2′-Fucosyllactose and its metabolites propionate/butyrate suppress inflammation through a shared TLR4/p38 MAPK-dependent pathway in vitro and in vivo

TL;DR

2'-Fucosyllactose and its metabolites reduce inflammation by targeting the TLR4/p38 MAPK pathway in both lab and animal models.

Contribution

This study reveals a shared anti-inflammatory mechanism of 2'-FL and its metabolites through TLR4/p38 MAPK signaling.

Findings

2'-FL increases propionate and butyrate production and enriches specific gut bacteria.

2'-FL, propionate, and butyrate suppress inflammation via TLR4/p38 MAPK signaling in cells and mice.

Molecular docking shows 2'-FL and its metabolites bind to TLR4–MD2 with specific affinity.

Abstract

2′-Fucosyllactose (2′-FL), the most abundant human milk oligosaccharide (HMO), exhibits potent anti-inflammatory properties, yet the underlying mechanisms remain incompletely understood. This study employed an integrated approach combining in vitro fermentation systems, Caco-2 cell assays, in vivo murine colitis models, and molecular docking simulations to elucidate the dual mechanisms of 2′-FL action. In vitro fermentation revealed that 2′-FL selectively enhanced propionate and butyrate production while enriching Bacteroides acidifaciens and Allobaculum stercoricanis, key short-chain fatty acids (SCFAs)-producing taxa. In enterotoxigenic Escherichia coli (ETEC)-infected Caco-2 cells, 2′-FL suppressed TLR4/p38 MAPK signaling and inflammation, an effect mirrored by propionate and butyrate. Similarly, in the dextran sulfate sodium (DSS)-induced mouse colitis model, 2′-FL inhibited this signaling pathway, alleviated inflammation, and, in parallel, specifically elevated levels of propionate and butyrate. Molecular docking analyses also indicated that 2′-FL and its metabolites, propionate and butyrate, exhibit binding to the TLR4–MD2 complex with specific affinity. These results support a coordinated mechanism whereby 2′-FL alleviates intestinal inflammation through putative TLR4 antagonism and the microbiota-derived SCFAs, both of which converge to modulate the TLR4/p38 MAPK-dependent signaling axis. Our work provides a mechanistic foundation for 2′-FL as a precision prebiotic targeting gut inflammation via host-microbe crosstalk.

Image 1

•2′-Fucosyllactose specifically upregulates the levels of propionate and butyrate.•2′-Fucosyllactose, propionate, and butyrate modulate Toll-like receptor 4 activity.•2′-Fucosyllactose, propionate, and butyrate share the p38 signaling pathway.

2′-Fucosyllactose specifically upregulates the levels of propionate and butyrate.

2′-Fucosyllactose, propionate, and butyrate modulate Toll-like receptor 4 activity.

2′-Fucosyllactose, propionate, and butyrate share the p38 signaling pathway.

Figures

Fig. 7Effect of 2′-FL on the mitigation of DSS-induced colitis of mice. (A) Animal experimental design; (B) Mice daily weight gain (n = 10); (C-D) Colon length and statistics (n = 6); (E) DAI score (n = 6); (F) H&E and PAS staining (n = 6); (G) Histological Score (n = 6); (H-I) Analysis of relative ZO-1 and Claudin-1 expression in colon tissue by indirect immunofluorescence assay (n = 3); (J-L) Detection of IL-6, IL-8 and TNF-α, expression levels in mouse serum by ELISA (n = 3). Data were represented as mean ± SD. Different lowercase letters above each column indicate significant differences b

Full text

Introduction

1

Human milk oligosaccharides (HMOs) are bioactive carbohydrate structures naturally present in human milk. They serve as bifidus factors in breast milk, ranking third in abundance, following lactose and fat (Dinleyici et al., 2023; Kassai and de Vos, 2024). These unique sugars are not only abundant but also exhibit significant structural diversity, with over 200 distinct structures identified to date (Gonsalves et al., 2025; Zhu et al., 2023). This structural diversity confers multiple roles on HMOs in promoting human health (Farhadihosseinabadi et al., 2020; Kang et al., 2020; Li et al., 2024b). Advancements in synthetic biology have enabled the amplification and mass production of certain HMOs through biosynthesis, facilitating their incorporation as food ingredients in infant formula (Liu et al., 2025; Yue et al., 2025; Zhang et al., 2025a). The highest concentration of oligosaccharides in human milk is attributed to 2′-fucosyllactose (2′-FL), constituting approximately 30% of total HMOs (Vandenplas et al., 2018). Its high abundance contributes to its significance and extensive research in the field. 2′-FL is recognized for its roles in regulating gut microbiota, inhibiting pathogen adhesion, modulating immune responses, and promoting brain development in infants, rendering it a topic of considerable interest in health research (Zhang et al., 2022; Zhu et al., 2022). As a key representative of HMOs, 2′-FL was one of the first HMOs to receive approval for use in both infant formula and dietary supplements (Liu et al., 2022b).

Structurally, 2′-FL is synthesized through the linkage of one lactose molecule and one fucose molecule via an α-1,2-glycosidic bond (Bode, 2012). As a functional oligosaccharide, 2′-FL contributes significantly to gut homeostasis via diverse biological pathways, including modulation of gut microbial composition and their metabolic activities (Cai et al., 2025; Chen et al., 2025a). Emerging evidence demonstrates that 2′-FL exerts systemic biological effects, such as ameliorating age-related osteoporosis via gut microbiota regulation (Li et al., 2025) and enhancing neurodevelopment in murine models through gut microbiota-derived serotonin (5-HT)-mediated pathways (Zhu et al., 2025). The crosstalk between 2′-FL and the host is partially elucidated by microbial metabolites derived from 2′-FL (Zhang et al., 2023). 2′-FL in synergy with Bifidobacterium bifidum DNG6 ameliorates intestinal injury via the SCFAs-TLR4/NF-κB pathway (Sun et al., 2025). Notably, short-chain fatty acids (SCFAs), a major class of microbial metabolites, are recognized mediators of immunomodulation through several well-characterized pathways. Butyrate, for instance, alleviates ulcerative colitis via activation of the Nrf2/GPX4 axis (Chen et al., 2024), whereas propionate attenuates intestinal ischemia–reperfusion injury by modulating the AhR/Notch1 pathway (Liang et al., 2026).

In addition to indirect effects via microbial metabolism, 2′-FL directly modulates the activity of specific host receptors to exert its functional properties. For example, 2′-FL has demonstrated efficacy in preventing adult-onset colitis by transactivating the epidermal growth factor receptor (EGFR) in intestinal epithelial cells (Kaur et al., 2026). Moreover, 2′-FL exerts intestinal immunomodulatory effects through receptor interactions, particularly via TLR4-mediated signaling pathways, demonstrating therapeutic potential in inflammatory bowel diseases through alleviation of ulcerative colitis and prevention of necrotizing enterocolitis (Chen et al., 2025b; Sodhi et al., 2021). In contrast to conventional oligosaccharides such as galacto-oligosaccharides (GOS) and fructo-oligosaccharides (FOS), the full mechanistic basis for the anti-inflammatory effects of 2′-FL remains elusive (Shivakoti et al., 2026), especially concerning the potential synergistic interaction between its direct receptor-mediated functions and its indirect effects mediated through the gut microbiota.

