Effects of synthetic breast milk on the gut metagenome and whole blood transcriptome in lambs

Abstract

Early postnatal nutrition is crucial for the growth and development of lambs, and artificial milk formulas are widely used as alternatives to breast milk in intensive sheep production. However, the molecular and microbial mechanisms underlying the differences between breast milk and formula feeding remain unclear. This study aimed to compare the fecal metagenomic and whole blood transcriptomic profiles of lambs fed breast milk (BF group) and commercial formula (FF group) from 4 to 45 days of age, to provide a theoretical basis for optimizing formula compositions. A total of 6 lambs were randomly divided into two groups (n = 3 per group), with body weight and body dimensions measured at 45 days of age, followed by fecal metagenomic sequencing and whole blood transcriptomic sequencing. The results showed that BF lambs had significantly higher body weight, body length, heart girth, and chest width than FF lambs. Metagenomic analysis revealed that at the phylum level, Bacteroidetes was enriched in FF lambs, whereas Firmicutes predominated in BF lambs. Differential abundance was also observed at the genus level (higher Desulfovibrio in FF lambs) and the pathway level, with BF lambs enriched in quorum sensing and FF lambs showing higher abundances of pathways related to ubiquinone and other terpenoid-quinone biosynthesis. Moreover, transcriptomic analysis identified 3290 differentially expressed genes (DEGs) between the two groups, with DEGs mainly enriched in metabolic pathways, mTOR signaling pathway, osteoclast differentiation, B cell receptor signaling pathway and MAPK signaling pathway. Collectively, compared with FF, BF enhanced lamb growth, optimized gut microbiome structure and modulated blood transcriptomic profiles related to metabolism, signaling and immunity. These findings highlight the key microbial taxa and functional pathways modulated by breastfeeding, providing valuable insights for the development of more effective milk formula alternatives.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12917-026-05460-5.

Introduction

Early postnatal nutrition is a critical determinant of growth performance and physiological development in lambs, with breastfeeding (BF) long recognized as the optimal nutritional strategy for neonatal ruminants due to its balanced composition of nutrients, bioactive factors, and immune components [1]. However, in intensive livestock production, factors such as maternal milk scarcity and and multiple births often necessitate the use of artificial milk formulas as alternatives [2, 3]. Despite continuous optimization, commercial lamb formulas frequently fail to replicate the growth-promoting and health-beneficial effects of breast milk, leading to differences in growth performance, gut microbial homeostasis, and immune function between formula-fed (FF) and BF lambs [4, 5].

The gut microbiome plays a pivotal role in nutrient metabolism [6], immune development [7], and host-microbe crosstalk [8], and its composition is highly sensitive to early nutritional interventions. Extensive research in human infants has demonstrated that BF shapes a gut microbial community enriched in beneficial taxa such as Bifidobacterium and Bacteroidetes, which in turn contribute to efficient nutrient utilization and intestinal barrier function [9]. In contrast, FF in this population often leads to alterations in microbial diversity and abundance, potentially increasing the risk of metabolic and immune disorders [10]. While these infant-focused findings have laid a foundational understanding of the link between early feeding strategies and gut microbiome dynamics, translating and validating such insights in livestock species—particularly in lambs, a key agricultural animal with distinct digestive physiology—remains an important knowledge gap. Specifically, the detailed differences in fecal metagenomic profiles—including taxonomic composition, functional gene abundance, and metabolic pathway potential—between BF and FF lambs remain incompletely characterized, limiting the targeted optimization of formula compositions for this species.

Parallel to microbial changes, early nutrition also modulates host transcriptional regulation, with blood transcriptome profiles reflecting systemic physiological responses to dietary cues [11, 12]. Transcriptional differences in genes related to metabolism, immunity, and growth signaling pathways may underpin the phenotypic disparities between BF and FF lambs. Integrating fecal metagenomic and blood transcriptomic analyses can provide a comprehensive understanding of the molecular and microbial mechanisms linking feeding regimens to growth outcomes, yet such multi-omics studies in neonatal lambs are scarce.

