ABSTRACT This study provided a comprehensive exploration of the nutritional regulation of volatile fatty acids (VFAs) on hindgut microbial metabolism and epithelial homeostasis in dairy goats. Twenty-four goats were orally administered sodium acetate (SA) at 0.8 g/kg of body weight (BW), propionate (SP) at 0.8 g/kg of BW, butyrate (SB) at 0.5 g/kg of BW, or saline (CON) before morning feeding for 12 days (n = 6/group). Serum and hindgut epithelial tissues were collected to measure antioxidant capacity, inflammatory cytokines, and tight junctions. Cecal contents were collected for 16S rRNA sequencing and metabolome analysis, and colonic epithelial cells were harvested for transcriptome sequencing. The data demonstrated that VFAs positively affected hindgut homeostasis. SB reduced serum malondialdehyde levels (P = 0.042), while SA and SP increased intestinal interleukin-10 concentration compared with CON (P < 0.001). All three VFAs enhanced gut barrier functions by increasing tight junctions compared with CON (P < 0.05). The data revealed distinct bacterial abundances and diversities associated with VFA administration, with notable responders including Rikenellaceae dgA-11 gut group, Christensenellaceae R-7 group, and Bacteroides. Metabolome analysis indicated significant changes in metabolic processes, such as purine, arachidonic acid, tyrosine, and tryptophan metabolism. Transcriptome analysis showed that SA and SP influenced endocrine and digestive functions, metal ion homeostasis, and muscle development in the colonic epithelium, with specific immune response pathways enriched in SB. Correlation analysis suggested interactions between hindgut bacteria and derived metabolites and epithelial homeostasis. In short, our study suggested potential strategies for improving gut health and overall well-being in goats through dietary interventions. IMPORTANCE The volatile fatty acids (VFAs), mainly produced by rumen microbiota, play an important role in ruminal metabolic functions and epithelial health, but their impact on the hindgut has received limited attention. Our study highlighted the significant role of VFAs in hindgut bacterial metabolism and homeostasis, providing novel insights into the role of VFAs in regulating hindgut metabolism and physiological homeostasis beyond the rumen. KEYWORDS: volatile fatty acids, dairy goats, hindgut, gut homeostasis, gut bacteria INTRODUCTION In the modern livestock industry, there is an increasing emphasis on animal welfare and healthy production to meet the demand for high-quality animal products. Ruminant animals, including cattle, sheep, and goats, play important roles in food production and provide 16% of global protein and 8% of global energy consumption through milk and meat ([36]1). Goat milk, characterized by its high digestibility due to concentrations of monounsaturated and polyunsaturated fatty acids, as well as minerals like zinc, iron, and magnesium ([37]2, [38]3), offers superior qualities such as alkalinity, probiotic carrier ability, and buffering capacity compared to cow milk ([39]4). Despite these benefits, goats contribute only 2.3% of global milk production, a fraction compared to 82.6% from cattle ([40]1, [41]5). Improving gastrointestinal tract (GIT) health in ruminants is crucial for enhancing milk yield and quality for human consumption. The rumen and hindgut, key components of the GIT, serve as primary sites for nutrient digestion and absorption by GIT microorganisms ([42]6). Moreover, the epithelial lining of the GIT acts as a critical barrier, preventing harmful substances from entering the portal circulation ([43]7, [44]8). While recent studies have extensively investigated the regulation of ruminal bacteria, their unique digestive functions, mechanisms of nutrient absorption through the ruminal epithelium, and their impacts on ruminal health and homeostasis ([45]9, [46]10), the regulation of the hindgut has received less attention despite its essential role. The hindgut is mainly for digesting starch and other fermentable substrates that are not fully degraded in the rumen ([47]11, [48]12), providing approximately 10% of dietary energy through fermentation by the cecal microbiota ([49]13). Significant differences in colonization patterns of the cecal microbiome compared to the rumen microbiome have been observed ([50]8, [51]14). Additionally, the cecal microbiota influences gut immunity, inflammation, and intestinal barrier function through host-microbe interactions, which is crucial for health and efficient production ([52]15). For instance, the hindgut microbiome affects host oxidative stress by influencing glutathione synthesis in postpartum dairy cows ([53]16). Unlike the multilayered rumen epithelium, although the hindgut epithelium has only a single layer of epithelial cells, it modulates inflammation and establishes a barrier against pathogens through tight junctions ([54]17, [55]18). Volatile fatty acids (VFAs), primarily acetate, propionate, and butyrate, are essential products of rumen microbial fermentation ([56]10). They serve as important energy sources, providing over 70% of the energy utilized by ruminants ([57]19). Studies in the rumen have demonstrated that supplementing VFAs, particularly butyrate, enhances ruminal fermentation, stimulates the development of rumen papillae and epithelial cells ([58]20, [59]21), and exhibits anti-inflammatory, antitumorigenic, and antimicrobial effects ([60]22). Notably, the hindgut contributes approximately 12% of the VFAs available to the organism ([61]23). However, the specific effects of VFAs on the cecal bacteria profiles, their metabolic processes, and their interaction with homeostasis and molecular functions in the gut epithelium remain unclear. Our hypothesis was that different VFAs can regulate the succession of cecal bacteria, influence their metabolic processes, and engage in crosstalk with epithelial cells to promote hindgut development and maintain homeostasis in goats. To investigate this hypothesis, we employed a dairy goat model and administered the animals with sodium acetate (SA), propionate (SP), and butyrate (SB), with saline serving as the control group. Through this approach, we aimed to generate comprehensive data encompassing VFA concentrations of cecal contents, serum biochemistry and antioxidant capacity parameters, and epithelial inflammatory cytokines and tight junctions. Additionally, we analyzed the bacterial and metabolome profiles of cecal contents and performed transcriptome sequencing of the colonic epithelial tissue. MATERIALS AND METHODS Animal management and experimental design This animal experiment was conducted using female Guanzhong dairy goats at the experimental farm of Yangzhou University in Jiangsu Province, China, from November to December 2019. A total of 24 dairy goats