To bridge these knowledge gaps, the present study utilized an integrated, multidisciplinary approach incorporating in vitro gut fermentation systems, inflammatory cell models, a murine dextran sulfate sodium (DSS)-induced colitis model, and molecular docking analyses. This comprehensive strategy was designed to delineate the direct and indirect mechanisms underlying the anti-inflammatory effects of 2′-FL in intestinal inflammation, thereby offering critical insights to inform its therapeutic development for gastrointestinal health.

Materials and methods

2

Reagents, bacterial strain, and colon cells

2.1

2′-FL was biosynthesized and provided by Cabio Biotech (Wuhan) Co., Ltd. (Wuhan, China), and GOS were obtained from Shanghai Macklin Biochemical Technology Co., Ltd. (Shanghai, China). The mass-to-charge ratio of 2′-FL [M-H] is 487.1664, with a purity of ≥94%, and GOS displays a mass-to-charge ratio of 377.0853, with a purity of ≥98%. The mass spectra of 2′-FL and GOS are shown in Supplementary Fig. S1. Standard compounds of propionic acid (PA, CAS No.79-09-4) and butyric acid (BA, CAS No.107-92-6) were obtained from Shanghai Aladdin Biochemical Technology Co., Ltd. In our laboratory, a strain of enterotoxigenic E. coli BE311 was isolated and characterized, and its mCherry-tagged variant, BE311-mCherry, was subsequently developed in a previous study (Lu et al., 2022). Caco-2 cells were purchased from Wuhan Pricella Biotechnology Co., Ltd. (Wuhan, China). DSS (molecular weight 40 kDa) was purchased from Aladdin Biochemical Technology Co., Ltd. (Shanghai, China).

In vitro fermentation experiment

2.2

The experiment was carried out as previously described (Zhou et al., 2025) with slight modifications to suit our experimental conditions. Briefly, fresh feces were collected from 8 specific-pathogen-free (SPF) C57BL/6 mice (4 weeks of age), with equal numbers of males and females (n = 4 per sex). After a 3-day acclimation, 1 g of feces was suspended in an anaerobic tube containing 9 mL of sterile phosphate-buffered saline (PBS), vortexed to homogeneity, and allowed to stand for 10 min. The supernatant was collected and diluted 5-fold with sterile PBS, then inoculated into the fermentation medium at a 2% (v/v) inoculum.

The fermentation medium was prepared as described previously (Li et al., 2019) with the following composition per 1 L of distilled water: 3 g peptone, 3 g tryptone, 4 g yeast extract, 0.50 g L-cysteine, 0.50 g bile salt, 0.02 g heme, 0.20 g MgCl_2_·6H_2_O, 2 g NaCl, 0.10 g CaCl_2_·6H_2_O, 1 g KCl, 10 μL vitamin K1, 0.40 g KH_2_PO_4_, and 1 mL Tween-80. The medium was dispensed into anaerobic tubes, sparged with N_2_, and sterilized at 121 °C for 30 min.

The in vitro fermentation experiment comprised three groups: a blank control (CON), a reference control (GOS), and a treatment group (2′-FL). For both GOS and 2′-FL, five substrate concentrations (0.1, 0.5, 1.0, 2.0, and 5.0 mg/mL) were evaluated across 12, 24, and 36 h. Guided by OD_600nm_ and pH as primary readouts, the optimal fermentation condition for 2′-FL was established at 2.0 mg/mL for 24 h (Figs. S2 and S3).

Cell culture

2.3

Caco-2 cells were cultured in 6-well plates in minimum essential medium (MEM; PM150410, Pricella, China) supplemented with 10% (v/v) fetal bovine serum (FBS; 164210-50, Pricella, China) and 1% (v/v) penicillin–streptomycin solution (BL505A, Biosharp, China) at 37 °C in a 5% CO_2_ incubator. At approximately 80% confluence, cells were pre-treated with 2′-FL or GOS at 50 μg/mL for 12 h. SCFAs were included at the same pre-treatment concentration (2 mM) and under the same conditions. After three washes with sterile PBS, cells were infected with ETEC at 10^6^ CFU/mL for 6 h and then processed for subsequent cellular assays.

Animal experiment

2.4

Fifty 4-week-old male C57BL/6 mice of specific pathogen-free (SPF) status were obtained from Three Gorges University (Yichang, China). Animals were housed in the Laboratory Animal Research Center of South Central Minzu University under controlled conditions (23 ± 2 °C; 12 h light/dark cycle). Following a 3-day acclimatization with ad libitum access to food and water, all procedures complied with national guidelines and were approved by the Scientific Ethics and Safety Committee of South Central Minzu University (No. 2024 SCUEC 034, Wuhan, China).

Mice were randomly assigned to four groups (n = 10 per group): control, DSS induced colitis, GOS reference control, and 2′-FL treatment. After acclimatization, GOS and 2′-FL were each administered at 0.5 mg/mL in PBS for 21 days. To induce colitis, mice in the DSS, GOS, and 2′-FL groups received 2.5% (w/v) DSS in PBS for 7 days; control mice received PBS throughout the study (Fig. 7A). Disease Activity Index (DAI), as defined by Chen et al. (2021), was determined on day 28. On day 29 mice were euthanized for collection of colon tissue and blood, which were stored at −80 °C for subsequent analyses. Mice were allocated to different experimental groups based on assay requirements: three were randomly selected for blood collection and enzyme-linked immunosorbent assay (ELISA); six for colon tissue harvesting and histology; and three for immunofluorescence and Western blot (WB) analyses. This stratified random-sampling strategy ensured sufficient sample sizes for each endpoint while maintaining statistical power.

Hematoxylin-Eosin (H&E) and Periodic Acid-Schiff (PAS) staining

2.5

Colonic tissues from mice were fixed with 4% paraformaldehyde, dehydrated, and embedded in paraffin to generate tissue blocks. Sections of 5 μm thickness were cut, deparaffinized in xylene, and sequentially stained with hematoxylin and eosin (H&E) for morphological evaluation or periodic acid–Schiff (PAS) for detection of glycoproteins. After staining, sections were rinsed, air-dried, and mounted with neutral resin. Colonic tissue injury was assessed using a Nikon ECLIPSE E100 optical microscope (Nikon, Japan), enabling detailed microscopic analysis of tissue architecture and integrity. Histological scoring was performed as described previously (Chen et al., 2021).

Indirect immunofluorescence assay

2.6

Mouse colon tissues were embedded in paraffin, sectioned at 6 μm, and deparaffinized in xylene (Servicebio, Wuhan, China) for 30 min. Sections were rehydrated through a graded ethanol series (100%, 90%, 75%) and rinsed with distilled water. Antigen retrieval was performed by boiling the sections in citrate buffer (pH 6.00) containing 0.05% Tween-20 for 30 min. After cooling, sections were blocked with 5% bovine serum albumin (BSA) for 90 min, followed by incubation with primary antibodies against Claudin-1 and ZO-1 at room temperature for 5-6 h. After PBS washes, sections were incubated with the corresponding fluorescently labeled secondary antibody for 2 h, then washed again to remove unbound antibody. Nuclei were counterstained with DAPI, and the expression levels of the tight junction proteins Claudin-1 and ZO-1 in colon tissue were visualized under a fluorescence microscope. Using ImageJ software, fluorescence intensity was measured, and the IntDen/Area values were derived and subjected to statistical analysis as described (Wang et al., 2025).