Given the practical need for high-quality artificial feeding alternatives in sheep production, this study aimed to systematically compare the fecal metagenomic and whole blood transcriptomic differences between BF and FF lambs from 4 to 45 days of age. We hypothesized that breastfeeding would induce distinct microbial and transcriptional signatures associated with superior growth performance, and that identifying these signatures would reveal key functional pathways and microbial taxa to inform formula optimization. The results of this study are expected to provide a theoretical basis for developing more effective milk formulas that mimic the nutritional and functional benefits of breast milk, ultimately improving the growth and health of artificially fed lambs.

Materials and methods

Experimental animals and feeding regimens

A total of 6 healthy 4-day-old lambs were randomly divided into two groups with 3 lambs each: the breastfeeding (BF) group and the formula feeding (FF) group. Lambs in the BF group were allowed ad libitum nursing from their maternal dams, while those in the FF group were fed a commercial lamb formula (Table S1, the formula is primarily composed of whole milk powder, multiple vitamins, and trace elements) following the same feeding frequency (4 times daily) and duration (30 min per feeding) as the BF group to ensure consistent nutritional intake conditions. The entire feeding trial was conducted from 4 days to 45 days of age, with lambs housed in individual pens under controlled environmental conditions (temperature 18–22 °C, humidity 50–60%) and provided with free access to clean drinking water throughout the experiment.

Body measurement assessment

At 45 days of age, all lambs were fasted for 12 h prior to body measurement to eliminate the influence of feed in the digestive tract. Body weight was determined using an electronic scale. Body dimension parameters, including height at withers (vertical distance from the highest point of the withers to the ground), body length (horizontal distance from the occipital bone to the ischial tuberosity), heart girth (circumference of the chest measured at the posterior margin of the scapula, taken at the deepest part of the chest), chest width (linear distance between the left and right shoulder joints, measured at the widest point of the chest), and cannon bone circumference (circumference of the middle of the metacarpal bone), were measured using a flexible measuring tape and vernier calipers according to standard livestock breeding measurement protocols. Each measurement was repeated three times, and the average value was recorded for statistical analysis.

Sample collection

Fresh fecal samples (approximately 5 g per lamb) were collected immediately after defecation in the morning before feeding using sterile disposable gloves and placed into sterile cryogenic vials, which were snap-frozen in liquid nitrogen within 5 min of collection to preserve microbial activity and stored at -80 °C until metagenomic sequencing. For blood sample collection, lambs were restrained gently, and 5 mL of whole blood was collected from the jugular vein using EDTA-anticoagulant vacuum blood collection tubes. For the BF and FF groups, the whole blood samples from three lambs in each group were pooled together, respectively. Blood samples were placed on ice immediately after collection and transported to the laboratory within 2 h. Total RNA was extracted promptly to avoid RNA degradation, ensuring the quality of transcriptomic sequencing data.

Fecal metagenomic sequencing and analysis

Total genomic DNA was extracted from fecal samples using the QIAamp Fast DNA Stool Mini Kit (Qiagen, Hilden, Germany) following the manufacturer’s instructions, including steps of mechanical lysis, chemical digestion, and column purification to obtain high-purity DNA. DNA quality and concentration were assessed using a NanoDrop 2000 spectrophotometer (Thermo Fisher Scientific, USA) and 1% agarose gel electrophoresis. Genomic DNA libraries were constructed by fragmenting DNA into 350 bp fragments, end-repairing, adding A-tails, and ligating sequencing adapters, followed by purification and PCR amplification. Libraries were sequenced on an Illumina NovaSeq 6000 platform (Illumina, USA) to generate 2 × 150 bp paired-end reads. Raw sequencing data were quality-filtered using Trimmomatic (v0.39) to remove adapter sequences, low-quality reads (Q-value < 20), and short reads (< 50 bp) to obtain clean reads. Clean reads were assembled into contigs using Megahit (v1.2.9), and open reading frames (ORFs) were predicted using MetaGeneMark (v3.38). Redundant sequences were removed using CD-HIT (v4.8.1) to construct a non-redundant gene catalog. Taxonomic annotation was performed by aligning the gene catalog to the Silva (v138) and NCBI NR databases using BLASTP (E-value < 1e-5). Functional annotation was conducted using the COG (Cluster of Orthologous Groups) and KEGG (Kyoto Encyclopedia of Genes and Genomes) databases. Alpha diversity indices (Shannon, Simpson) were calculated using QIIME2 (v2021.4), and beta diversity was analyzed by principal coordinate analysis (PCoA) based on Bray-Curtis distances. Differential abundance analysis of taxa and functional pathways between groups was performed using LEfSe (LDA score > 2, P < 0.05) and DESeq2.

