan 2011). Statistical significance of the correlations was evaluated using the “cor.test” function in R, and p‐values were reported for each lncRNA–mRNA pair. LncRNA Annotation and Functional Enrichment Analysis The genomic positions of DE lncRNAs and their interacting genes were used to search Single Nucleotide Polymorphism (SNP) located within these regions, extending 5 kb upstream and downstream of each gene. SNPs were retrieved and annotated using the Goat Short Variants dataset from the Ensembl Variation Release 115 database (https://ftp.ensembl.org/pub/release‐115/variation/vcf/capra_hircus/). Variant extraction and annotation were performed using the BioMart tool available through Ensembl (https://www.ensembl.org/biomart). The Variants Effect Predictor (VEP) analysis was performed on a set of 147 SNPs located within or surrounding genomic regions containing DE lncRNAs and their interacting genes, with the aim of characterizing the predicted functional consequences of these variants. Additionally, Gene Ontology (GO) enrichment analysis was conjointly performed to identify biological processes, molecular functions, and cellular components related to genetic variants of genes associated with the DE lncRNAs. The GALLO (Genomic Annotation in Livestock positional candidate Loci; v.1.0) R package was utilized to annotate quantitative trait loci (QTLs) located within candidate regions or in the nearest surrounding regions (Fonseca et al. 2020). This annotation followed a genome‐wide approach and explicitly accounted for the limited number of QTLs currently available for goats (total = 3115) in the Goat QTLdb (https://www.animalgenome.org/cgi‐bin/QTLdb/CH/summary). Results Table 1 shows the three differentially expressed novel long non‐coding RNAs (lncRNAs) identified in the skeletal muscle of offspring exposed to maternal feed restriction during different gestational stages. Among these, lncRNA_10996.1, located on chromosome 17, exhibited a significant upregulation in the RM group compared to the MR group, with a Fold Change (FC) of 4.43 (FDR = 2.10E‐05). In the other hand, the lncRNA_11801.1 (FDR = 3.97E‐02) on chromosome 18 and lncRNA_7992.1 (FDR = 1.16E‐02) on chromosome 11 were downregulated in the RM group, with FCs of −4.35 and −5.54, respectively. These findings indicate that maternal nutritional intervention during gestation is associated with alterations in lncRNA expression profiles in offspring skeletal muscle, suggesting potential involvement of these lncRNAs in pathways related to muscle development and metabolic adaptation. TABLE 1Differentially expressed novel non‐coding RNA (lncRNA) in the skeletal muscle of offspring subjected to maternal feed restriction at different gestational stages.lncRNA code a Position b Length (bp) c FC (Log2)MR (FPKM)RM (FPKM)FDRlncRNA_10996.117:15792398–15 793 5619444.431.9541.952.10E‐05lncRNA_11801.118:50396895–50 402 211935−4.3554.402.683.97E‐02lncRNA_7992.111:97723600–97 727 898621−5.54179.993.871.16E‐02Abbreviations: FC, fold change (log2); FDR, false discovery rate (p‐value correction); FPKM, fragments per kilobase million; MR, treatment maintenance‐restriction; RM, treatment restriction‐maintenance. a Given name to novel lncRNA identified in this study. b Chromosome position for differentially expressed (DE) lncRNAs. c DE lncRNA transcript lengths in base pair. Differentially expressed novel non‐coding RNA (lncRNA) in the skeletal muscle of offspring subjected to maternal feed restriction at different gestational stages. Abbreviations: FC, fold change (log2); FDR, false discovery rate (p‐value correction); FPKM, fragments per kilobase million; MR, treatment maintenance‐restriction; RM, treatment restriction‐maintenance. Given name to novel lncRNA identified in this study. Chromosome position for differentially expressed (DE) lncRNAs. DE lncRNA transcript lengths in base pair. The classification of novel long non‐coding RNAs (lncRNAs) differentially expressed in the skeletal muscle of ruminant offspring revealed several predicted interactions with nearby protein‐coding genes based on their genomic context relative to annotated genes (Table 2). The lncRNA_10996.1 exhibited antisense interactions with genes ENSCHIG00000020055 and ENSCHIG00000024132, indicating both intergenic and genic relationships. Among these, the strongest predicted co‐localization interaction was identified with transcript ENSCHIT00000029747, an mRNA isoform of the ENSCHIG00000024132 gene, which encodes a GTP‐binding protein involved in chromatin condensation and cell cycle control. This association was also supported by a moderate negative correlation (r s = −0.38; Table S3). However, this relationship was not statistically significant (p = 0.18) and should therefore be interpreted with caution. The observed trend may reflect the limited