ELISA analysis of serum levels of IL-6, IL-8, and TNF-α

2.7

Fresh mouse blood was collected and centrifuged at 4 °C at 2500×g for 10 min to obtain serum. The serum supernatant was carefully aspirated and stored at −80 °C until analysis. Serum levels of the inflammatory cytokines TNF-α, IL-8, and IL-6 were measured using ELISA kits (Shanghai Enzyme-linked Biotechnology Co., Ltd.).

Quantification of SCFAs level in samples

2.8

The fecal fermentation supernatant and mouse feces samples were processed and analyzed independently, with appropriate modifications to the method described previously (Feng et al., 2024). For the fermentation supernatant, 50 μL was mixed with 100 μL of 0.5% (v/v) phosphoric acid and 750 μL of methyl tert-butyl ether (MTBE) containing internal standards. For feces, 100 mg was homogenized with 100 μL of 0.5% (v/v) phosphoric acid and 750 μL of MTBE containing internal standards. Each mixture was vortexed for 3 min, sonicated for 5 min, and then centrifuged at 4 °C and 12 000×g for 8 min. The resulting supernatants were collected and subjected to GC-MS/MS analysis according to the method of Cheng et al. (2025), using an Agilent 7890B gas chromatograph coupled to 7000D triple quadrupole mass spectrometer (Agilent Technologies, USA).

Full-length 16S rRNA gene sequencing of microbiota in fermentation broth

2.9

According to the manufacturer's protocol, total bacterial DNA was extracted from 24 h fermented fecal bacterial cultures using the DNeasy® PowerSoil® Kit (Qiagen, Germany). DNA purity and concentration were assessed by Nanodrop spectrophotometry, and fragment size distribution was evaluated by agarose gel electrophoresis. The full-length 16S rRNA gene was amplified using the universal primer pair 27F and 1492R. Amplicon quantification was performed with a Qubit fluorometer, followed by end-repair and dA-tailing. Subsequent steps comprised adapter ligation, library purification with magnetic beads, and final sequencing on a PromethION P48 sequencer (Oxford Nanopore Technologies, Oxford, UK).

Flow cytometric analysis of intracellular ROS levels in Caco-2 cells

2.10

The pretreated Caco-2 cells were washed three times with sterile PBS. According to the manufacturer's instructions for the Reactive Oxygen Species Assay Kit (S0033S, Beyotime, China), cells were detached with 0.25% trypsin–EDTA (25200056, Gibco, USA), collected, and centrifuged at 1000×g for 5 min. The supernatant was discarded, and the cell pellet was resuspended in DCFH-DA at a final concentration of 10 μmol/L, with the cell density adjusted to 1 × 10^6^ cells/mL. Cells were then incubated at 37 °C for 20 min, gently inverting the tube every 5 min to ensure complete probe uptake. After incubation, cells were washed three times with PBS to remove unbound probe, filtered through a 300-mesh nylon mesh, and analyzed by flow cytometry (BD Accuri™ C6 Plus, USA).

Analysis of ABTS radical scavenging capacity, MDA content, and SOD activity in Caco-2 cells

2.11

For the measurement of ABTS radical-scavenging capacity and malondialdehyde (MDA) content, cells were lysed with RIPA lysis buffer (G2002, Servicebio, China); for superoxide dismutase (SOD) activity, cells were lysed with SOD sample preparation buffer. All cell lysates were centrifuged at 12 000×g for 10 min, supernatants were collected, and protein concentrations were determined using a BCA assay kit (BL521A, Biosharp, China).

ABTS radical-scavenging capacity, SOD activity, and MDA content were measured with the Total Antioxidant Capacity Assay Kit (ABTS method; S0119, Beyotime, China), Total Superoxide Dismutase Assay Kit (WST-8 method; S0101S, Beyotime, China), and Lipid Peroxidation MDA Assay Kit (S0131S, Beyotime, China), respectively.

All measurements were carried out strictly according to the manufacturers’ instructions. Briefly, for the ABTS assay, samples were added to a 96-well plate containing ABTS working solution, incubated at room temperature for 5 min, and absorbance was recorded at 734 nm. For the SOD assay, WST-8/enzyme working solution and reaction-initiating solution were dispensed into the plate, followed by the sample; the mixture was incubated at 37 °C for 30 min and absorbance measured at 450 nm. For the MDA assay, MDA detection reagent was added to the sample, heated at 100 °C for 60 min, cooled to room temperature, centrifuged at 1000×g for 10 min, and the supernatant was assayed at 532 nm.

RT-qPCR analysis

2.12

Total cellular RNA was extracted using trizol reagent (Invitrogen, USA). The concentration and purity of RNA were determined using a NanoDrop 2000 spectrophotometer (Thermo Fisher Scientific, USA). 1 μg of total RNA was taken and reverse transcribed into first-strand cDNA using the HiScript II RT SuperMix for qPCR reverse transcription kit (R223-01, Vazyme, China). Real-time quantitative PCR (RT-qPCR) was performed using the ChamQ Universal SYBR qPCR Master Mix kit (Q711-02, Vazyme, China) on a CFX Connect™ Optics Module real-time PCR instrument (Bio-Rad, USA). The total volume of the reaction system was 15 μL. The amplification program was set as follows: pre-denaturation at 95 °C for 30 s; followed by 40 cycles of denaturation (95 °C for 10 s) and annealing/extension (60 °C for 30 s). Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) was used as an internal reference gene for data normalization, and the relative expression of target genes was calculated using the 2^−ΔΔCt^ method. Primer sequences (Table 1) were designed using the Primer-BLAST tool from the National Center for Biotechnology Information (NCBI, www.ncbi.nlm.nih.gov) and synthesized by Beijing Qingke Biotechnology Co., Ltd.Table 1. Real-time quantitative PCR Primer Sequences.Table 1. Forward PrimerReverse PrimerGAPDH5′-GTCTCCTCTGACTTCAACAGCG-3′5′-ACCACCCTGTTGCTGTAGCCAA-3′TLR45′- TTCCGTGGCATTTTTGCTGG-3′5′- ATGCCCTCTGGGATACCTGT-3′IL-1β5′-TTCATTGCTCAAGTGTCTGA-3′5′-TTCATCTGTTTAGGGCCATC-3′TNF-α5′-ACCTCTCTCTAATCAGCCCT-3′5′-TCAGCTTGAGGGTTTGCTAC-3′IL-65′-TTCTCCACAAGCGCCTTC-3′5′-GCGGCTACATCTTTGGAATC-3′