Whole blood transcriptomic sequencing and analysis

Total RNA was extracted from whole blood samples using the RNAiso Plus Kit (TaKaRa, Dalian, China) following the manufacturer’s protocol, with additional DNase I digestion to remove genomic DNA contamination. RNA quality was evaluated using a NanoDrop 2000 spectrophotometer (A260/A280 ratio 1.8-2.0) and Agilent 2100 Bioanalyzer (RNA Integrity Number > 7.0). RNA libraries were constructed using the TruSeq Stranded Total RNA Library Prep Kit (Illumina, San Diego, USA) by rRNA depletion, fragmentation, first-strand cDNA synthesis, second-strand cDNA synthesis, end-repair, adapter ligation, and PCR amplification. Libraries were sequenced on an Illumina HiSeq X Ten platform (Illumina, USA) to generate 2 × 150 bp paired-end reads. Raw reads were filtered using Fastp (v0.20.1) to remove adapters, low-quality bases, and short reads. Clean reads were aligned to the ovine reference genome (Oar_v4.0) using Hisat2 (v2.2.1), and gene expression levels were quantified as Fragments Per Kilobase of transcript per Million mapped reads (FPKM) using StringTie (v2.1.6). Differentially expressed genes (DEGs) analysis between the BF and FF groups was conducted using DEGseq (with TMM normalization, Poisson distribution for p-value calculation, and BH procedure for FDR adjustment), and differentially expressed genes were screened by the special criteria of |log2(FoldChange)| > 1 and qvalue < 0.005. To further assess the robustness of DEGs identified by DEGseq, an additional analysis was performed using GFOLD (Generalized Fold Change), a Bayesian model-based method that estimates posterior fold change distributions and is suitable for transcriptomic data without biological replicates. The input count tables were prepared following GFOLD’s required format, and genes with |GFOLD| > 1 were considered differentially expressed. Gene Ontology (GO) enrichment analysis and KEGG pathway enrichment analysis of DEGs were performed using the clusterProfiler R package (v4.2.2), with a corrected P < 0.05 considered statistically significant. Hierarchical clustering of DEGs was conducted using the pheatmap R package, and visualization of results was performed using ggplot2.

Statistical analysis

All phenotypic data (body weight and body dimensions) were expressed as mean ± standard error of the mean (SEM) and analyzed using Student’s t-test in SPSS 26.0 software (IBM, USA). For metagenomic data, statistical significance was determined using one-way analysis of variance (ANOVA) followed by Benjamini-Hochberg correction to control for false discovery rate. A P < 0.05 was considered statistically significant, and P < 0.001 was considered highly statistically significant. All experiments included three biological replicates, and statistical graphs were generated using R software (v4.1.2) and GraphPad Prism 8.0.