sample size used in this study (n = 14). TABLE 2Classification of novel non‐coding RNA (lncRNA) differentially expressed in the skeletal muscle of ruminant offspring and their associated genes.lncRNA code a Partner gene b Partner transcript c Direction d TypeDistance e SubtypeLocationlncRNA_10996.1ENSCHIG00000020055ENSCHIT00000029747AntisenseIntergenic7322ConvergentDownstreamENSCHIG00000024132*ENSCHIT00000036607*AntisenseGenic0ContainingExoniclncRNA_11801.1ZNF546ENSCHIT00000016686AntisenseIntergenic762DivergentUpstreamPSMC4*ENSCHIT00000023798*AntisenseIntergenic497ConvergentDownstreamlncRNA_7992.1SNORA65ENSCHIT00000008359SenseIntergenic1695Same strandDownstreamZNF79*ENSCHIT00000023597*AntisenseGenic0OverlappingExonicLRSAM1ENSCHIT00000016616AntisenseIntergenic5522DivergentUpstreamENSCHIT00000016626AntisenseIntergenic7378DivergentUpstreamENSCHIT00000016608AntisenseIntergenic5522DivergentUpstreamENSCHIG00000023017ENSCHIT00000034693SenseIntergenic994Same strandDownstream a Given name to novel lncRNA identified in this study. b Gene Partner = “*” means that it is the DE novel lncRNA best RNA interaction, which is defined by if intergenic lncRNA (lincRNA), the best RNA partner is the closest to the lincRNA, and if genic lncRNA, the best RNA partner is by rule of priority exonic (then the fraction of exonic overlap) > intronic > containing. c Transcripts of the partner genes associated with DE lncRNA. d Transcription genome direction. e Distance between the DE lncRNA and their gene interaction. Classification of novel non‐coding RNA (lncRNA) differentially expressed in the skeletal muscle of ruminant offspring and their associated genes. Given name to novel lncRNA identified in this study. Gene Partner = “*” means that it is the DE novel lncRNA best RNA interaction, which is defined by if intergenic lncRNA (lincRNA), the best RNA partner is the closest to the lincRNA, and if genic lncRNA, the best RNA partner is by rule of priority exonic (then the fraction of exonic overlap) > intronic > containing. Transcripts of the partner genes associated with DE lncRNA. Transcription genome direction. Distance between the DE lncRNA and their gene interaction. The lncRNA_11801.1 was predicted to interact with mRNA isoforms of ZNF546 and PSMC4 genes through antisense associations located 762 bp upstream and 497 bp downstream, respectively. The strongest predicted co‐localization interaction for this lncRNA was observed with transcript ENSCHIT00000023798 (PSMC4‐201). However, Spearman correlation analysis indicated that the strongest expression‐based association was observed with the ENSCHIT00000016686 (ZNF546‐201) transcript (r s = 0.24, p‐value = 0.41; Table S3). This result reflects a weak positive correlation that was not statistically significant, suggesting no strong evidence of coordinated expression between these transcripts. It is important to note that the lack of statistical significance may be influenced by the limited sample size in this study (n = 14), which reduces statistical power and may hinder the detection of subtle biological relationships. Therefore, these findings should be interpreted with caution. Future studies involving larger sample sizes and complementary experimental approaches, such as gene knockdown or overexpression assays, are warranted to further investigate the potential functional relationship between lncRNA_11801.1 and its predicted target genes. Additionally, lncRNA_7992.1 demonstrated multiple predicted interactions based on genomic location, including both sense and antisense relationships with SNORA65, a snoRNA transcript, as well as with protein‐coding genes, ZNF79 and LRSAM1. The strongest predicted co‐localization interaction for lncRNA_7992.1 was identified with transcript ENSCHIT00000023597 (ZNF79‐201). However, co‐expression analysis revealed a weak and non‐significant positive correlation (r s = 0.20, p = 0.50; Table S3), indicating no clear evidence of coordinated expression between these transcripts. It is also important to highlight that several predicted target transcripts (ENSCHIT00000008359, ENSCHIT00000016616, ENSCHIT00000016626, and ENSCHIT00000034693) were not expressed in goat skeletal muscle under the experimental conditions evaluated in this study, impeding correlation analysis. This lack of detectable expression limits the ability to assess potential co‐expression relationships involving lncRNA_7992.1 and its predicted co‐localized targets and suggests that these interactions may be tissue‐specific or condition‐dependent. Given these limitations, including the relatively small sample size and the lack of detectable expression of predicted target transcripts, the observed associations should be interpreted with caution. Although, lncRNA_7992.1 may potentially be involved in regulatory processes, there is insufficient