Western blotting

2.13

Caco-2 cells and mouse colonic tissue were lysed on ice for 30 min in RIPA lysis buffer (G2002, Servicebio, China) containing 1% Phenylmethylsulfonyl fluoride (PMSF, G2008, Servicebio, China), then centrifuged at 12 000×g and 4 °C for 15 min, and the supernatant was collected. Protein concentration was measured using a BCA protein assay kit (BL521A, Biosharp, China), and 30 μg of protein was separated by sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS-PAGE) and then transferred to a 0.45 μm polyvinylidene fluoride (PVDF) membrane (Merck Millipore, Germany). The membrane was blocked at room temperature for 2 h in Tris-Buffered Saline with Tween-20 (TBST, G0004, Servicebio, China) containing 5% nonfat milk; primary antibody against the target protein was then added and incubated overnight at 4 °C. After washing three times with TBST buffer, the membrane was incubated with secondary antibody at room temperature for 1 h. Bands were visualized using enhanced chemiluminescence (ECL) substrate (BMU102, Abbkine Scientific, China), images were captured with a ChemiDoc™ XRS Imaging System equipped with a Universal Hood II (Bio-Rad, USA). and band densitometry was performed with ImageJ software. Antibodies for the study were purchased from ABclonal Biotech Co., Ltd and Proteintech Group, Inc, including monoclonal mouse anti-GAPDH (Cat. No. 60004-1-lg), rabbit anti-Claudin-1 (Cat. No. 13050-1-AP), mouse anti-Occludin (Cat. No. 66378-1-lg), rabbit anti-p38 (Cat. No. A4771), polyclonal rabbit anti-p-p38 (Cat. No. AP0526), polyclonal rabbit anti-TLR4 (Cat. No. A17436).

Molecular docking

2.14

Molecular docking of 2′-FL, together with the gut microbial metabolites propionate and butyrate, to the TLR4-MD2 complex was performed using AutoDock Vina (version 1.2.0). The three-dimensional structural coordinates of the macromolecular target TLR4-MD2 (PDB ID: 2Z63) were retrieved from the Protein Data Bank (PDB; https://www.rcsb.org). The structure was visualized and prepared using PyMOL (version 2.5.0), which included the addition of hydrogen atoms and assignment of charges, repair of missing atoms and side chains, optimization of residue conformations, and verification of solvent, ion, and water molecule integrity. The structures of the ligands 2′-FL, propionate, and butyrate were constructed using KingDraw (Education Edition 4.0) and subsequently converted to mol2 format with Open Babel (version 3.1.1). Hydrogen atoms were added, partial charges were assigned, and energy minimization was performed prior to exporting the ligands as PDBQT files for docking.

For docking, a grid box was defined in AutoDock Vina with its center at X = −6.6, Y = −20.6, Z = −22, and dimensions of X = 30 Å, Y = 28 Å, Z = 36 Å. An exhaustiveness parameter of 8 was used, and each ligand was docked 10 times. The conformation with the lowest binding energy was selected, and the resulting binding mode was visualized in PyMOL.

Statistical analysis

2.15

A minimum of three independent experimental replicates were performed, and data are presented as mean ± standard deviation (SD). Visualization was carried out using R (version 4.4.5). Statistical analysis was performed by one-way ANOVA followed by Tukey's multiple comparisons test, with all computations conducted in R. Densitometric quantification of protein bands was performed using ImageJ (version 1.8.0), and differences were considered statistically significant at p < 0.05.

Results

3

Effect of 2′-FL and GOS on pH, bacterial growth and production of SCFAs in in vitro fermented broth

3.1

After a 24-h fermentation, both 2′-FL and GOS significantly promoted the growth of fecal microbiota relative to the control group (p < 0.05). The optical density at 600 nm (OD_600 nm_) of the 2′-FL group was lower than that of the GOS group (Fig. 1A). In addition, both treatments decreased the pH of the fermentation medium (Fig. 1B), although the pH in the 2′-FL group remained significantly higher than in the GOS group (p < 0.05). To evaluate the metabolic output of 2′-FL fermentation, we quantified SCFAs in fermented broth. Fermentation with 2′-FL increased total SCFAs levels, whereas GOS decreased overall SCFAs production in contrast to the control group (Fig. 1C). The major SCFAs identified were acetate, propionate, and butyrate. Compared with the control, 2′-FL significantly increased the concentrations of propionate and butyrate (p < 0.05). In contrast, neither 2′-FL nor GOS markedly altered acetate levels; however, GOS significantly decreased propionate and butyrate levels (Fig. 1D–F).Fig. 1Effect of 2′-FL and GOS on the growth and metabolism of murine fecal microbiota in vitro. (A) OD_600 nm_; (B) pH value; (C) SCFAs; (D-F) acetate, propionate and butyrate; Data were represented as mean ± SD (n = 6). Different lowercase letters above each column indicate significant differences between groups (p < 0.05). 2′-FL: 2′-fucosyllactose; GOS: galacto-oligosaccharides; OD_600 nm_: optical density at 600 nm; SCFAs: short-chain fatty acids. SD: standard deviation.Fig. 1

Effect of 2′-FL and GOS on microbial diversity during in vitro murine fecal fermentation

3.2

To determine the impact of 2′-FL on fecal microbial diversity during in vitro fermentation, we performed comprehensive 16S rRNA gene sequencing across multiple experimental groups. Fig. 2A presents a Venn diagram of amplicon sequence variant (ASV) overlaps among CON, GOS, and 2′-FL groups. While GOS exhibited a markedly higher number of unique ASVs (92) consistent with its known prebiotic activity, the 2′-FL group harbored 13 unique ASVs, indicating a distinct shift in community composition relative to CON. Shared ASVs further reflected both common and group-specific taxa, underscoring compositional alterations induced by 2′-FL. Alpha-diversity metrics are shown in Fig. 2B. Compared with CON, GOS displayed significantly higher Shannon diversity and Simpson evenness, confirming the expected prebiotic effect of GOS. In contrast, 2′-FL did not significantly alter Shannon, Chao1, or ACE indices, although it significantly reduced the Simpson index, suggesting a shift toward a less even community structure. The lack of increase in richness indices implies that 2′-FL modulates diversity differently than GOS, primarily affecting community evenness rather than overall taxonomic breadth. Fig. 2C displays a hierarchical clustering dendrogram based on microbial composition. Samples segregated predominantly by group, with GOS forming a well-separated cluster as anticipated for a strong prebiotic. The 2′-FL group clustered distinctly from CON, albeit with closer proximity than GOS, indicating that 2′-FL induces a recognizable reconfiguration of the microbiota that differs from both CON and the robust GOS response. Fig. 2D shows principal coordinate analysis (PCoA) of microbial community dissimilarities. Notably, the GOS group was markedly separated from CON and 2′-FL, validating the sensitivity of the system to known modulators. Importantly, 2′-FL also differed significantly from CON, demonstrating that it elicits a distinct community shift, albeit of a different magnitude and pattern than GOS. These results highlight that 2′-FL reshapes fecal microbial composition during in vitro fermentation, primarily by reducing community evenness, whereas GOS exerts broader effects on both richness and diversity, serving as an effective reference control.Fig. 2Analysis of alpha and beta diversity of fecal microbiota after fermentation by 2′-FL. (A) Venn diagram; (B) Alpha diversity index; (C) Cluster tree analysis; (D) PCoA. Data were represented as mean ± SD (n = 6). Different lowercase letters above each column indicate significant differences between groups (p < 0.05). 2′-FL: 2′-fucosyllactose; PCoA: principal coordinates analysis; SD: standard deviation.Fig. 2

2′-FL specifically altered fecal microbiota composition during in vitro fermentation