Results

Body weight and dimensions differences between BF and FF Lambs

Lambs under breastfeeding and formula feeding were subjected to phenotypic assessment, fecal metagenomics, and whole blood transcriptomics to investigate the impacts of feeding regimens (Fig. 1A).Body weight and dimensions, which are core indicators of growth performance, were significantly affected by the feeding regimen. As shown in (Fig. 1B), BF lambs achieved a mean body weight of 13.27 ± 0.64 kg at 45 days of age, whereas FF lambs had a mean body weight of 6.47 ± 0.36 kg, and statistical analysis confirmed that the body weight of BF lambs was significantly higher than that of FF lambs with a highly significant difference (P < 0.001). To evaluate skeletal and somatic development, six key body dimension parameters were measured. No significant difference was observed in height at withers between the two groups (BF: 40.37 ± 1.23 cm; FF: 40.13 ± 2.67 cm, P > 0.05, Fig. 1C) and cannon bone circumference (BF: 6.07 ± 0.09 cm; FF: 5.70 ± 0.42 cm, P > 0.05, Fig. 1G). In contrast, BF lambs exhibited a significantly longer body length (45.00 ± 0.78 cm vs. 35.80 ± 2.36 cm, P < 0.05, Fig. 1D), a greater heart girth (55.67 ± 1.25 cm vs. 45.33 ± 1.25 cm, P < 0.001, Fig. 1E), and a wider chest width (15.17 ± 1.13 cm vs. 11.17 ± 0.21 cm, P < 0.05, Fig. 1F) compared to FF lambs.

Fig. 1.

Experimental design and growth performance of lambs fed breast milk or formula. A Experimental design and workflow of fecal metagenome and whole blood transcriptome analysis. B–G Bar plots depicting body weight (B), height at withers (C), body length (D), heart girth (E), chest width (F), and cannon bone circumference (G) of BF and FF lambs. ns: not significant, *P < 0.05, **P < 0.001

Sequencing data quality and gene abundance profiles

The fecal metagenome sequencing generated a large dataset for microbial community and functional analysis. Quality control of the fecal metagenomic sequencing data is summarized in Table S2-S3. A total of six samples (BF_1, BF_2, BF_3, FF_1, FF_2, FF_3) were sequenced. After removal of low-quality reads and host sequences, clean bases ranged from 4.55 Gb to 10.48 Gb across samples. The GC content ranged from 45.5% to 48.1%, and the Q20 and Q30 scores were consistently above 95.8% and 88.9%, respectively. The gene abundance distribution across samples is displayed, where BF and FF groups showed distinct patterns in the abundance ranges of genes (Fig. 2A). The sequence length distribution and gene statistics are summarized. A total of 1,512,382 unique genes were identified, with a total sequence length of 1,148,313,615 bp and an average gene length of 759.3 bp. Sequences were categorized into multiple length intervals, with the highest number of sequences observed in the SL-500 interval, indicating the prevalence of short genes in the fecal metagenome (Fig. 2B).

Fig. 2.

Fecal metagenome sequencing quality and microbial community similarity. A Bar plot of gene abundance distribution across BF and FF samples. B Circular plot summarizing sequence length distribution and gene statistic. C Heatmap displaying sample similarity based on gene abundance. D Venn diagram illustrating core and unique genes across samples. E–F PCoA plots illustrating microbial community structure at phylum (E) and genus (F) levels

For whole blood transcriptome sequencing, quality control data are presented in Table S3. The BF and FF samples generated 66.3 million and 74.2 million raw reads, corresponding to 9.94 Gb and 11.13 Gb of raw bases, respectively. After quality filtering, 65.3 million (BF) and 69.5 million (FF) clean reads were obtained, with clean bases of 9.79 Gb and 10.42 Gb, respectively. The error rate was 0.03% for BF and 0.02% for FF. Q20 and Q30 scores exceeded 97.9% and 94.2%, respectively, and GC content was 56.1% for BF and 54.9% for FF. Reads alignment to the ovine reference genome (Oar_v4.0) is detailed in Table S4 and Table S5. The overall mapping rate was 92.79% for BF and 95.05% for FF, with uniquely mapped reads accounting for 89.16% and 92.03%, respectively. These results demonstrate the high quality and reliability of the transcriptome sequencing data.

Sample similarity and core gene analysis

A correlation heatmap showing the similarity between samples is presented. BF samples (BF_1, BF_2, BF_3) clustered closely together, and FF samples (FF_1, FF_2, FF_3) also formed a distinct cluster, suggesting strong intra-group similarity and significant inter-group differences in microbial gene composition (Fig. 2C). A Venn diagram of core and unique genes is shown. A total of 61,314 core genes were shared across all samples, representing the conserved microbial gene pool in lamb feces. Meanwhile, each group harbored a substantial number of unique genes: BF_1 had 397,930 unique genes, BF_2 had 648,978, BF_3 had 406,146, FF_1 had 605,302, FF_2 had 485,036, and FF_3 had 569,651(Fig. 2D).