with a mean body weight (BW) of 47.44 ± 3.38 kg at 1.5 years old in the early-lactation period were selected for this study. All goats were maintained under the same farming conditions and were fed a standard diet containing 60% forage and 40% concentrate mix ad libitum for 14 days at 1.50 kg to ensure stable feeding. Goats were individually housed in separate pens and fed twice daily at 8:00 and 18:00, with ad libitum access to feed and water. The feed ingredients and chemical composition are shown in [62]Table 1 on a dry matter (DM) basis. The feed ingredients were composed of 41% oat hay, 29.5% corn grain, 14.5% soybean meal, 10% wheat bran, and 5% limestone powder, CaH[2]PO[4], NaCl, and premix. The content of nutrient components was composed of 15.56% crude protein (CP), 35.26% neutral detergent fiber (NDF), 19.63% acid detergent fiber (ADF), 2.89% ether extract (EE), 0.41% Ca, and 0.47% P. The diet was formulated to meet the current feeding recommendations with a digestible energy (DE) of 10.25 MJ/kg. During the adaptation period, goats were manually milked before each feeding time in their tie stalls twice a week, and the milk yield was recorded. Milk samples were analyzed for composition and quality at Yangda Kangyuan Dairy Co., Ltd. (Yangzhou, Jiangsu, China). The average milk yield of the selected goats was 1.45 ± 0.20 kg/day, with average milk fat and protein compositions of 3.28% ± 0.30% and 3.05% ± 0.21%, respectively. All goats were vaccinated and had no history of antimicrobial agent (antibiotics, antifungals, or antivirals) administration or clinical signs of diseases. TABLE 1. Feed ingredients and chemical composition (DM basis, %) Items Diet Ingredients (% of DM)  Oat hay 41.00  Corn grain 29.50  Soybean meal 14.50  Wheat bran 10.00  Limestone powder 0.35  CaH[2]PO[4] 0.15  NaCl 0.50  Premix[63]^a 4.00 Chemical composition (% of DM)  CP 15.56  NDF 35.26  ADF 19.63  EE 2.89  Ca 0.41  P 0.47 DE (MJ/kg) 10.25 [64]Open in a new tab ^^a The premix provided the following per kg of diet: VA, 200,000 IU; VD3, 70,000 IU; VE, 350 IU; Fe, 1.6 g; Cu, 1.7 g; Zn, 8.2 g; Mn, 2.5 g; and Se, 40 mg. After adaptation, goats were randomly assigned to four groups (n = 6/group) to receive oral administration treatments. The BW and milk production were balanced among the groups, with no significant differences in these parameters. Goats were orally administered SA at 0.8 g/kg of BW per day (Sigma-Aldrich, USA, >99.0% purity), SP at 0.8 g/kg of BW per day (Sigma-Aldrich, >99.0% purity), SB at 0.5 g/kg of BW per day (Sigma-Aldrich, >98.0% purity), or normal saline (CON) at 1.0 L/day as the control group ([65]Fig. 1A). The dosage of VFA administration solution was determined based on the optimal concentration selected according to our previous research and related studies in ruminants ([66]24[67]–[68]27). Briefly, each goat was weighed, and the required amount of each VFA was calculated before daily morning feeding. The sodium powders were dissolved in 1.0 L of normal saline, and all solutions were calibrated to a consistent pH using a pH meter. Solutions were then orally administered through acid-resistant tubing with adjustable syringes 1 h before morning feeding, within 10 min per goat. The experiment lasted for 13 days, consisting of a 12-day administration period and a 1-day sampling period, which ensured that the animal management and dietary formulation were consistent with those in the adaptation period. Fig 1. [69]Experimental design of goat groups treated with different short-chain fatty acids. Beta diversity plot indicates distinct microbial profiles across groups. Bar graphs display shifts in gut microbiota composition at phylum, family, and genus levels. [70]Open in a new tab The schematic representation of experimental design, and the bacterial beta diversity and abundance in cecal contents. (A) The schematic representation of experimental design; (B) bacterial beta diversity using principal coordinate analysis (PCoA) plot calculated by unweighted UniFrac distances; (C) component proportions of cecal bacteria at phylum, family, and genus levels of four groups; n = 6 for 16S rRNA sequencing. Analyses performed on feed samples were as follows: DM, according to AOAC official method 930.15 by drying them at 105°C for 2 h, and the ash content was determined by combustion at 550°C for 4 h (AOAC official method 942.05); CP, according to AOAC official method 990.02 by multiplying the nitrogen concentration by 6.25 using a Kjeltec Auto Analyzer (K9860, Jinan Hanon Instruments Co., Ltd, Jinan, Shandong, China); EE, according to AOAC official method 920.39 using a Soxtec Auto Analyzer (SOX500, Jinan Hanon Instruments Co., Ltd, Jinan, Shandong, China). The contents of NDF and ADF were determined according to the Van Soest method ([71]28). The contents of Ca and P were measured according to AOAC official methods 968.08 and 965.17. DE was calculated from tabulated feed values according to NRC. Sample collection All goats were fasted for 12 h on sampling day, and then the BW was recorded. Blood samples were collected from the coccygeal vein using procoagulant tubes and centrifuged at 3,000 × g for 15 min at 4°C to collect serum. Samples were then stored at −80°C for serum biochemistry and antioxidant capacity measurements. Subsequently, goats were anesthetized and euthanized by jugular vein puncture at the slaughterhouse of the experimental farm. The middle section of the cecum tissue was scraped using a microscope slide, and cecal contents were immediately collected and stored at −80°C for 16S rRNA sequencing and metabolome analysis. The entire colon, cecum, and rectal tissues were carefully separated, emptied of digesta, rinsed in phosphate-buffered saline (PBS), and weighed. Another 5.0 g of colonic and cecal epithelial tissues from the middle sections was harvested, rapidly frozen in liquid nitrogen, and stored at −80°C for measurements of inflammatory cytokines, tight junctions, and transcriptome sequencing analysis. Concentration of VFAs of cecal contents The concentration of VFAs was determined using a gas chromatograph (GC-9A; Shimadzu, Kyoto, Japan). Briefly, 0.5 g of cecal content was diluted with 1.0 mL of distilled water. Then, 0.3 mL of metaphosphoric acid containing 20% of 60.0 mmol/L crotonic acid was added, vortexed, centrifuged, and filtered through a 0.25-μm-pore-size syringe filter. The supernatant was then collected for subsequent analyses. Portions (0.5 µL) of test samples and standard solution mix were run through a CP-WAX capillary column (length, 30.0 m; inner diameter, 0.53 mm; and film thickness, 1.0 mm) in a gas chromatograph. Program settings and calculations followed the method of our laboratory ([72]27). Serum biochemistry and antioxidant capacity parameters Serum concentrations of lactate dehydrogenase (LDH; BC0685), creatinine (CREA; BC4915), blood urea nitrogen (BUN; BC1535), triglyceride (TG; BC0625), cholesterol (CHOL; BC1985), high-density lipoprotein cholesterol (HDL-C; BC5325), and low-density lipoprotein cholesterol (LDL-C; BC5335) were measured using commercial kits (Beijing Solarbio Science & Technology Co., Ltd., Beijing, China) following the manufacturer’s protocols. For serum antioxidant capacity parameters, the total antioxidant capacity (T-AOC; BC1315), glutathione peroxidase (GSH-Px; BC1195), catalase (CAT; BC0205), superoxide dismutase (SOD; BC0170), and malondialdehyde (MDA; BC0025) were measured using commercial assay kits from Solarbio according to the manufacturer’s instructions. Concentration of inflammatory cytokines and tight junctions of hindgut epithelial tissue To measure concentrations of total protein (TP), inflammatory cytokines, and tight junctions, 0.1 g of colonic and cecal epithelial tissue was homogenized with 1 mL PBS and then centrifuged at 13,400 × g at 4°C for 10 min to collect the supernatant. TP concentration was measured using a commercial assay kit (A045-2, Nanjing Jiancheng Bioengineering Institute, Nanjing, China). Concentrations of interleukin-1β (IL-1β), interleukin-6 (IL-6), interleukin-10 (IL-10), tumor necrosis factor α (TNF-α), Claudin1, occludin, and tight junction protein 1 (ZO-1) were measured using enzyme-linked immunosorbent assay (ELISA) with a multifunctional microplate reader (SpectraMax M5, Molecular Devices, Sunnyvale, CA, USA) following the manufacturer’s instructions (MLBio, Shanghai, China). Parameters were normalized by TP to determine concentrations per milligram protein. 16S rRNA sequencing and data processing of cecal contents Total microbial DNA from cecal contents was extracted using FastPure Bacteria DNA Isolation Mini Kit (DC103, Vazyme Biotech Co., Ltd., Nanjing, China) for 16S rRNA sequencing. DNA quality was assessed by agarose gel electrophoresis and quantified using a UV spectrophotometer. High-throughput sequencing of the V3-V4 region of the 16S rRNA gene was performed using an Illumina NovaSeq PE250 (Illumina, San Diego, CA, USA) at LC-Bio Technology Co., Ltd. (Hangzhou, Zhejiang, China). The V3-V4 region was amplified with primers 341F (5′-CCTACGGGNGGCWGCAG-3′) and 805R (5′-GACTACHVGGGTATCTAATCC-3′) ([73]29). For data processing, paired-end reads were merged using FLASH software (version 1.2.7) ([74]30). Quality filtering was then performed under specific filtering conditions to obtain the high-quality clean reads using Fqtrim software (version 0.9.4). Chimeric sequences were filtered using VSEARCH software (version 2.3.4) ([75]31). The feature table and feature sequence were obtained after dereplication using DADA2 software (version 1.10.1) ([76]32). Bacterial alpha diversity (Chao 1, Shannon, and Simpson) was used to assess the complexity of species diversity in QIIME2 software (version 2019.4) ([77]33). Beta diversity was calculated using principal coordinate analysis (PCoA) based on Unweighted UniFrac distance ([78]34). The amplicon sequence variant (ASV) profiling analysis and microbial relative abundance analysis were performed using R software (version 4.2). Given that the microbial alpha diversity parameters and the relative abundances of bacterial phyla and genera were not normally distributed, the Wilcoxon rank-sum test was used to analyze the differences between the treatments. The significant taxonomic differences in response to VFA administration at genus classification were also selected using linear discriminant analysis effect size (LEfSe) analysis (LDA score > 3) ([79]35). LC-MS metabolome and data processing of cecal contents An accurately weighed 0.1 g sample of cecal contents was used for LC-MS metabolome analysis. Metabolome analysis was performed on a Vanquish UHPLC System (ThermoFisher Scientific, Santa Clara, CA, USA) using an ACQUITY UPLC HSS T3 column (150 mm × 2.1 mm, 1.8 µm) (Waters, Milford, MA, USA) at a column temperature of 40°C, a flow rate of 0.25 mL/min, and an injection volume of 2 µL. The LC-ESI (+)-MS and LC-ESI (−)-MS analyses followed a previously described protocol ([80]36). Then, mass spectrometric detection of metabolites was also performed on a Q Exactive HF-X (Thermo Fisher Scientific, USA) with an ESI ion source. The parameters were set up following a previous study ([81]37). Metabolome analysis was performed at Panomix Biomedical Tech Co., Ltd. (Suzhou, Jiangsu, China). For data processing, the raw data were converted to mzXML format by MSConvert using ProteoWizard software (version 3.0.8789). Data were then processed using XCMS for feature detection, retention time correction, and alignment ([82]38, [83]39). The metabolites were identified by accurate mass (<30 ppm) and MS/MS data, which were matched using HMDB, MassBank, LipidMaps, mzCloud, and KEGG databases. Then, normalized data were imported into SIMCA software (version 14.1) (AB Umetrics, Umea, Sweden) and preprocessed using PAR scaling and mean centering before principal components analysis (PCA) ([84]40). The orthogonal partial least squares discrimination analysis (OPLS-DA) was performed to allow the determination of discriminating metabolites using the variable importance on projection (VIP) value. The OPLS-DA model was tested for overfitting with 200 permutation tests. Then, nonparametric tests were performed on non-normally distributed metabolomic data using the Wilcoxon-Mann-Whitney test to calculate the P-value. The P-value, VIP, and fold change (FC) were applied to discover the contributable variable for classification. Finally, P-value < 0.05 and VIP > 1 were considered to be statistically significant metabolites. The functional enrichment analysis was performed using the KEGG database at the MetaboAnalyst 5.0 website ([85]41). Transcriptome sequencing and data processing of colonic epithelial tissue Briefly, total RNA of colonic epithelial tissue sample of each goat was isolated and purified using FastPure Cell/Tissue Total RNA Isolation Kit V2 (RC112, Vazyme Biotech Co., Ltd., Nanjing, China) for transcriptome sequencing analysis. The RNA integrity was assessed by Agilent 2100 with RIN number > 7.0. The cDNA library was constructed and then sequenced on the DNBSEQ-T7 sequencer platform (BGI, Shenzhen, China) at Novogene Co., Ltd. (Beijing, China). For data processing, the Fastp software (version 0.23.1) ([86]42) was used to perform quality control on the raw data and obtain clean data. Then, the HISAT2 software (version 2.1.0) ([87]43) was used to align the obtained clean data to the latest reference genome of goats (Capra hircus, ARS1.2) ([88]44). The Samtools software (version 1.10) was used to sort and convert the SAM files to BAM format ([89]45), while the Stringtie software (version 2.2.1) was used to assemble and quantify the genes based on