3.3

Fig. 3A and B illustrates the relative abundance of gut microbiota at the phylum level. Compared with the control group, GOS intervention significantly elevates the abundance of Firmicutes while reducing that of Bacteroidetes; conversely, 2′-FL exerts no discernible effect on these dominant phyla. When examining species-level abundance (Fig. 3C and D), 2′-FL markedly enhances the relative abundance of Parabacteroides distasonis, Allobaculum stercoricanis, and Bacteroides acidifaciens, whereas GOS preferentially boosts Lactobacillus johnsonii, Lactobacillus reuteri, and Lactobacillus murinus. Volcano plot analysis (Fig. 3E) further identified Lactobacillus johnsonii and Lactobacillus gasseri taxa significantly upregulated under 2′-FL treatment (p < 0.05). Linear discriminant analysis (LDA) in Fig. 3F reveals that the 2′-FL-enriched microbiota are characterized by signature clades including Tannerellaceae, Parabacteroides distasonis, Allobaculum stercoricanis, and Bacteroides acidifaciens.Fig. 3Effect of 2′-FL on fecal microbiota composition during in vitro fermentation. (A-B) Phylum level; (C-D) Species level; (E) Volcano plot; (F) LDA; (G) Analysis of correlations between metabolites and microbiota; (H) RDA. Data were represented as mean ± SD (n = 6). Different lowercase letters above each column indicate significant differences between groups (p < 0.05). 2′-FL: 2′-fucosyllactose; LDA: linear discriminant analysis; RDA: redundancy analysis; SD: standard deviation.Fig. 3

To elucidate the metabolic implications of these microbial shifts, we correlated the top 20 SCFAs-associated taxa with treatment groups (Fig. 3G). Visualization of the heatmap showed marked associations between Enterococcus, Bacteroides, and Parabacteroides taxa and profiles of SCFAs. Redundancy analysis (RDA; Fig. 3H) indicated that treatment groups explain 78.45% (RDA1) and 11.73% (RDA2) of microbial variance. Specifically, 2′-FL supplementation correlated with elevated concentrations of propionate, butyrate, and isobutyrate, whereas GOS primarily modulated acetate levels.

2′-FL and its metabolites propionate and butyrate enhance intestinal barrier function and alleviates intestinal inflammation in Caco-2 cells

3.4

ETEC infection significantly impaired intestinal barrier integrity, as evidenced by reduced protein expression of Occludin and Claudin-1 in Western blotting (Fig. 4A–E). However, supplementation with 2′-FL reversed ETEC-induced Occludin and Claudin-1 downregulation (both p < 0.05), while GOS had no significant effect on the expression of either Occludin or Claudin-1. Similarly, both propionate and butyrate mitigated ETEC-induced tight-junction disruption: butyrate elevated both Occludin and Claudin-1 levels (p < 0.05), whereas Propionate selectively increased Occludin (p < 0.05) but not Claudin-1.Fig. 4Effect of 2′-FL and its metabolites propionate/butyrate on the cellular barrier function and inflammatory response of Caco-2 cells challenged by ETEC. (A, E) Analysis of Occludin and Claudin-1 expression levels via western blotting; (B-D, F-H) Analysis of IL-6, TNF-α, and IL-1β expression levels via RT-qPCR. Data were represented as mean ± SD (n = 3). Different lowercase letters above each column indicate significant differences between groups (p < 0.05). PA: propionic acid; BA: butyric acid; 2′-FL: 2′-fucosyllactose; Caco-2 cells: human colorectal adenocarcinoma cells; ETEC: enterotoxigenic Escherichia coli; IL-6: interleukin-6; TNF-α: tumor necrosis factor-alpha; IL-1β: interleukin-1 beta; RT-qPCR: reverse transcription quantitative polymerase chain reaction; SD: standard deviation.Fig. 4

Regarding inflammation, ETEC infection markedly upregulated pro-inflammatory cytokines (IL-6, TNF-α, and IL-1β) at the mRNA level (Fig. 4B–D, F–H). Intervention with 2′-FL and GOS significantly attenuated this inflammatory response (p < 0.05). Consistently, propionate and butyrate also suppressed ETEC-induced inflammation (p < 0.05).

Effect of 2′-FL and its metabolites propionate/butyrate on the alleviation of oxidative stress injury in Caco-2 cells challenged by ETEC

3.5

To investigate the effects of 2′-FL and its potential metabolites, propionate and butyrate, on ETEC-induced oxidative stress in Caco-2 cells, we used GOS as a reference control. Intracellular ROS levels, ABTS radical-scavenging capacity, SOD activity, and MDA content were systematically evaluated. Compared with the control group, ETEC infection led to significant changes in oxidative stress-related indices. Specifically, intracellular ROS levels were markedly elevated, while ABTS scavenging capacity and SOD activity were decreased, and MDA levels were increased (p < 0.05). Treatment with 2′-FL or GOS significantly reversed these oxidative changes. As shown in Fig. 5A–E, they restored the redox balance. Similarly, both propionate and butyrate alleviated ETEC-evoked oxidative injury. They suppressed ROS overproduction (Fig. 5F and G), enhanced ABTS radical-scavenging capacity (Fig. 5H), increased SOD activity (Fig. 5I), and reduced MDA accumulation (Fig. 5J) (p < 0.05). These results suggest that 2′-FL, along with its metabolites propionate and butyrate, may have protective effects against ETEC-induced oxidative stress in Caco-2 cells (p < 0.05).Fig. 5Effect of 2′-FL and its metabolites propionate/butyrate on alleviating ETEC-induced oxidative stress in Caco-2 cells. (A, F) Intracellular ROS levels by flow cytometry; (B-E, G-J) ABTS radical scavenging ability, MDA content, and SOD activity analysis. Data were represented as mean ± SD (n = 3). Different lowercase letters above each column indicate significant differences between groups (p < 0.05). PA: propionic acid; BA: butyric acid; 2′-FL: 2′-fucosyllactose; ETEC: enterotoxigenic Escherichia coli; Caco-2 cells: human colorectal adenocarcinoma cells; ROS: reactive oxygen species; ABTS; 2,2′-Azino-bis (3-ethylbenzothiazoline-6-sulfonic acid); MDA, malondialdehyde; SOD: superoxide dismutase; SD: standard deviation.Fig. 5

2′-FL and its metabolites propionate and butyrate regulated TLR4 expression to inhibit the p38 MAPK signaling pathway in Caco-2 cells