Microbial community structure differentiation

To assess the overall microbial community structure, principal coordinate analysis (PCoA) was performed at the phylum and genus levels. At the phylum level, PCoA1 accounted for 73.07% of the variation, and PCoA2 accounted for 26.02%. Samples from BF and FF groups were clearly separated, indicating a significant difference in phylum-level community composition between the two feeding regimens (Fig. 2E). Similarly, at the genus level, PCoA1 explained 34.52% of the variation, and PCoA2 explained 30.34%, with BF and FF samples forming distinct clusters (Fig. 2F).

Taxonomic composition at phylum and genus levels

To characterize the microbial community composition, taxonomic analysis was performed at the phylum and genus levels. At the phylum level, both BF and FF groups were dominated by Firmicutes and Bacteroidetes, yet BF lambs had a higher relative abundance of Firmicutes while FF lambs showed a greater proportion of Bacteroidetes, with other phyla like Proteobacteria present at lower abundances (Fig. 3A). At the genus level, Bacteroides was the most abundant genus in both groups, but FF lambs had higher abundances of Prevotella and Clostridium whereas BF lambs showed increased levels of Ruminococcus and Bifidobacterium (Fig. 3B).

Fig. 3.

Taxonomic composition of fecal microbiota at phylum and genus levels. A Stacked bar plot showing phylum-level abundance in BF and FF groups. B Stacked bar plot showing genus-level abundance in BF and FF groups

Microbial community dissimilarity and differential taxa

The ANOSIM plot showed a dissimilarity rank of 0.222 (P = 0.1) between BF and FF groups, suggesting a trend of microbial community differentiation associated with feeding regimen (Fig. 4A). When analyzing differentially abundant taxa, at the phylum level, Bacteroidetes was more abundant in FF lambs and Firmicutes was more prevalent in BF lambs (Fig. 4B). At the genus level, genera such as Desulfovibrio were more abundant in FF lambs than in BF lambs (Fig. 4C).

Fig. 4.

Microbial community dissimilarity and differential taxa analysis. A ANOSIM plot showing rank of dissimilarity between BF, FF, and inter-group samples. B–C Dot plots showing differential phyla (B ) and genera (C) between BF and FF groups

Functional gene abundance and pathway analysis

Functional gene abundance based on COG classification showed that BF and FF groups display a consistent proportion in terms of gene distribution trends, and the top five categories in terms of gene abundance were ranked in the following order, translation, ribosomal structure and biogenesis; replication, recombination and repair, amino acid transport and metabolism, cell wall/membrane/envelope biogenesis and carbohydrate transport and metabolism (Fig. 5A). Further analysis of functional categories and pathways revealed that for COG categories, no statistically significant difference was observed between the BF and FF lambs (Fig. 5B), and at the pathway level, BF lambs were enriched in quorum sensing whereas FF lambs had higher abundances of pathways related to ubiquinone and other terpenoid-quinone biosynthesis (Fig. 5C).

Fig. 5.

Functional gene abundance and pathway enrichment of fecal metagenome. A Stacked bar plot showing COG functional category abundance in BF and FF groups. B–C Dot plots showing differential COG functional categories (B) and KEGG pathways (C) between BF and FF groups

Transcriptome sequencing quality and differential gene expression

The whole blood transcriptome sequencing data were analyzed to investigate transcriptional differences between BF and FF lambs. High-quality transcriptome data with satisfactory QC metrics and high genome mapping rates were obtained, as shown in Tables S4 and S5. The boxplot of FPKM values showed distinct expression profiles between the two groups, with BF lambs exhibiting a slightly lower median FPKM value compared to FF lambs (Fig. 6A). A volcano plot further revealed 3290 DEGs between FF and BF groups, with 1901 upregulated and 1389 downregulated genes in FF lambs (Fig. 6B, Table S6). GFOLD analysis identified 5216 DEGs between BF and FF groups, with 3574 upregulated and 1642 downregulated in FF lambs (Table S7). The results showed substantial overlap with the DEGseq-derived DEGs (Figure S1).