evidence to support a role in gene suppression or activation based on the current data. Interestingly, among the evaluated co‐expression interactions, lncRNA_10996.1 (located on chromosome 17) showed a moderate negative correlation with transcript PSMC4‐201 (r s = −0.46; p = 0.09; located on chromosome 18) and a significant positive correlation with LRSAM1‐202 (r s = 0.65; p = 0.01; located on chromosome 11). Although these transcripts were not associated with lncRNA_10996.1 through co‐localization analysis, PSMC4‐201 and LRSAM1‐202 were co‐localized with lncRNA_11801.1 and lncRNA_7992.1, respectively. Overall, these findings suggest that differentially expressed lncRNAs may participate in diverse regulatory mechanisms within skeletal muscle, with many predicted interactions occurring in the antisense orientation relative to neighboring transcripts (Table 2). However, as these associations are primarily based on genomic proximity, co‐localization predictions, and co‐expression trends, they should be considered exploratory. Therefore, further functional validation, such as gain or loss of function experiments, transcript knockdown or overexpression assays, and integrative multi‐omics analyses is required to confirm their biological relevance. A total of 147 genetic variants were identified in genomic regions associated with differentially expressed lncRNAs in the skeletal muscle of ruminants (Table S1). These genetic variants were located on chromosomes 11, 17, and 18, with the highest concentration observed on chromosome 11. Among the identified genetic variants, chromosome 11 harbored 76 single nucleotide polymorphisms (SNPs) spanning positions 97 723 693–97 727 870 bp, while chromosome 18 contained 68 SNPs distributed across positions 50 396 982–50 402 180 bp. Chromosome 17 exhibited a lower number of variants, with only three SNPs detected within a short region between 15 793 474 and 15 793 554 bp. The Variant Effect Predictor (VEP) analysis was performed to characterize the 76 genetic variants (Table S1) identified within a genomic region on chromosome 11 harboring differentially expressed (DE) lncRNAs detected in offspring muscle. A total of 416 variant consequences associated with these SNPs were identified and are presented in Figure 1A.1. Among these consequences, a subset (1.44%) affected coding protein regions, as highlighted in Figure 1A.2. Additionally, analysis of SNPs located within targeted regions on chromosomes 17 and 18 (Figure 2A) revealed 114 variant consequences. Detailed information regarding the SNPs identified within the genomic regions, containing DE lncRNAs can be found in Table S2—Spreadsheets 1 and 2. Variant Effect Predictor analysis of genomic regions on Chromosome 11 harboring differentially expressed long non‐coding RNAs in ruminant muscle. (A.1) Summary of genetic variant consequences identified in differentially expressed lncRNA‐associated regions on Chromosome 11, based on the Capra hircus ARS1.112 reference genome. (A.2) Genetic variants affecting coding regions within differentially expressed lncRNA‐associated regions on Chromosome 11. Genetic variants and quantitative trait loci in genomic regions harboring differentially expressed lncRNAs in ruminant muscle. (A) Variant Effect Predictor analysis of genomic regions on Chromosomes 17 and 18 associated with differentially expressed lncRNAs, based on the Capra hircus ARS1.112 reference genome. (B) Summary of QTLs identified in genomic regions on Chromosomes 11, 17, and 18 harboring differentially expressed lncRNAs in ruminant muscle. A total of five genome windows containing the lncRNAs, lncRNA_10996.1 and lncRNA_11801.1, as well as three of their interacting genes, showed statistically significant association (p < 0.05) with two QTLs (Figure 2B and Table S2—Spreadsheet 3). Notably, one QTL identified on chromosome 17 was associated with lncRNA_10996.1 and its co‐located interacting gene ENSCHIT00000036607. This QTL was classified as an exterior QTL, specifically related to feet and leg conformation traits. On chromosome 18, one QTL was identified within the genomic regions containing lncRNA_11801.1 and its co‐localization interacting genes ENSCHIT00000023798 and ENSCHIT00000016686. This QTL was categorized as a production QTL associated with body weight traits. Although these findings suggest a potential relationship between DE lncRNAs, their interacting genes, and economically relevant traits, these associations should be interpreted as exploratory and hypothesis‐generating, as they do not establish causality. Therefore, further functional and validation studies are necessary to confirm the biological relevance of these genomic regions and their potential regulatory roles. Discussion Maternal nutrition plays a crucial role in fetal skeletal muscle development by