3.6

To define the anti-inflammatory mechanisms of 2′-FL, we further employed Western blotting and RT-qPCR to analyze associated signaling pathways in the Caco-2 cells. ETEC infection significantly increased the protein expression of phosphorylated p38 (p-p38) and TLR4 in Caco-2 cells; however, treatment with 2′-FL effectively reversed this upregulation (Fig. 6A). Consistent with these protein-level changes, RT-qPCR analysis revealed that ETEC infection elevated TLR4 mRNA expression, a response that was similarly mitigated by 2′-FL supplementation (Fig. 6B). Beyond 2′-FL, its metabolites propionate and butyrate also suppressed the ETEC-induced overexpression of TLR4 and p-p38 at the protein level (Fig. 6C and D).Fig. 6Effect of 2′-FL and its metabolites propionate/butyrate on regulating TLR4/p38 MAPK signaling in ETEC-infected Caco-2 cells. (A, C) Western blotting analysis of TLR4 and p38 MAPK expression levels; (B, D) RT-qPCR analysis of TLR4 expression levels. Data were represented as mean ± SD (n = 3). Different lowercase letters above each column indicate significant differences between groups (p < 0.05). PA: propionic acid; BA: butyric acid; 2′-FL: 2′-fucosyllactose; ETEC: enterotoxigenic Escherichia coli; Caco-2 cells: human colorectal adenocarcinoma cells; TLR4: toll-like receptor 4; p38 MAPK: p38 mitogen-activated protein kinase; RT-qPCR: reverse transcription quantitative polymerase chain reaction; SD: standard deviation.Fig. 6. Fig. 7Effect of 2′-FL on the mitigation of DSS-induced colitis of mice. (A) Animal experimental design; (B) Mice daily weight gain (n = 10); (C-D) Colon length and statistics (n = 6); (E) DAI score (n = 6); (F) H&E and PAS staining (n = 6); (G) Histological Score (n = 6); (H-I) Analysis of relative ZO-1 and Claudin-1 expression in colon tissue by indirect immunofluorescence assay (n = 3); (J-L) Detection of IL-6, IL-8 and TNF-α, expression levels in mouse serum by ELISA (n = 3). Data were represented as mean ± SD. Different lowercase letters above each column indicate significant differences between groups (p < 0.05). 2′-FL: 2′-fucosyllactose; DSS: dextran sulfate sodium; DAI: disease activity index; H&E: hematoxylin-eosin; PAS: periodic acid-Schiff; ZO-1: zonula occludens-1; IL-6: interleukin-6; IL-8: interleukin-8; TNF-α: tumor necrosis factor-alpha; ELISA: enzyme-linked immunosorbent assay; SD: standard deviation.Fig. 7

Confirmation of 2′-FL's protective role in attenuating DSS-induced colitis symptoms in mice

3.7

The mice in the DSS-treated group exhibited a marked decrease in body weight, and this reduction was statistically significant (p < 0.05) when compared with the control group (CON). Notably, supplementation with both GOS and 2′-FL effectively attenuated this weight loss, as evidenced by the statistical significance (p < 0.05) (Fig. 7B). Subsequent to DSS administration, a significant shortening of the colon length was observed, along with the presence of edema (p < 0.05). Concurrently, the DAI score demonstrated a substantial increase (p < 0.05). In contrast, the administration of either 2′-FL or GOS led to the restoration of colon length, amelioration of edema, and a significant reduction in the DAI score (Fig. 7C–E; p < 0.05). Histological examinations, utilizing H&E and PAS staining of mouse colon tissues, revealed that both 2′-FL and GOS treatments significantly mitigated inflammatory cell infiltration, mucosal damage, and glycogen depletion (Fig. 7F). Quantitative histological scoring further confirmed that DSS challenge induced a dramatic elevation in colon histological scores (p < 0.05), while treatment with GOS or 2′-FL significantly reduced the histological damage scores (Fig. 7G; p < 0.05).

Immunofluorescence staining results indicated that the DSS challenge led to a significant downregulation of the levels of ZO-1 and claudin-1 (p < 0.05). While 2′-FL treatment restored the levels of these two tight-junction proteins to near - control levels (p < 0.05), the GOS treatment did not yield a significant restorative effect on their expression (Fig. 7H and I). Analysis of serum cytokines demonstrated that both 2′-FL and GOS significantly attenuated the elevation of pro-inflammatory cytokines, namely TNF-α, IL-8, and IL-6, in mice with colitis (Fig. 7J–L; p < 0.05). These findings confirm that both 2′-FL and GOS protect against DSS-induced colitis, with 2′-FL more effectively restoring tight-junction protein levels than GOS.

2′-FL alleviates colitis by specifically enriching fecal propionate and butyrate and inhibiting the TLR4/p38 MAPK pathway

3.8

To further explore the mechanistic basis by which 2′-FL ameliorates DSS-induced colitis in mice, we conducted comprehensive analyses of SCFAs levels in murine fecal samples and the activation states of relevant signaling pathways. Our findings demonstrated that 2′-FL selectively enhanced the fecal concentrations of propionate and butyrate, while exerting no significant influence on acetate levels (Fig. 8A–D; p < 0.05). Concurrently, 2′-FL treatment led to a marked downregulation of TLR4 expression within the colonic tissues of DSS-exposed mice, which in turn attenuated the activation of the p38 MAPK signaling cascade (Fig. 8E; p < 0.05).Fig. 82′-FL alleviates colitis by specifically enriching propionate and butyrate and inhibiting the TLR4/p38 MAPK pathway. (A-D) SCFAs content in mice feces (n = 6); (E) Western blotting analysis of TLR4 and p38 MAPK expression levels (n = 3). Data were represented as mean. Different lowercase letters above each column indicate significant differences between groups (p < 0.05). 2′-FL: 2′-fucosyllactose; TLR4: toll-like receptor 4; p38 MAPK: p38 mitogen-activated protein kinase; SD: standard deviation.Fig. 8

Molecular docking of 2′-FL, propionate and butyrate with TLR4-MD2

3.9

To confirm whether 2′-FL can directly bind to TLR4, we performed molecular docking simulations of 2′-FL with the TLR4-MD2 complex using AutoDock Vina. The results revealed that 2′-FL binds to the TLR4-MD2 complex with an affinity of −6.60 kcal mol^−1^. The sugar hydroxyl groups of 2′-FL form direct hydrogen bonds with Thr319, Tyr292, and Ser118. Additionally, the side-chain amino group of Lys362 forms an extra hydrogen bond with a sugar oxygen atom. This synergistic network of hydrogen bonds firmly anchors 2′-FL within the MD-2 pocket (Fig. 9A).Fig. 9Molecular docking of 2′-FL and its metabolites propionate/butyrate with the TLR4-MD2 complex. Molecular docking of 2′-FL(A), propionate and butyrate (B-C) to the TLR4-MD2 complex as illustrated in the 3D docking visualization. 2′-FL: 2′-fucosyllactose; TLR4-MD2: toll-like receptor 4-myeloid differentiation factor 2; 3D: three-dimensional.Fig. 9

Furthermore, we carried out additional molecular docking studies to examine the interactions of propionate and butyrate with the TLR4-MD2 complex. The binding affinities of propionate and butyrate were −3.56 kcal/mol and −4.00 kcal/mol, respectively (Fig. 9B and C). Propionate forms a single hydrogen bond between its carboxylate oxygen and Lys341. In contrast, butyrate forms hydrogen bonds with both Thr319 and Lys341. This increase in the number of interaction sites enhances the binding stability of butyrate compared to propionate.

Discussion

4

Although the anti-inflammatory properties of 2′-FL have been extensively investigated both in vitro and in vivo (Li et al., 2024a; Lv et al., 2023; Pak et al., 2023), a comprehensive mechanistic framework encompassing both its direct effects and the indirect contributions of its microbial metabolites remains incompletely elucidated. This study provides a holistic elucidation of the anti-inflammatory mechanisms of 2′-FL, demonstrating a synergistic interplay between its direct receptor-mediated actions and its indirect effects mediated through gut microbial metabolites.