Fig. 6.

Whole blood transcriptome sequencing quality and differential gene expression. A Box plot showing FPKM distribution in BF and FF groups. B Volcano plot showing DEGs with |log₂(fold change)| ≥ 2 and adjusted P < 0.05. C Bar plot showing the most enriched GO terms of DEGs

Gene ontology (GO) enrichment analysis, hierarchical clustering and pathway enrichment

The most enriched GO terms for DEGs between FF and BF groups, categorized into biological process and cellular component, are presented in (Fig. 6C). In the biological process category, terms like metabolic process, cellular metabolic process, and primary metabolic process were significantly enriched, while in the cellular component category, terms such as cell, cell part, and intracellular were highly represented. The hierarchical clustering heatmap showed distinct expression patterns of DEGs between BF and FF groups, with genes like TENT5A, ARL11, GALNS, VWA2, NR2E1, C3AR1, CYP20A1, MID1IP1, FILIP1L, RSAD1, BEST4 and DCUN1D4 differentially expressed and clear clustering distinguishing the two feeding groups (Fig. 7A). The bubble plot of KEGG pathway enrichment indicated that DEGs were significantly enriched in pathways such as metabolic pathways, mTOR signaling pathway, osteoclast differentiation, B cell receptor signaling pathway, MAPK signaling pathway, other types of O-glycan biosynthesis, lysosome, inositol phosphate metabolism, thermogenesis, Fc epsilon RI signaling pathway, with the rich factor and q-value showing these pathways were functionally relevant to the transcriptional differences (Fig. 7B, Table S8).

Fig. 7.

Hierarchical clustering and pathway enrichment of DEGs. A Heatmap showing hierarchical clustering of DEGs in BF and FF groups. B Bubble plot showing KEGG pathway enrichment of DEGs

Discussion

The findings of the present study indicated that breastfeeding exerted a pronounced positive effect on the growth performance of lambs, as reflected by the significantly higher body weight, body length, heart girth, and chest width in breastfed lambs relative to those receiving formula feeding. This observation aligns with a broader body of evidence highlighting breastfeeding’s advantageous impacts on early-life development-though the specific outcomes may vary by species and trait. This aligns with previous findings in ruminants and non-ruminants, where breast milk’s balanced nutrient composition-including bioactive peptides, fatty acids, and growth factors-promotes efficient nutrient utilization and somatic growth [13, 14]. In contrast, commercial formulas often lack these bioactive components or have suboptimal nutrient ratios, leading to reduced growth efficiency [15]. The non-significant differences in height at withers and cannon bone circumference suggest that early feeding regimens primarily affect somatic and thoracic muscle development rather than appendicular skeletal growth. This may be attributed to breast milk’s enrichment of factors that regulate muscle protein synthesis, such as insulin-like growth factor 1 (IGF-1), which is less abundant in artificial formulas. Breastfed infants show a positive correlation between plasma IGF-1 levels and their weight and length growth rates [16].

The distinct gut microbial community structures between BF and FF lambs highlight the profound impact of early nutrition on gut microbiome assembly. BF lambs were enriched in Firmicutes. In contrast, FF lambs had a higher proportion of Bacteroidetes and Desulfovibrio. This significant taxonomic shift may be the core driver of the decreased growth performance in FF lambs. Firmicutes offer a key advantage through their functional strains, which efficiently degrade complex carbohydrates such as cellulose and starch [17]. These strains provide the host with a critical energy source by producing volatile fatty acids. Additionally, they contribute to the repair and maintenance of the intestinal mucosal barrier, a process essential for nutritional metabolism and intestinal health in young animals [18]. In contrast, the increased proportion of Bacteroidetes in FF lambs may reflect a shift toward a microbiota composition with distinct metabolic capabilities. It is important to note that Bacteroidetes also play beneficial roles in the gut, including the degradation of dietary fiber and production of short-chain fatty acids such as propionate and acetate, which contribute to energy harvest and intestinal homeostasis [18, 19]. However, in the context of the FF group, this shift was accompanied by an overgrowth of Desulfovibrio. This genus comprises sulfate-reducing bacteria that produce hydrogen sulfide, a metabolite with known cytotoxic effects on intestinal epithelial cells [20]. Consequently, this combination may disrupt the metabolic balance of the flora and reduce the efficiency of nutrient digestion and absorption.