influencing molecular and cellular processes that regulate muscle growth and differentiation (Yan et al. 2013; Costa et al. 2022; Nascimento et al. 2024; Dean et al. 2024; Sousa et al. 2024). Studies have shown that maternal feed restriction during gestation can alter muscle development, which may lead to change in patterns of fiber formation and muscle structure in offsprings (Costa et al. 2019, 2021). In this context, the present study investigated the effects of maternal feed restriction at different gestational stages on long non‐coding RNA (lncRNA) expression in goat skeletal muscle. Our results indicate that nutritional restriction during early (RM group) and late gestation (MR group) is associated with changes in lncRNA expression profile, suggesting that maternal nutrition may contribute to transcriptional regulatory processes involved in fetal muscle development. Long non‐coding RNAs (lncRNA) have emerged as regulators of skeletal muscle development, functioning through various mechanisms, including chromatin remodeling, transcriptional regulation, and post‐transcriptional modifications (Han and Chang 2015; Sweta et al. 2019; Mattick et al. 2023). In the present study, three differentially expressed lncRNAs were identified (|FC| > 2|and FDR ≤ 0.05) when comparing offspring from does fed at 50% of maintenance requirements during the first trimester of gestation, followed by 100% of maintenance requirements until parturition (RM), with offspring from does fed at 100% of maintenance requirements during the first trimester, then restricted to 50% of maintenance requirements until parturition (MR). These findings suggest that gestational stage‐specific nutritional restriction may be associated with differential transcriptional responses in fetal muscle tissue, potentially involving lncRNA associated regulatory mechanisms. Functional annotation of predicted lncRNA target genes suggested potential associations with biological processes involved in fatty acid metabolism, transcriptional regulation, proteolysis, and protein synthesis. For instance, the upregulated lncRNA_10996.1 was associated with genes ENSCHIG00000020055 and ENSCHIG00000024132, which encode Acyl‐CoA dehydrogenase family member 10 (ACAD10) and a GTP‐binding nuclear protein, respectively. They genes have been associated with biological processes related to fatty acid metabolism, chromatin condensation, and cell cycle regulation. The ACAD10 gene is an enzyme that catalyze the first step of mitochondrial β‐oxidation of fatty acids (Rashan et al. 2025). While direct studies on ACAD10 expression in ruminant muscle are limited, it is been linked to metabolic disturbances such as impaired glucose tolerance, insulin resistance, and abnormal weight gain in human and mice (Bian et al. 2010; Bloom et al. 2018, 2020). Costa et al. (2021), using the same animal set used in our study, identified genes that indicate impaired insulin sensitivity and changes in the composition of intramuscular fat. Additionally, a proteomic investigation showed that proteins that catalyzes reactions of β‐oxidation (ACAT1 and ECHS1) were upregulated in the skeletal muscle of kids from the RM group, corroborating to our finding of ACAD10 and suggesting enhancement in fatty acid degradation and acetyl‐CoA accumulation (Costa et al. 2022). Therefore, we hypothesize that lncRNA_10996.1 may positively regulate ACAD10 expression, which should be further investigated through functional interaction assays. This hypothesis is supported by previous studies demonstrating that lncRNAs often exhibit tissue‐specific and condition‐specific expression patterns, highlighting their potential roles as regulators and potential biomarkers of muscle energy metabolism (Wang et al. 2019; Statello et al. 2021; Park and Kim 2023). The downregulated lncRNAs (lncRNA_11801.1 and lncRNA_7992.1) are linked to zinc finger protein encoders such as ZNF79 (Zinc finger protein 79) and ZNF546 (Zinc finger protein 546), as well as PSMC4 (proteasome 26S subunit, ATPase 4), LRSAM1 (leucine rich repeat and sterile alpha motif containing 1), and the ENSCHIG00000023017 gene, which protein matches for RPL12 (Large ribosomal subunit protein uL11). These genes are involved in transcriptional and translational regulation, blastocyst development, proteolysis, and DNA‐binding transcription factor activity and RNA polymerase II‐specific functions. These associations suggest that lncRNAs may contribute to regulatory pathways associated with metabolic and structural responses to maternal nutritional stress. However, it is important to emphasize that these interactions are inferred based on genomic proximity predictions, and additional functional interaction assays are required to validate these potential relationships. In addition, a notable finding of this study is the