For a direct comparison of 2′-FL and GOS fermentation, we used a controlled murine fecal fermentation model (Kang et al., 2022; Liu et al., 2021), as it offers a more standardized platform than the variable human fecal inocula, which remain the gold standard for general HMO studies (Bajic et al., 2023). This model reduces inter-individual variability, avoids ethical constraints, and provides stable, comparable microbial communities (Zhou et al., 2025), offering a robust platform to delineate mechanistic differences between the two prebiotics. Meanwhile, to facilitate a comparative investigation with 2′-FL, we employed GOS as a reference prebiotic. This choice was informed by GOS's status as a widely accepted and extensively studied prebiotic, with a robust body of evidence demonstrating its positive effects on gut health (Endika et al., 2025; Salli et al., 2019).

Our in vitro fermentation analyses demonstrated that 2′-FL not only functioned as a metabolic substrate for gut microbiota but also significantly modulated the microbial metabolic landscape. A key finding was the 2′-FL-induced augmentation of specific short-chain fatty acids (SCFAs), particularly propionate and butyrate, whereas GOS did not elicit this selective enhancement. This points to a distinct metabolic fate and functional outcome of 2′-FL in comparison to GOS. SCFAs—including acetate, propionate, and butyrate—are pivotal microbial metabolites derived from carbohydrate fermentation, playing critical roles in regulating energy metabolism, immune responses, and overall metabolic health (Hays et al., 2024; Kimura et al., 2020). Consistent with prior reports (Chen et al., 2022; Ge et al., 2024), the 2′-FL-mediated increase in propionate and butyrate levels is likely driven by specific compositional changes in the gut microbiota, potentially underlying these metabolic alterations. In the study, supplementation with GOS unexpectedly reduced the total SCFA pool, particularly propionate and butyrate. This effect may be attributed to the selective stimulation of lactic acid bacteria (LAB) by GOS, whose metabolism favors lactic acid production over direct SCFA synthesis. Accumulation of lactic acid subsequently resulted in a lowered pH environment (Punia Bangar et al., 2022; Nuryana et al., 2019), and a decline in pH to approximately 5.0 has been shown to inhibit the enzymatic activities responsible for propionate and butyrate formation (Belenguer et al., 2007). Compared with 2′-FL, GOS appears to promote a lactate-fermentation pathway rather than acetate/propionate/butyrate production in certain microbial contexts (Louis and Flint, 2017). As a result, the SCFA producers could not efficiently harness the available lactate, leading to the observed decrease in SCFA levels.

At the species level, 2′-FL selectively enriched taxa with established roles in SCFA biosynthesis, including Parabacteroides distasonis, Bacteroides acidifaciens, and Allobaculum stercoricanis. These latter two species have been previously implicated in propionate and butyrate production (Jacobson et al., 2018; Wang et al., 2024), and our redundancy analysis confirmed a positive correlation between their abundance and elevated fecal levels of these SCFAs. These results substantiate a functional linkage between the 2′-FL-modulated microbiota and its metabolic output, positioning these taxa as potential keystone regulators of the metabolic benefits associated with 2′-FL consumption. In contrast, GOS preferentially stimulated the growth of lactic acid-producing Lactobacillus species (L. johnsonii, L. reuteri, L. murinus).

Collectively, these findings of in vitro fermentation underscore that the prebiotic efficacy of 2′-FL is underpinned by a distinctive modulation of gut microbial communities, which markedly differs from the broader ecological alterations induced by GOS. While GOS robustly enhanced microbial alpha diversity and substantially altered community structure—particularly through the proliferation of Lactobacillus spp.—2′-FL exhibited a more selective mode of action. It primarily influenced community evenness without significantly altering overall microbial abundance or the relative composition of dominant phyla such as Firmicutes and Bacteroidetes. This pattern is consistent with previous observations (Ge et al., 2024; Renwick et al., 2025; Salli et al., 2021), suggesting that 2′-FL confers prebiotic benefits while minimally disrupting the foundational architecture of the gut microbiota. Such selectivity may confer advantages in terms of long-term safety and the stability of host–microbe interactions. Although both prebiotics benefit gut microbial ecology, they act through divergent mechanisms. GOS induced broad-spectrum proliferation of specific genera like Lactobacillus, leading to lactic acid-dominant metabolic profiles. 2′-FL executed nuanced modulation by enhancing community evenness and enriching SCFA-producing species, resulting in elevated propionate and butyrate levels. This differential modulation likely underlies their distinct impacts on intestinal homeostasis, with 2′-FL's ability to promote beneficial SCFAs while preserving microbial structural integrity reflecting a targeted, functionally oriented prebiotic activity. Such selectivity may offer advantages in modulating host metabolism and immune responses with a reduced risk of dysbiosis, highlighting the need for further research to delineate the molecular pathways through which these microbial shifts translate into systemic health outcomes.

The functional relevance of 2′-FL and its microbial metabolites was further substantiated using in vitro cellular models. It's noted that the ETEC-infected Caco-2 cell model was employed to investigate intestinal infection and host immune responses in this study. This in vitro system is widely utilized for evaluating the prebiotic activities of various bioactive compounds and elucidating their underlying physiological mechanisms (Lu et al., 2023; Xie et al., 2025; Zhou et al., 2025). ETEC infection induces intestinal epithelial inflammation, typically accompanied by disruption of the epithelial barrier function and elevated oxidative stress (Jin et al., 2024); consequently, this model provides a relevant platform to assess potential mitigation strategies. Using the ETEC-challenged Caco-2 model, we examined the protective effects of GOS, 2′-FL, and their microbial SCFA metabolites against the pathological sequelae of infection. Both 2′-FL and its principal SCFA metabolites, propionate and butyrate, conferred significant protection against ETEC-induced cellular damage. Treatment with 2′-FL and its metabolites propionate/butyrate markedly restored epithelial barrier integrity, as evidenced by upregulation of tight junction proteins, and suppressed secretion of pro-inflammatory cytokines. In contrast, GOS intervention did not modulate tight junction protein expression but selectively attenuated the ETEC-driven elevation in pro-inflammatory cytokines, indicating a distinct immunoregulatory profile that diverges from the dual-functional attributes of 2′-FL. Subsequent experiments demonstrated that propionate and butyrate alone were sufficient to ameliorate barrier dysfunction and reduce inflammation in the ETEC model, thereby establishing a plausible mechanistic link between the microbiota shifts induced by 2′-FL and its prebiotic functionality.

To investigate whether 2′-FL and its metabolites, propionate and butyrate, individually contribute to mitigating oxidative stress, we assessed oxidative stress markers in Caco-2 cells challenged with ETEC. This analysis confirmed that both 2′-FL and its metabolites (propionate/butyrate) alleviated oxidative stress injury. Further support for this metabolite-mediated protective effect came from assessments of cellular antioxidant capacity, which demonstrated their collective role in enhancing cellular defense against oxidative damage. This result aligns with the established role of SCFAs in maintaining intestinal homeostasis (Kim, 2023), and it also highlights a distinctive feature of 2′-FL: its ability to preserve gut health through a dual-pathway strategy—directly modulating host cells and indirectly regulating via microbial-derived metabolites. Such an integrated approach may confer greater resilience to the intestinal ecosystem compared to prebiotics acting primarily through a single mechanism.