Functional metagenomic analysis further revealed group-specific metabolic potentials, with BF lambs enriched in genes related to quorum sensing and FF lambs showing enrichment in ubiquinone and other terpenoid-quinone biosynthesis. This result is highly correlated with the previously described intestinal microbiota taxonomic differences and growth performance phenotypes, further elaborating the regulatory mechanism of microbiota structure on lamb growth and development from the perspective of functional metabolism. For BF lambs, the enrichment of Firmicutes accompanied by high-abundance quorum sensing -related genes has important physiological significance. Quorum sensing, a key mechanism for microbial communities to perceive cell density and regulate group behaviors, can enhance the efficiency of synergistic microbial metabolism [21]. Functional Firmicutes strains with the capacity to degrade cellulose and starch achieve metabolic synchronization via the quorum sensing pathway [22]. This process synergistically upregulates the synthesis and secretion of carbohydrate-degrading enzymes, enhances the decomposition efficiency of complex carbohydrates, and promotes the massive production of volatile fatty acids, thereby supplying ample energy to the host and forming a positive function [23, 24]. Meanwhile, quorum sensing maintains intestinal microbial homeostasis by regulating the colonization, competition and defense behaviors of microbiota [25]. This regulatory effect helps Firmicutes form a stable dominant structure and inhibits the invasion of harmful microbes. In turn, it indirectly preserves the integrity of the intestinal mucosal barrier, thereby providing a healthy microenvironment for lamb growth. Thus, the enrichment of quorum sensing -related genes is an important molecular basis for BF lambs to exert efficient nutritional metabolism and homeostasis maintenance functions, and a key support for their superior growth performance.

In contrast, the enrichment of ubiquinone and terpenoid-quinone biosynthesis genes in FF lambs is not a manifestation of optimized microbial functions, but a compensatory metabolic response to intestinal microenvironment stress. Ubiquinone (coenzyme Q) and terpenoid-quinones are key substances involved in oxidative phosphorylation and antioxidant stress, which participate in cellular energy synthesis and scavenge reactive oxygen species (ROS) to alleviate oxidative damage [26, 27]. The increased proportion of Bacteroidetes and excessive proliferation of Desulfovibrio in FF lambs lead to the production of hydrogen sulfide. This hydrogen sulfide damages intestinal epithelial cells and induces low-grade chronic inflammation, during which a large number of ROS are generated to cause oxidative damage to cells [28, 29]. It can be inferred that the enrichment of related biosynthesis genes serves as a compensatory mechanism initiated by intestinal microbiota and host cells, which accelerates the synthesis of antioxidant substances to resist oxidative stress and repair damaged cells.

The identified DEGs between BF and FF lambs reflect extensive transcriptional regulation by feeding patterns, and these transcriptional changes are closely associated with the aforementioned gut microbial functional shifts (presented as testable hypotheses due to sample pooling constraints). DEGs are mainly enriched in metabolic processes, cellular components and immune-related pathways, among which mTOR, MAPK signaling pathways and immune-related pathways (B cell receptor signaling pathway, osteoclast differentiation) are significantly enriched, highlighting the profound effect of feeding-induced microbiota changes on the physiological state of the host. The mTOR signaling pathway is a core regulator of cell growth, proliferation, and energy metabolism [30]. Its modulation by DEGs in BF lambs may promote cell growth and protein synthesis by integrating nutrient signals, thereby facilitating overall growth and development. The MAPK signaling pathway, a key mediator of stress and inflammatory responses [31], is likely activated in FF lambs in response to intestinal mucosal damage induced by hydrogen sulfide from Desulfovibrio. The enrichment of DEGs in this pathway may amplify the inflammatory response cascade, further exacerbating intestinal barrier impairment and growth retardation. Additionally, the significant enrichment of DEGs in two key biological pathways, namely the B cell receptor signaling pathway and osteoclast differentiation, implies that dietary feeding regimens exert a non-negligible regulatory effect on both the immune system and skeletal development of lambs. Specifically, the B cell receptor signaling pathway is a critical component of adaptive immunity in lambs. This pathway mediates B cell activation, proliferation, and antibody secretion, thereby mounting targeted immune responses against pathogenic invasions in the intestinal tract [32, 33]. Meanwhile, osteoclast differentiation is not only closely linked to bone remodeling and mineral metabolism but also plays a vital role in immune regulation by modulating the activity of immune cells such as T lymphocytes and macrophages [34, 35]. Collectively, these findings highlight the multifaceted impacts of nutritional strategies on the physiological homeostasis of lambs, bridging the gap between dietary interventions, immune competence, and skeletal health.

