Hyperoside, a dietary flavonoid, protects against endometritis via gut microbiota-dependent production of hydroxyphenyllactic acid and the gut–uterus axis

Abstract

Endometritis, primarily caused by Escherichia coli (E. coli) infection, poses significant therapeutic challenges due to rising antibiotic resistance. The associated pro-inflammatory cytokines cause persistent endometrial damage, thereby leading to infertility, pregnancy loss, and other gynecological complications, which impose substantial long-term medical and socioeconomic burdens. Hyperoside, a flavonol glycoside abundant in various common fruits (e.g., hawthorn) and vegetables, exhibits significant anti-inflammatory activity, highlighting its potential as a functional food or nutraceutical. Our present study firstly demonstrated that hyperoside could alleviate E. coli-induced endometritis in mice through a gut–uterus axis mechanism. Specifically, hyperoside remodeled the gut microbiota by enriching beneficial genera, such as Lactobacillus and Prevotella, which subsequently elevated the production of the metabolite hydroxyphenyllactic acid (HPLA). Crucially, antibiotic treatment and fecal microbiota transplantation (FMT) experiments further confirmed that gut microbiota restructuring was essential for the anti-endometritic effect of hyperoside. Mechanistically, HPLA enters systemic circulation and targets uterine tissue, where it is directly bound to TLR4 to suppress the activation of the TLR4/NF-κB pathway and then the release of inflammatory cytokines. The present study provides the first systematic evidence of the gut–uterus axis, establishing microbiota-derived HPLA as a key effector against E. coli-induced endometritis, offering a novel nutritional intervention strategy for inflammatory reproductive disorders.

---

Introduction

Endometritis, a prevalent inflammatory disorder of the reproductive system during the perinatal period, causes reproductive dysfunction in both humans and animals, resulting in huge economic losses.^1,2 Endometritis is characterized by persistent inflammatory responses and compromised barrier integrity in endometrial tissues, which is primarily triggered by pathogenic microorganism infection.^3,4Escherichia coli (E. coli) is recognized as one of the most prevalent pathogenic bacteria in uterine diseases.⁵ Following E. coli invasion, the bacterial outer membrane component lipopolysaccharide (LPS) is recognized by toll-like receptor 4 (TLR4), which is a critical innate immune receptor expressed in endometrial epithelial cells.⁶ This binding induces TLR4 dimerization and recruits the adaptor protein MyD88, initiating a downstream signaling cascade that culminates in NF-κB activation.⁷ The transcriptionally active NF-κB then translocates to the nucleus and upregulates the expression of pro-inflammatory cytokines, including TNF-α, IL-1β and IL-6, thereby triggering an inflammatory response.⁸ Therefore, controlling the activation of TLR4 signaling can effectively suppress the inflammatory cascade and alleviate endometrial inflammatory damage.

The gut microbiota, as a critical regulator of host immune homeostasis, plays a crucial role in the pathogenesis of various diseases.⁹ The imbalance of the gut microbiota promotes the development of diverse pathological processes by disrupting the integrity of the intestinal barrier and inducing bacterial translocation, which elicits sustained systemic inflammation.¹⁰ Recent studies have revealed that the gut microbiota not only affects the local intestinal microenvironment through metabolites and immunomodulatory signaling pathways, but also regulates the inflammatory response and metabolic balance in distal organs via a multi-organ crosstalk network. The gut microbiota has been found to modulate the level of LPS and secondary bile acids via the gut–liver axis, thereby playing a crucial role in triclosan-induced liver injury.¹¹ A recent report from Zhang Q. et al.¹² illustrated that the severity of influenza infection was associated with heterogeneous responses of the gut microbiota, and endogenous B. animalis can enhance resistance against lethal influenza infection. In addition, it has been demonstrated in endometritis models that the gut microbiota and metabolites exhibit a protective effect against pathogenic bacterial infections through activation of anti-inflammatory factors.^13–15