A central finding of this study is the identification of the TLR4/p38 MAPK signaling pathway as a convergent mechanistic node mediating the protective effects of both 2′-FL and its microbial metabolites. As a pivotal hub in intestinal inflammatory responses, the TLR4/p38 MAPK axis orchestrates the transcriptional upregulation of pro-inflammatory cytokines and induces redox imbalance, ultimately compromising epithelial barrier integrity (Antunes et al., 2023). Our results show that pretreatment with 2′-FL markedly inhibited ETEC-driven overexpression of TLR4 and abrogated downstream phosphorylation and activation of p38 MAPK, thereby attenuating the propagation of inflammatory signals. This observation is corroborated by independent studies reporting similar inhibitory effects of 2′-FL on the TLR4/p38 MAPK pathway (Chen et al., 2025b). Importantly, propionate and butyrate recapitulated this inhibitory effect, reducing both receptor expression and kinase activation to an extent comparable to that observed with the parent prebiotic.

The in vitro inhibition of TLR4/p38 MAPK signaling by 2′-FL and its metabolites was faithfully recapitulated in vivo in a DSS-induced murine colitis model. The results of the mouse study demonstrated that oral administration of either 2′-FL or GOS ameliorated the severity of colitis. A distinctive feature of 2′-FL, relative to GOS, is its capacity to promote intestinal barrier repair, as demonstrated by a marked upregulation of tight junction proteins in the colonic mucosa of colitic animals. This phenomenon, reported in earlier investigations (Liu et al., 2022a), is likely attributable to the ability of 2′-FL to selectively enhance microbial production of propionate and butyrate. Butyrate, in particular, serves as the principal energy source for colonic epithelial cells, driving ATP synthesis necessary for the biosynthesis and functional maintenance of tight junction complexes, thereby directly reinforcing barrier integrity (Chen et al., 2024). In support of this mechanism, we detected elevated concentrations of propionate and butyrate in colonic contents following 2′-FL supplementation—a metabolic shift consistent with documented alterations in SCFA profiles induced by 2′-FL (Ge et al., 2024) and congruent with our prior in vitro fermentation data.

Crucially, the observed protective effects were accompanied by parallel increases in fecal propionate and butyrate levels and concurrent downregulation of the TLR4/p38 MAPK pathway in colonic tissue. Beyond its metabolic actions, 2′-FL suppressed hyperactivation of this signaling cascade in the inflamed colon, mirroring the inhibitory effect established in our cellular infection model. The convergence of these two complementary mechanisms—targeted attenuation of inflammatory signaling and SCFA-dependent reinforcement of epithelial barrier function—suggests a synergistic mode of action through which 2′-FL mitigates experimental colitis. Such dual-pathway engagement may explain the superior intestinal barrier preservation afforded by 2′-FL compared to prebiotics operating predominantly via a single regulatory axis. It's noted that GOS was initially included as a positive control to benchmark the anti-inflammatory efficacy of 2′-FL. However, as the primary objective of this study was to elucidate the mechanism of action of 2′-FL, GOS was subsequently excluded from all mechanistic analyses.

Molecular docking analyses revealed that 2′-FL binds favorably to the TLR4-MD2 complex (−6.60 kcal/mol), forming hydrogen bonds with key residues (Thr319, Lys362) in a manner predicted to interfere with receptor dimerization and MyD88-dependent signaling. This structural antagonism is consistent with the reduced TLR4 overexpression and p38 MAPK activation observed in ETEC-challenged Caco-2 cells and DSS-induced colitis mice. Intriguingly, microbial metabolites propionate and butyrate were also found to dock into the same pocket, with butyrate displaying even greater binding affinity, likely due to enhanced van der Waals contacts and additional polar interactions with adjacent residues. These results suggest that both 2′-FL and its microbial catabolites can act as molecular antagonists of TLR4, thereby obstructing the initial step of innate immune activation triggered by pathogen-associated molecular patterns (PAMPs).

Although previous studies have suggested analogous mechanisms for oligosaccharides in general (Na et al., 2022), the present work provides the first direct evidence that 2′-FL and its metabolic products can act in a TLR4 receptor-dependent manner to mitigate intestinal inflammation. This dual-action mechanism helps explain the potent anti-inflammatory effects of 2′-FL observed both herein and in earlier reports (Li et al., 2022; Zhang et al., 2025b). Of particular significance is the selective enrichment of propionate and butyrate induced by 2′-FL, as these SCFAs not only strengthen intestinal barrier function (Wang et al., 2022) but also act as signaling molecules that modulate TLR4 activity. This finding broadens our understanding of how prebiotics can influence host physiology via both direct receptor modulation and microbiota-mediated pathways.

It is important to note, however, that the evidence supporting a direct mechanistic role for TLR4 in this process is currently correlative. Definitively assigning a causal role to this receptor will require future experimentation with TLR4-specific agonists, inhibitors, or genetic ablation models. Furthermore, the Caco-2 model used in this study, selected for its robustness in modeling epithelial TLR4 signaling, does not mirror the complete innate immune landscape orchestrated by professional immune cells. To directly address these limitations, future investigations ought to integrate gene editing technologies, organoid models, and clinical sample analyses. This multi-pronged approach will further refine the theoretical framework of 2′-FL's anti-inflammatory mechanisms and provide more robust experimental substantiation for its therapeutic application.

In summary, we identify a novel dual-pathway mechanism whereby 2′-FL exerts its anti-inflammatory effects: functioning dually in a TLR4 receptor-dependent manner and as a prebiotic that selectively augments anti-inflammatory metabolite generation. These pathways converge on the TLR4/p38 MAPK axis, thereby suppressing inflammatory responses. This finding significantly advances our knowledge of HMO bioactivity and establishes a strong rationale for exploring 2′-FL's therapeutic utility in inflammatory gut disorders.

Author contributions

Kai Na: Conceptualization; methodology; Writing—original draft. Cheng Chao: Methodology; Validation; Formal analysis. Tianfei Yu: Software; Formal analysis. Liqun Wang: Validation; Formal analysis. Xiaotong Liu: Validation; Investigation. Li Zhang: Visualization; Supervision. Kejue Feng: Data Curation; Project administration. Junhua Lei: Validation; Formal analysis. Yu Xiong: Investigation; Formal analysis. Xiaohua Guo: Conceptualization; Methodology; Writing - Review & Editing; Funding acquisition. All authors have read and agreed to the published version of the manuscript.

Ethics approval statement

Ethical approval for all animal procedures detailed herein was granted by the Scientific Ethics and Safety Committee of South Central Minzu University (No. 2024 SCUEC 034, Wuhan, China). Animal experiments in this study complied with international guiding principles for biomedical research involving animals.

Notes

The authors declare no conflict of interest.

Funding

This work was supported by the Fund for the Hubei Province Engineering Master's and PhD Joint Industry-University Training Practice Project (grant Number: HZY25113; recipient: Kai Na); the Hubei Province Undergraduate New Engineering Practice Base Construction Project (grant Number: XGK04019); the Academic Innovation Teams of South-Central Minzu University (grant Number: XTZ24023); and the Fundamental Research Funds for the Central Universities, South-Central Minzu University (grant Number: CZQ24013).

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.