The enrichment of the B cell receptor signaling pathway in the transcriptomic analysis is particularly noteworthy, as this pathway plays a central role in B lymphocyte activation, proliferation, and differentiation into antibody-secreting plasma cells [36]. Among the antibodies produced, immunoglobulin A (IgA) is of special interest in the context of early nutrition, as it serves as the primary mediator of mucosal immunity in the gastrointestinal tract [37]. IgA contributes to intestinal homeostasis by neutralizing pathogens, modulating gut microbiota composition, and reinforcing the intestinal epithelial barrier. Although direct measurement of IgA levels was not feasible in this study due to the complete utilization of blood samples for transcriptome sequencing, the transcriptional enrichment of the B cell receptor signaling pathway in BF lambs suggests enhanced humoral immune responsiveness. This observation aligns with the enrichment of beneficial gut microbiota, such as Bifidobacterium, in BF lambs—a genus known to promote IgA production through interactions with gut-associated lymphoid tissue [38]. Collectively, these transcriptomic signatures provide a molecular basis for the hypothesis that breastfeeding may enhance mucosal immune function in lambs. Future studies incorporating direct measurements of immune parameters, including IgA levels, are warranted to validate these transcriptional findings and to elucidate the functional implications of the observed pathway enrichments.

Conclusions

This study found that BF enhanced lambs’ growth performance, shaped a more beneficial gut microbiome (Firmicutes) and Quorum sensing function, and modulated whole blood transcriptomic profiles related to metabolism, signal transduction and immunity compared to FF. Our findings demonstrate that breastfeeding not only enhances growth performance but also shapes a more beneficial gut microbiome and modulates transcriptional profiles related to metabolism, signal transduction and immunity in lambs. These results provide valuable insights into the molecular and microbial mechanisms underlying the growth-promoting effects of breastfeeding in early postnatal lambs, which could inform strategies to optimize artificial feeding regimens for neonatal ruminants.

Nevertheless, certain limitations in this study are acknowledged, including the relatively small sample size and the pooling of transcriptomic samples within each group, which meant that biological replicates at the individual level were not available and may have influenced the statistical robustness of the differentially expressed genes identified.

Authors' contributions

Zhenliang Zhang: Writing–original draft, Data curation. Jiachen Bai: Writing–review & editing. Yucheng Liu: Writing–Data curation, Software. Jingjing Wang: Writing–Data curation, Software. Zhanmin Lv: Resources, Methodology. Ling Tang: Validation. Ruigang Wang: Resources. Lei Gao: Review & editing, Changbin Liu: Writing–review & editing, Shouliang Lu: Review & editing, Xiangwei Fu: Writing–review & editing, Visualization. Jianhong Ni: Writing–review & editing, Visualization. Pengcheng Wan: Writing–review & editing, Visualization.

Data availability

This raw data of this article will be available by National Genomics Data Center with identifier (Project ID: PRJCA060776).

Declarations

Ethics approval and consent to participate

Sample collection was carried out by the animal owners with their consent. This study was conducted in accordance with ethical guidelines and regulations and was approved by the Ethical Committee of the Xinjiang Academy of Agricultural and Reclamation Sciences (Approval Letter No. 2024-001).

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.