read counts ([90]43). The PCA was then generated in R software (version 4.2). Finally, the identification of differentially expressed genes (DEGs) was estimated using DESeq2 software (version 1.36.0) based on the normalized gene count data. DEGs were identified using a threshold of P < 0.05 and |log2 fold change| > 1. Then, a Venn plot, a volcano plot, and a dot plot of all DEGs were generated. Finally, the functional enrichment analysis was performed using KEGG and GO databases. Statistical analysis and visualization The Statistical Package for Social Sciences (SPSS 25.0, SPSS, Inc., Chicago, IL, USA) software was used for the statistical analyses except for the multi-omics data. The data were fitted into a general linear model for a completely randomized design. The different VFAs and saline treatments were considered as the main factor, while the animals were considered as a random factor. The normality of data distribution was verified by the Kolmogorov-Smirnov test. The comparison of the differences of the normally distributed data, such as BW, hindgut weight, VFA concentrations, serum biochemistry and antioxidant capacity parameters, inflammatory cytokines, and tight junctions, was subjected to the one-way analysis of variance (one-way ANOVA) and post hoc Duncan’s test. Significant and extremely significant differences were declared at P < 0.05 and P < 0.01, respectively. Correlation analysis between selected parameters and multi-omics data was calculated using Pearson’s or Spearman correlation coefficients and Mantel’s test. R package ggplot2, Pheatmap, ggcor, and GraphPad Prism 6.0 software (GraphPad, California, USA) were used for graphics. RESULTS Body weight, hindgut weight, and the concentration of VFAs in cecal contents The BW, hindgut weight, and concentration of VFAs in cecal contents are presented in [91]Table 2. Oral administration of the three VFAs and saline did not impact the BW of all goats (P = 0.089) but significantly affected hindgut weight. Specifically, SA and SB significantly increased cecal weight compared with CON and SP (P < 0.001). Additionally, SA increased rectal weight compared with the other three groups (P < 0.001). However, VFA administration did not affect colonic weight (P = 0.125). Regarding the concentration of VFAs in cecal contents, SA significantly increased the concentration of total VFAs (P = 0.017), acetate (P = 0.005), and other VFAs (valerate, isobutyrate, and isovalerate) (P = 0.008) compared with the other three groups. SB also significantly increased the concentration of butyrate compared with CON and SP (P = 0.024). However, the concentration of propionate did not differ significantly among the four groups (P = 0.585). TABLE 2. Body weight, hindgut weight, and the concentration of VFAs in colonic and cecal contents[92]^a Items CON SA SP SB SEM P-value Body weight and hindgut weight  Body weight (kg) 48.77 48.83 51.11 50.69 0.373 0.089  Colonic weight (g) 146.52 146.13 130.05 126.64 3.825 0.125  Cecal weight (g) 239.23b 296.14a 224.62b 270.50a 6.612 <0.001  Rectal weight (g) 182.20b 208.98a 168.43b 167.02b 3.856 <0.001 Concentration of VFAs in cecal contents  Total VFAs (μmol/g) 31.64b 40.38a 30.61b 34.03b 1.281 0.017  Acetate (μmol/g) 23.69b 30.75a 22.81b 24.49b 1.071 0.005  Propionate (μmol/g) 5.45 5.96 5.22 5.25 0.205 0.585  Butyrate (μmol/g) 1.34b 1.87ab 1.31b 2.85a 0.223 0.024  Other VFAs[93]^b (μmol/g) 1.16b 1.81a 1.27b 1.44b 0.081 0.008 [94]Open in a new tab ^^a Values with different letters within a row mean statistically significant (P < 0.05 or P < 0.01). ^^b Other VFAs included valerate, isobutyrate, and isovalerate. Serum biochemistry and antioxidant capacity parameters The serum biochemistry and antioxidant capacity parameters are shown in [95]Table 3. SP increased the concentration of LDH compared with the other three groups (P < 0.001), while both SA and SP increased the concentration of CREA compared with CON and SB (P = 0.003). No significant difference was observed among the four groups for other serum parameters, including BUN (P = 0.405), TG (P = 0.230), CHOL (P = 0.207), HDL-C (P = 0.602), and LDL-C (P = 0.340). For serum antioxidant capacity parameters, VFA administration did not impact serum T-AOC (P = 0.379), GSH-Px (P = 0.513), and CAT (P = 0.417) compared with CON. However, the concentration of SOD was significantly decreased in SA and SP compared with CON and SB (P = 0.038), and SB also decreased the concentration of MDA compared with SP (P = 0.042). TABLE 3. Serum biochemistry and antioxidant capacity parameters[96]^a Items CON SA SP SB SEM P-value Serum biochemistry parameters  LDH (U/L) 235.33b 242.44b 370.06a 234.44b 13.646 <0.001  CREA (μmol/L) 95.83b 107.08a 108.14a 98.63b 1.576 0.003  BUN (mmol/L) 4.69 4.56 3.93 4.66 0.178 0.405  TG (mmol/L) 0.08 0.07 0.06 0.09 0.005 0.230  CHOL (mmol/L) 2.08 1.62 1.73 1.91 0.081 0.207  HDL-C (mmol/L) 1.38 1.18 1.38 1.43 0.066 0.602  LDL-C (mmol/L) 0.58 0.48 0.50 0.57 0.024 0.340 Serum antioxidant capacity parameters  T-AOC (μmol/mL) 1.29 1.17 1.31 1.09 0.049 0.379  GSH-Px (U/mL) 240.11 203.80 260.88 215.48 14.145 0.513  CAT (U/mL) 5.78 4.20 4.99 4.37 0.358 0.417  SOD (U/mL) 9.16a 6.26b 6.72b 7.39ab 0.397 0.038  MDA (nmol/mL) 0.65ab 0.66ab 0.85a 0.21b 0.085 0.042 [97]Open in a new tab ^^a Values with different letters within a row mean statistically significant (P < 0.05 or P < 0.01). Cecal bacterial structure and selection of significant taxonomic differences [98]Table 4 and [99]Fig. 2 illustrate the hindgut bacterial patterns and alterations in their diversity and taxonomy of cecal contents. Regarding alpha diversity indexes ([100]Table 4), SA significantly decreased the Chao 1 index compared with CON and SP (P = 0.016); however, no significant difference was found in the Shannon (P = 0.367) and Simpson (P = 0.373) indexes among the four groups. Bacterial beta diversity, calculated using the PCoA diagram based on unweighted UniFrac distance, revealed milder separation among the four groups ([101]Fig. 1B), indicating that VFA administration influenced cecal bacterial succession patterns and community structure. At the phylum level, 10 phyla were identified, with the dominant bacterial phyla including Firmicutes, Bacteroidetes, Verrucomicrobia, Proteobacteria, etc. ([102]Fig. 1C). The dominant bacterial families were Oscillospiraceae, Rikenellaceae, Akkermansiaceae, UCG-010, Ruminococcaceae, Eubacterium coprostanoligenes group, Lachnospiraceae, Christensenellaceae, etc. ([103]Fig. 1C). The dominant bacterial genera included Akkermansia, Oscillospiraceae UCG-005, Rikenellaceae RC9 gut group, UCG-010 unclassified, Eubacterium coprostanoligenes group, Christensenellaceae R-7 group, etc. ([104]Fig. 1C). TABLE 4. Bacterial alpha diversity indexes in cecal content[105]^a Items CON SA SP SB SEM P-value Chao 1 index 493.30a 