co‐expression pattern observed for lncRNA_10996.1. Although this lncRNA is located on chromosome 17, it exhibited correlations with transcripts located on different chromosomes, including PSMC4‐201 (chromosome 18) and LRSAM1‐202 (chromosome 11), suggesting potential trans‐regulatory associations. A significant positive correlation was observed with LRSAM1‐202 (Spearman's ρ = 0.65, p = 0.01), indicating coordinated expression between these transcripts, while a moderate negative correlation was detected with PSMC4‐201 (ρ = −0.46, p = 0.09). These results suggest that lncRNA_10996.1 may participate in regulatory networks extending beyond its local genomic context, however, these relationships are based on co‐expression analysis and should be interpreted as putative, requiring further functional validation to confirm potential trans‐acting mechanisms. In contrast, predicted interactions involving lncRNA_11801.1 and lncRNA_7992.1 showed weak and non‐significant correlations, providing limited support for coordinated expression with their predicted targets. Additionally, several predicted target transcripts were not expressed in skeletal muscle under the experimental conditions evaluated, precluding correlation analysis. This lack of detectable expression suggests that some interactions may be tissue‐specific or condition‐dependent. Along with the relatively small sample size (n = 14), these factors reduce statistical power and limit the ability to detect subtle biological relationships. Consequently, the observed associations should be interpreted with caution and warrant further investigation in future studies of these interactions. Among the genes associated with DE downregulated lncRNAs, PSMC4 (also known as Rpt3) is an essential subunit of the 26S proteasome, required for the degradation of most proteasomal substrates. Its absence is lethal in mice before implantation due to defects in blastocyst development (Sakao et al. 2000). Additionally, muscle‐specific deletion of Rpt3 impaired muscle growth and reduced force production. Preliminary results from Safain et al. (2024) indicated that fetal muscle of feed‐restricted heifers during early gestation affected the expression of RPL12, a crucial gene involved in translation initiation and protein metabolism. Moreover, the findings highlight the tissue‐specific metabolic adaptations in response to maternal nutrient restriction, emphasizing the complex balance between fetal energy demands and mitochondrial function during early development (Safain et al. 2024). Gionbelli et al. (2018) investigated the effects of maternal nutrition on fetal skeletal muscle development and growth potential in cattle. They observed gene expression changes in the skeletal muscle of offspring from overnourished cows, including increased Zinc Finger protein expression at mid‐gestation. In our study, we identified two Zinc Finger protein‐encoding genes (ZNF79 and ZNF546) interacting with differentially expressed lncRNAs, suggesting that maternal feed restriction, particularly during early gestation, may be associated with Zinc Finger gene expression and related regulatory elements, which could contribute to shaping the biological mechanisms involved in fetal muscle development. Overall, this indicates that gestational stage‐specific nutritional stress may trigger distinct transcriptional responses in fetal muscle tissue, potentially involving lncRNA related regulatory mechanisms. The integration of genomic variant analysis further identified SNPs within regions harboring differentially expressed lncRNAs, including variants with potential functional impact on genes involved in protein synthesis and metabolism. Within a genomic window on chromosome 11 (97723600–977 278 981), ten SNPs were identified, including high‐impact variants such as a frameshift mutation (rs689852198) and a missense variant (rs650562014), both affecting the RPL12 gene. This gene encodes the large ribosomal subunit protein uL11 and was associated with the downregulated lncRNA_7992.1, which may suggest a link between these variants and altered protein synthesis or translational control mechanisms. Frameshift variants, known to disrupt the translational reading frame (Clark and Pazdernik 2013), can lead to loss of function or trigger nonsense‐mediated decay, while missense variants may impact protein efficiency (Petrosino et al. 2021). In the context of this study, where goats were subjected to feed restriction at different stages of gestation, the impact of these variants on gene expression in offspring muscle tissue becomes particularly relevant. Once, nutritional stress during gestation can influence epigenetic regulation and transcriptomic patterns in developing muscle, potentially modulating how these SNPs affect gene function, potentially impacting muscle growth or