286.00b 494.36a 346.08ab 30.239 0.016 Shannon index 7.42 6.61 7.45 7.29 0.188 0.367 Simpson index 0.97 0.96 0.98 0.98 0.006 0.373 [106]Open in a new tab ^^a Values with different letters within a row mean statistically significant (P < 0.05 or P < 0.01). Fig 2. [107]LEfSe analysis identifies differentially abundant taxa between treatment groups and control, with distinct microbial markers enriched in SA, SP, and SB groups. Heatmap visualizes genus-level abundance patterns, revealing clustering and microbial shifts. [108]Open in a new tab Selection of candidate cecal bacteria in response to VFA administration and different bacterial taxonomy. (A) LEfSe analysis of candidate cecal bacteria and different bacterial taxonomy in response to VFA administration at the genus level, only LDA score greater than 3 was marked; (B) cluster heatmap of candidate cecal bacteria of four groups. The LEfSe method was used to identify candidate bacteria in response to VFA administration and significant bacterial taxonomy differences ([109]Fig. 2). Compared with CON, SA significantly decreased Akkermansia, Monoglobus, Phascolarctobacterium, Prevotellaceae UCG-001, Aeriscardovia, Enterorhabdus, Lachnospiraceae NK4B4 group, and Eubacterium oxidoreducens group while significantly increasing Bacteroides, Christensenellaceae R-7 group, Rikenellaceae dgA-11 gut group, and Desulfosporosinus ([110]Fig. 2A). Then, Oscillospiraceae UCG-005, Prevotellaceae UCG-004, Christensenellaceae R-7 group, Slackia, Rikenellaceae dgA-11 gut group, Candidatus Soleaferrea, and Frisingicoccus were enriched in SP, while Monoglobus, Phascolarctobacterium, Treponema, Prevotellaceae UCG-001, Oxobacter, Prevotellaceae UCG-003, Clostridium sensu stricto 1, Lachnospiraceae NK4B4 group, Blautia, and Succinivibrio were enriched in CON ([111]Fig. 2A). Finally, SB decreased Akkermansia, Monoglobus, Erysipelatoclostridium, Oxobacter, Lachnospiraceae NK4B4 group, Aeriscardovia, and Eubacterium oxidoreducens group while increasing Oscillospiraceae UCG-005, Rikenellaceae RC9 gut group, Rikenellaceae dgA-11 gut group, Christensenellaceae R-7 group, Bacteroides, Candidatus Saccharimonas, Defluviitaleaceae UCG-011, and Papillibacter compared with CON ([112]Fig. 2A). The cluster heatmap visualized in [113]Fig. 2B shows the higher relative abundance of significant taxonomic differences at the bacterial genus level in each group compared to others. Cecal metabolome profiling Cecal metabolome analysis using LC-MS was performed to study microbial metabolism profiling in response to VFA administration ([114]Fig. 3). A total of 410 small metabolites were identified among the four groups and QC samples. The PCA score plot indicated no separation among the groups and QC samples ([115]Fig. 3A). However, the OPLS-DA plot demonstrated that metabolic patterns in cecal contents were significantly separated among the groups, indicating that VFA administration significantly altered microbial metabolism patterns in the hindgut ([116]Fig. 3A). Fig 3. [117]PCA and OPLS-DA plots reveal metabolomic separation across treatment groups. Bar chart quantifies up- and downregulated metabolites. VIP plots identify key differential metabolites per treatment. Heatmap displays abundance patterns across all samples. [118]Open in a new tab The LC-MS metabolome profiles in cecal contents and selection of differential metabolites. (A) The distributions of all metabolites using PCA and OPLS-DA score plots of groups and QC samples; (B) the number of differential metabolites between each VFA and CON in cecal contents; (C) the VIP and log[2](FC) of differential metabolites; (D) the cluster heatmap of differential metabolites of four groups in cecal contents; n = 3 for LC-MS metabolome analysis. Metabolites with VIP > 1 and P < 0.05 were considered significantly different. In SA, 12 metabolites, including L-proline, aminoadipic acid, 3′-AMP, AMP, and chenodeoxycholic acid, were significantly decreased, while 13 metabolites, including isopyridoxal, L-homophenylalanine, androsterone, chlorpromazine, and urocanic acid, were significantly higher than those in CON ([119]Table S1; [120]Fig. 3B and C). In SP, 14 metabolites were decreased, including hydroxypyruvic acid, ectoine, citric acid, 5,6-DHET, and creatine, while seven metabolites, including adipate semialdehyde, p-hydroxyphenylacetic acid, O-acetylcarnitine, deoxycytidine, and GMP, were increased compared with CON ([121]Table S1; [122]Fig. 3B and C). In SB, nine metabolites were decreased, including 3-methyloxindole, shikimic acid, androstenedione, 5,6-DHET, and 2-isopropylmalic acid, while 13 metabolites, including 3,4-dihydroxymandelic acid, bitertanol, chelirubine, urocanic acid, and phenylpyruvic acid, were increased compared with CON ([123]Table S1; [124]Fig. 3B and C). The cluster heatmap identified the abundances of selected differential metabolites ([125]Fig. 3D). A correlation heatmap showed significant correlations between differential bacteria and metabolites, suggesting that alterations in hindgut metabolic patterns induced by VFA administration may be mediated by gut microbes ([126]Fig. 4A). Fig 4. [127]Correlation heatmap links specific metabolites to microbial genera. KEGG enrichment analysis reveals significantly altered metabolic pathways. Pathway heatmaps and comparison plots highlight treatment-specific metabolic shifts. [128]Open in a new tab Correlation analysis between differential metabolites and candidate bacteria, and functional enrichment analysis of differential metabolites. (A) Correlation analysis between differential metabolites and candidate bacteria. The analysis is based on the Spearman correlation coefficient. *P < 0.05; ^+P < 0.01; and ^#P < 0.001; (B) functional enrichment analysis of all differential metabolites using KEGG database of four groups; (C) heatmaps of the abundances of differential metabolites in metabolic pathways of four groups; (D) functional enrichment analysis of differential metabolites of each comparison using KEGG database. Functional enrichment analysis of differential metabolites Pathway enrichment analysis of cecal differential metabolites was performed ([129]Fig. 4). KEGG enrichment analysis of all differential metabolites after VFA administration revealed significant impacts on several metabolic pathways ([130]Fig. 4B). Differential metabolites were enriched in metabolic pathways, including purine metabolism (AMP, GMP, hypoxanthine, 3′-AMP, and 5-aminoimidazole-4-carboxamide); phenylalanine metabolism (phenylpyruvic acid, trans-cinnamate, p-hydroxyphenylacetic acid, and enol-phenylpyruvate); glycine, serine, and threonine metabolism (hydroxypyruvic acid, creatine, and ectoine); arachidonic acid metabolism (delta-12-prostaglandin J2, 5,6-EET, and 5,6-DHET); glyoxylate and dicarboxylate metabolism (citric acid and hydroxypyruvic acid); tyrosine metabolism (p-hydroxyphenylacetic acid and 3,4-dihydroxymandelic acid); and tryptophan metabolism pathway (kynurenic acid and 5-hydroxyindoleacetic acid). Functional enrichment analysis also revealed that, except for metabolic pathways, differential metabolites may also be enriched in other functional pathways. Differential metabolites were enriched in the longevity-regulating pathway, endocrine resistance, mTOR signaling pathway, and PI3K-Akt signaling pathway between SA and CON. Metabolites were enriched in the cGMP-PKG signaling pathway, olfactory transduction, and taste transduction between SP and CON. Most differential metabolites were enriched in biosynthesis of amino acids, serotonergic synapse, and intestinal immune network for IgA production between SB and CON ([131]Fig. 4D). Colonic epithelial transcriptome sequencing profiling We performed transcriptome sequencing of colonic epithelium samples to study the effect of microbiota and derived metabolites on epithelial metabolism, absorption, and signal transduction. The PCA using the combined data set of all genes indicated significant separation among the four groups ([132]Fig. 5A). Specifically, PCA of the SA, SP, and SB groups showed marked separation from the CON group ([133]Fig. 5A). The expression level of DEGs was further analyzed using a cutoff of P < 0.05. The number of DEGs between each VFA and CON is shown in [134]Fig. 5B, with detailed information in [135]Table S2. In detail, 1,466 DEGs were identified in SA, with 484 upregulated and 982 downregulated compared to CON. In the SP vs CON comparison, 2,009 DEGs were selected, with 708 upregulated and 1,301 downregulated. Finally, in the SB vs CON comparison, 696 DEGs were identified, with 368 upregulated and 328 downregulated. A Venn plot demonstrated the number of single- or co-expressed DEGs ([136]Fig. 5C; [137]Table S3). Among all comparisons, 136 DEGs, including FAM64A, IGFN1, IL17B, MYOM2, and TNFSF11, were screened, and their expression levels among the four groups are shown in [138]Fig. 5D. These genes can be suggested as marker genes for hindgut epithelial responses to VFAs. Fig 5. [139]PCA plots reveal distinct transcriptomic profiles among treatment groups. Volcano plot displays DEGs in each comparison. Venn diagram shows overlap of DEGs across treatments. Heatmap illustrates expression patterns of co-expressed DEGs. [140]Open in a new tab The transcriptome sequencing profiles in colonic epithelium. (A) The PCA using a combined data set of all genes of four groups; (B) the number of up- and downregulated DEGs between each VFA and CON; (C) Venn diagram of the number of single- and co-expressed DEGs between each VFA and CON; (D) heatmap of the expression level of the co-expressed DEGs in four groups; n = 3 for transcriptome sequencing analysis. Functional enrichment analysis of DEGs The up- and downregulated DEGs underwent functional enrichment analysis using KEGG and GO databases. Results revealed that the administration of each VFA contributed to different epithelial responses. For KEGG enrichment ([141]Fig. 6A), most DEGs between SA and CON were enriched in pathways such as salivary secretion, insulin secretion, mineral absorption, calcium signaling, and pancreatic secretion. Between SP and CON, most DEGs were enriched in pathways including salivary secretion, circadian entrainment, cholinergic synapse, insulin secretion, mineral absorption, and gastric acid secretion. Finally, DEGs between SB and CON were enriched in pathways related to viral protein interaction with cytokine and cytokine receptor, cytokine-cytokine receptor interaction, interleukin-17 (IL-17) signaling pathway, TNF signaling pathway, and T cell receptor signaling pathway. Fig 6. [142]Bubble plots display KEGG and GO enrichment results for differentially expressed genes across treatment groups. KEGG pathways include cytokine signaling and immune response, while GO terms depict immune activation, transport, and muscle-related processes. [143]Open in a new tab The KEGG and GO functional enrichment analyses of DEGs. (A) The functional enrichment analysis of DEGs using the KEGG database; (B) the functional enrichment analysis of DEGs using the GO database. The GO enrichment results also indicated varying degrees of biological process impact from the VFA administration ([144]Fig. 6B). For instance, most DEGs were enriched in behavior, muscle development and differentiation, and metal ion (such as calcium ion) homeostasis and transport-related pathways. Several immune response and inflammatory pathways were highly enriched between SB and CON. The protein concentration of inflammatory cytokines and tight junctions in hindgut epithelium The protein concentration of inflammatory cytokines and tight junctions in hindgut epithelium was measured using ELISA methods ([145]Table 5). For inflammatory cytokines in colonic epithelium, data revealed that administration of three VFAs had no effect on IL-1β (P = 0.196), IL-6 (P = 0.204), and TNF-α (P = 0.238), while SA and SP increased the concentration of IL-10 compared with CON and SB (P = 0.001). Regarding tight junctions in colonic epithelium, the concentration of Claudin1 in SB was higher than that in SA and SP (P = 0.01), while the concentration of ZO-1 in SA was higher than that in CON (P = 0.034). The differences in Occludin were not significant among the four groups (P = 0.934). Similar results were found in the cecal epithelium. SA and SP significantly increased the concentration of IL-10 compared with CON and SB (P = 0.001). Administration of three VFAs and saline did not affect the concentrations of IL-1β (P = 0.16), IL-6 (P = 0.135), and TNF-α (P = 0.277). For tight junctions in cecal epithelium, SA increased the concentration of Claudin1 (P = 0.01) compared with the other three groups, while the differences in Occludin (P = 0.766) and ZO-1 (P = 0.166) were not significant among the four groups. TABLE 5. The concentrations of inflammatory cytokines and tight junctions in colonic and cecal epithelial tissues[146]^a Items CON SA SP SB SEM P-value Colonic epithelial tissue  IL-1β (pg/mg protein) 14.88 15.47 15.54 16.90 0.342 0.196  IL-6 (pg/mg protein) 33.03 32.07 29.24 30.14 0.699 0.204  IL-10 (pg/mg protein) 40.91c 48.60a 45.74ab 44.07bc 0.785 0.001  TNF-α (pg/mg protein) 119.53 118.28 122.29 107.99 2.614 0.238  Claudin1 (pg/mg protein) 11.21ab 10.39bc 9.97c 11.58a 0.204 0.010  Occludin (ng/mg protein) 1.19 1.14 1.18 1.18 0.030 0.934  ZO-1 (ng/mg protein) 115.80b 134.55a 126.13ab 124.48ab 2.359 0.034 Cecal epithelial tissue  IL-1β (pg/mg protein) 27.30 29.14 26.34 26.93 0.462 0.160  IL-6 (pg/mg protein) 57.47 56.86 59.62 53.84 0.884 0.135  IL-10 (pg/mg protein) 19.34b 22.45a 22.11a 18.61b 0.427 0.001  TNF-α (pg/mg protein) 210.07 223.90 221.44 214.76 2.746 0.277  Claudin1 (pg/mg