function in the offspring. Although no statistically significant QTLs were identified in this region, the presence of potentially high‐impact SNPs may indicate a regulatory role not yet captured in current QTL databases, especially considering the limited annotation for goats. This gap highlights the need for integrating expression data to understand the biological relevance of such variants. The SNPs identified within windows on chromosomes 17 (15792398–15 793 561) and 18 (50396895–50 402 211), regions containing lncRNA_10996.1 and lncRNA_11801.1, were classified as modifier variants, often associated with non‐coding regions or lncRNAs, where their functional impact is less direct but still potentially significant in gene regulation. Interestingly, QTLs linked to feet and leg conformation (chromosome 17) and body weight (chromosome 18) were also identified in these regions, both of which are traits influenced by muscle development. All the SNPs identified within these windows on chromosomes 17 and 18 were classified as “modifier” variant consequences, which are variants usually located in non‐coding regions or affecting non‐coding genes (e.g., lncRNAs), where predictions are difficult or there is no evidence of impact. Therefore, given the limited QTL genomic annotation available for goats and the complexity of gene regulation, these associations should be regarded as exploratory and may not be indicative of direct causative relationships. The integration of genetic variant analysis and QTL mapping annotation highlights genomic regions that may be involved in regulating muscle development and performance traits in ruminants. If validated, these findings could contribute to a better understanding of the developmental programming of muscle‐related traits in response to maternal nutrition during gestation. Overall, the findings of this study suggest that differentially expressed lncRNAs may participate in complex regulatory networks underlying skeletal muscle development and metabolic adaptation in response to maternal nutrition. However, it is important to emphasize that the identified interactions are largely based on genomic proximity, co‐localization predictions, and co‐expression trends. As such, they should be considered exploratory and hypothesis generating. Future studies incorporating larger sample sizes, additional tissues, integrative multi‐omics approaches, and functional interaction assays will be essential to validate these findings. Functional experiments, including gain‐ and loss‐of‐function assays, transcript perturbation, and molecular interaction analyses, are required to confirm the biological roles of these lncRNAs and to elucidate their mechanisms of action. Therefore, this study provides a foundation for future research into the regulatory roles of lncRNAs in fetal programming, while highlighting the need for experimental validation and cautious interpretation of computational predictions. Conclusions Our study provides novel insights into the potential involvement of lncRNAs in the regulation of skeletal muscle development in response to maternal restriction during gestation. The differential expression patterns observed, together with predicted co‐localization and co‐expression relationships, suggest that lncRNAs may contribute to complex regulatory networks underlying fetal programming, including potential trans‐acting mechanisms. However, these findings should be interpreted as exploratory, as the interactions identified in this study are based on genomic proximity predictions and correlation analyses, and are further limited by the relatively small sample size and lack of detectable expression for some predicted targets. Importantly, this work establishes a valuable framework for future investigations, highlighting candidate lncRNAs and genomic regions for further study. Subsequent research integrating larger datasets and functional validation approaches will be essential to confirm these regulatory relationships and to better understand the role of lncRNAs in shaping muscle development and improving livestock production outcomes. Funding This research was supported by funding from CNPq—National Council for Scientific and Technological Development (Grant #175251/2023‐1; 152401/2025‐3; 174939/2023‐0; 313858/2021‐7), CAPES (#001), and NSERC—Natural Sciences and Engineering Research Council (Grant #401862). Ethics Statement All animal procedures were approved by the Ethical Committee on Animal Use of the Department of Animal Science at Federal University of Viçosa, Minas Gerais, Brazil (protocol number 09/2017). Conflicts of Interest The authors declare no conflicts of interest. Supporting information Table S1: Genetic variants identified in genomic regions harboring differentially expressed long noncoding RNAs in muscle of ruminants.