protein) 21.16b 22.54a 20.20bc 19.90c 0.281 0.001  Occludin (ng/mg protein) 2.21 2.18 2.18 2.27 0.030 0.766  ZO-1 (ng/mg protein) 213.42 217.07 223.05 199.51 3.847 0.166 [147]Open in a new tab ^^a Values with different letters within a row mean statistically significant (P < 0.05 or P < 0.01). The interactions between candidate bacteria and homeostasis status, metabolome, and transcriptome sequencing data The interactions between candidate bacteria and hindgut homeostasis, as well as multi-omics data in response to different VFA administrations, were studied using Pearson’s correlation and Mantel’s test ([148]Fig. 7). Most of these gut bacteria were correlated with each other, as well as differential metabolites and DEGs in gut epithelium; however, these bacteria were rarely correlated with inflammatory cytokines and tight junctions. For example, between SA and CON ([149]Fig. 7A), Christensenellaceae R-7 group was positively correlated with Desulfosporosinus (Pearson’s P < 0.01); meanwhile, the Christensenellaceae R-7 group was correlated with differential metabolites (Mantel’s P < 0.05) and DEGs (Mantel’s P < 0.01), while Desulfosporosinus was correlated with serum antioxidant capacity parameters (Mantel’s P < 0.05) and DEGs (Mantel’s P < 0.01). Between SP and CON ([150]Fig. 7B), Prevotellaceae UCG-004 was positively correlated with Frisingicoccus (Pearson’s P < 0.001), and both of these two bacteria were correlated with differential metabolites (Mantel’s P < 0.01) and DEGs (Mantel’s P < 0.05 or Mantel’s P < 0.01). Between SB and CON ([151]Fig. 7C), Akkermansia was negatively correlated with Oscillospiraceae UCG-005 (Pearson’s P < 0.01), and both of these two bacteria were correlated with DEGs (Mantel’s P < 0.01). Fig 7. [152]Correlation network plots for SA, SP, and SB vs. control link gut microbiota with serum, colonic parameters, metabolites, and gene expression. Edge thickness indicates Mantel's R strength; edges represent positive and negative associations, respectively. [153]Open in a new tab The interactions between candidate bacteria and homeostasis status, metabolome, and transcriptome sequencing data. (A) Correlation analysis between SA and CON; (B) correlation analysis between SP and CON; (C) correlation analysis between SB and CON; the correlation analysis of candidate bacteria was calculated using Pearson’s correlation coefficient. *P < 0.05; **P < 0.01; and ***P < 0.001. The correlation analysis between candidate bacteria and homeostasis status, metabolome, and transcriptome sequencing data was performed using Mantel’s test. Blue line indicates Mantel’s P < 0.05, red line indicates Mantel’s P < 0.01, and the gray line indicates Mantel’s P > 0.05. DISCUSSION In most studies reported in the literature, supplementation with VFAs has shown several advantages for ruminants. For example, SA has been demonstrated to increase milk yield ([154]46) and promote milk fat synthesis ([155]47, [156]48), while SP can improve nitrogen utilization, glucose metabolism, and protect the blood-milk barrier integrity ([157]49[158]–[159]51). SB has also been reported to promote rumen papillae development and reduce inflammation and autophagy in the rumen epithelium ([160]24, [161]52, [162]53). However, limited data are available on the role of VFAs in hindgut bacteria, metabolism, and epithelial homeostasis in dairy goats. In our study, we systematically evaluated the effects of VFAs on hindgut bacterial composition and metabolism, epithelial homeostasis parameters, and transcriptome profiling in dairy goats. It is well-known that the majority of VFAs produced in the rumen are absorbed through the rumen wall, but a portion does pass to the lower digestive tract ([163]54). Our study demonstrated that oral administration of VFAs into the rumen also had a significant impact on the concentration of VFAs in the hindgut. In detail, data showed that SA significantly increased the concentration of total VFAs, acetate, and other VFAs (valerate, isobutyrate, and isovalerate) in the cecal contents compared with the other groups. SB also significantly increased the concentration of butyrate compared with CON and SP. These results indicated that despite the primary absorption of VFAs through the rumen, sufficient quantities of administered VFAs reached the cecum, thereby influencing the microbial composition and metabolic activities. This is supported by the dynamics of VFA absorption and passage rates from rumen to cecum, and the variations in bacterial fermentation patterns and metabolic utilization of VFAs of the rumen from our previous study ([164]27). Furthermore, we reported that VFA administration decreased the expression levels of VFA absorption genes, such as MCT1 and MCT4 in the rumen epithelium. This may lead to changes in the systemic absorption and metabolism of VFAs, further affecting the production of VFAs in the cecum ([165]27). Moreover, the difference between the rumen and the hindgut of ruminants lies in the lack of factors such as saliva and protozoa, resulting in weak buffering ability, and the VFAs entering the hindgut cannot be absorbed by the intestinal epithelium in time, resulting in changes in VFA concentration ([166]12, [167]54). The 16S rRNA sequencing data demonstrated that VFA administration significantly impacted the succession of cecal bacteria. SA increased the relative abundance of Bacteroides, Christensenellaceae R-7 group, Rikenellaceae dgA-11 gut group, and Desulfosporosinus; SP increased the relative abundance of Oscillospiraceae UCG-005, Prevotellaceae UCG-004, Christensenellaceae R-7 group, Slackia, Rikenellaceae dgA-11 gut group, Candidatus Soleaferrea, and Frisingicoccus; and SB increased the relative abundance of Oscillospiraceae UCG-005, Rikenellaceae RC9 gut group, Rikenellaceae dgA-11 gut group, Christensenellaceae R-7 group, Bacteroides, Candidatus Saccharimonas, Defluviitaleaceae UCG-011, and Papillibacter. Previous studies have linked these bacteria, such as Bacteroides, Rikenellaceae dgA-11 gut group, Christensenellaceae, Papillibacter, and Rikenellaceae RC9 gut group, with the production of VFAs ([168]23, [169]55[170]–[171]57). Therefore, our data indicated that VFAs could significantly affect the abundances of VFA-producing bacteria, thereby altering microbial fermentation and metabolic patterns. Interestingly, administration with three VFAs increased the relative abundance of Christensenellaceae R-7 group in the cecum, a gut bacterium that is reported to be more abundant in healthy people than in people with inflammatory bowel disease, suggesting that VFA administration may benefit gut health in goats ([172]58). The SA-induced reduction in Chao1 index was consistent with a significant enrichment of specialized taxa, including Bacteroides, Christensenellaceae R-7 group, and Rikenellaceae dgA-11 gut group, accompanied by a decrease of mucin-degrading bacteria Akkermansia. Similar alterations in microbial abundances were also observed in the SB administration group, along with a downward trend in the Chao1 index in SB. These taxonomic shifts suggested that high acetate availability favors fast-growing, acetate-utilizing bacteria (such as Bacteroides), which may outcompete rare species through niche specialization, thereby reducing overall richness while promoting functional dominance ([173]59, [174]60). Our findings indicated that VFA administration had a limited effect on systemic lipid metabolism in goats, which was supported by insignificant differences in serum TG and CHOL concentrations among the four groups. This result could reflect the efficient energy utilization strategy in ruminants. The administered VFAs may be directly absorbed by the rumen and intestinal epithelium and further oxidized for energy supplementation rather than being used for lipid synthesis, thereby maintaining the systemic lipid homeostasis ([175]61). Furthermore, data revealed that VFA administration could enhance gut homeostasis by modulating inflammatory responses and tight junctions. For instance, SB decreased the MDA level in serum, which could alleviate systemic oxidative stress ([176]62). Additionally, SA and SP increased the level of IL-10 in both colonic and cecal epithelium, indicating their anti-inflammatory effects in gut epithelium, which further improve the resistance of the intestinal epithelium to pathogen invasion and enhance the barrier functions ([177]62, [178]63). The ELISA results also suggested that different VFA administrations increased the concentration of tight junctions. SB increased the protein concentration of Claudin1 in colonic epithelium, while SA increased the protein concentration of ZO-1 in colonic epithelium and Claudin1 in cecal epithelium. It has been previously reported that the Th1, Th2, and Th17 cytokines play an important role in maintaining gut homeostasis ([179]64, [180]65). Furthermore, VFAs, particularly butyrate, exert their protective roles in gut inflammation through the inhibition of histone deacetylases (HDACs, such as HDAC3) in gut epithelium ([181]66). VFAs can suppress the differentiation of Th1 and Th17 cells to reduce the production of pro-inflammatory cytokines (such as IFN-γ, TNF, and IL-17), while promoting Th2-mediated secretion of the anti-inflammatory cytokines (such as IL-10) ([182]67, [183]68). Our findings aligned with the observed dual immunomodulatory effects of VFAs. The reduction level of MDA and the increased concentrations of IL-10 and ZO-1/Claudin1 tight junctions demonstrated that the VFAs had positive roles in alleviating oxidative stress, mitigating inflammatory responses, and preserving the integrity of gut barrier functions in goats. These findings could provide guidance for the prevention and management of subacute acidosis induced by high-concentrate diet in ruminants ([184]69). VFA administration significantly reshaped hindgut metabolic profiles, with purine, tryptophan, phenylalanine, and tyrosine metabolism being the most responsive pathways. Purine metabolites regulate energy homeostasis and signaling transduction, while tryptophan derivatives (kynurenic acid and 5-hydroxyindoleacetic acid) modulate immunity and anti-inflammation functions ([185]70[186]–[187]73). Notably, SA treatment reduced cecal microbial richness (Chao1), driven by the dominance of acetate-utilizing taxa (Bacteroides and Christensenellaceae R-7 group) and the reduction of Akkermansia. This taxonomic shift correlated with decreased purine (3′-AMP, AMP), steroid (androstenedione), and amino acid (L-proline) metabolites, indicating reduced metabolic redundancy. However, SA upregulated anti-inflammatory and antioxidant metabolites, such as 5-hydroxyindoleacetic acid, urocanic acid, and 3,4-dihydroxymandelic acid (ROS scavenger), suggesting compensatory mechanisms to maintain gut homeostasis ([188]74, [189]75). These findings demonstrated that SA-induced microbial simplification prioritizes niche efficiency (antioxidant capacity, anti-inflammation, and barrier function) over metabolic versatility, offering a targeted strategy to optimize gut health by acetate-driven microbiota remodeling. The functional enrichment analysis of transcriptome sequencing data of colonic epithelium suggested that VFA administration had varying effects on nutrient absorption and transport, development, homeostasis, and immunity of colonic epithelium. SA and SP affected endocrine and digestive functions, metal ion (particularly calcium ion) homeostasis and transport, and muscle development and differentiation in colonic epithelium. Several immune response and inflammation-related pathways were especially enriched in SB. Intracellular calcium ions play a crucial role in maintaining biopotential, nerve conduction, and energy absorption in intestinal epithelial cells of ruminants ([190]76). Previous studies have shown that VFAs can activate intracellular calcium ion signaling in rats ([191]77). Our study reported that VFAs can also regulate calcium ion signaling pathways in the hindgut of dairy goats. Additionally, muscle development and differentiation pathways were enriched in SA compared with CON, indicating that VFAs, such as acetate, can regulate energy metabolism in the colon ([192]78). Finally, apart from SA and SP, our results explained that SB had its unique role in regulating intestinal epithelial homeostasis and immunity, similar to studies in the rumen ([193]79[194]–[195]81). Conclusions This study highlighted the significant roles of VFAs on hindgut bacterial metabolism and homeostasis and provided novel insights into the metabolic and physiological roles beyond the rumen. VFA administration positively affects hindgut epithelial health and barrier function. Specifically, SB reduced serum MDA levels, indicating a decrease in oxidative stress, whereas SA and SP increased IL-10 concentration, reflecting an enhanced anti-inflammatory response. VFA administration also contributed to gut barrier integrity by increasing the expression of tight junctions. Data also revealed that VFAs modulated microbial communities, and the changes in microbial composition were accompanied by significant shifts in metabolic pathways. Transcriptomic analysis further demonstrated that SA and SP influenced varying aspects of colonic epithelial function, including endocrine activity, digestive processes, metal ion homeostasis, and muscle development. Additionally, immune response pathways were enriched in SB. These findings provided a comprehensive understanding of the nutritional regulatory effects of VFAs and suggested potential strategies for improving gut health and overall well-being in ruminants through dietary interventions. ACKNOWLEDGMENTS