The flavonol glycoside hyperoside (Hyp), also known as quercetin-3-O-β-D-galactopyranoside, is abundant in various common fruits (such as hawthorn), vegetables, beverages, and other edible plants.^16,17Hypericum perforatum L. (St John's wort), which is rich in hyperoside, has been thoroughly tested and is commonly used in the form of oils, infusions, or crucial dietary supplements.¹⁸ It is recognized for its multiple health-beneficial properties, including potent anti-oxidant and anti-inflammatory activities.^17,19 In LPS-stimulated murine macrophages, hyperoside significantly reduces pro-inflammatory cytokine secretion by inhibiting NF-κB activation.²⁰ Another study also indicates that hyperoside exerts a protective effect against LPS-induced inflammatory responses in microglial cells by regulating the p38 and NF-κB pathways.²¹ Notably, hyperoside has also been demonstrated to modulate microRNA to mitigate blood–brain barrier damage induced by Streptococcus, another important pathogen implicated in endometritis.^22,23 Furthermore, hyperoside has been shown to ameliorate depressive-like behaviors in mice by modulating the gut–brain axis through regulation of the gut microbiota composition and short-chain fatty acid levels.²⁴ However, the therapeutic potential and underlying mechanisms of hyperoside in endometritis remain to be elucidated.

Given the crucial function of the gut microbiota and metabolites in inflammatory regulation, alongside the proven impact of hyperoside on the gut microbiome and inflammation, we propose a scientific hypothesis that its protective action against E. coli-induced endometritis is mediated through the gut–uterine axis. Our findings demonstrate that hyperoside can significantly remodel the compositions of the gut microbiota and promote the production of the gut microbiota-derived metabolite hydroxyphenyllactic acid (HPLA), thereby effectively alleviating E. coli induced-uterine inflammation by inhibiting the activation of the TLR4/NF-κB pathway. This study provides novel insights into the mechanism of hyperoside against endometritis and offers a new intervention strategy and potential drug candidate for endometritis treatment.

Materials and methods

Animals and strain

7-Week-old female ICR mice were purchased from SPF Biotechnology Co., Ltd (Beijing, China). After a one-week acclimatization period at room temperature (20–24 °C) with a 12 h dark–night cycle, the mice were subjected to experiments. All experimental procedures involving animals were approved by the Life Scientific Ethic Committee of Yunnan Agricultural University (Approval number: APYNAU202207009). Escherichia coli O111, generously gifted by Prof. Yaohong Zhu's lab (China Agricultural University), was inoculated into Luria–Bertani (LB) broth medium and cultured at 37 °C and 180 rpm for 12 h.

Endometritis model establishment and treatment

A mouse model of E. coli-induced endometritis was established as previously described.²⁵ Briefly, both uterine horns of specific pathogen-free (SPF) mice were infused with 1 × 10⁸ CFU mL⁻¹ of E. coli (30 μL per side) using a 100 μL syringe equipped with a 30-gauge blunt needle after anesthesia. Twenty-four hours post E. coli or PBS treatment, the uterine cervix and fallopian tubes were carefully transected under aseptic conditions. The entire uterine body along with bilateral uterine horns were harvested and immediately snap-frozen in a −80 °C freezer to preserve tissue integrity for subsequent analysis.

In the study of hyperoside treatment, the mice were randomly divided into four groups with six mice per group: control group, E. coli group, hyp group and E. coli + hyp (80 mg kg⁻¹) group. Hyperoside (purity > 95%, HPLC; Yuanye Bio-Technology Co., Ltd, China, Fig. S1) was solubilized using dimethyl sulfoxide (DMSO, Sigma, USA) to obtain a final concentration of 80 mg kg⁻¹. Twenty-four hours after E. coli infection, the mice were orally administered with hyperoside for 3 days. The uterine tissues, feces, and serum samples were harvested and used for further assays.

In the study of HPLA treatment, the mice were randomly divided into three groups with six mice per group and were orally administered with 20 mg per kg of body weight of HPLA (purity > 95%, HPLC; Yuanye Bio-Technology Co., Ltd, China) or equal phosphate-buffered saline (PBS) for 3 days. Subsequently, the mice were challenged with E. coli on day 4 to induce endometritis. The treatment groups were as follows: (1) control group: oral gavage of PBS for 3 days without infection; (2) E. coli group: oral gavage of PBS for 3 days with infection; and (3) HPLA + E. coli group: oral gavage of HPLA for 3 days with infection. The mice were euthanized on day 5, and the uterine and colon tissues were harvested for subsequent analysis.

In the study on the therapeutic effect of HPLA rather than that of hyperoside, mice pretreated with an antibiotic cocktail (ATB) were randomly divided into five groups with six mice per group and then were orally administered with hyperoside with or without HPLA supplementation for 3 days. The treatment groups were as follows: (1) ATB + Hyp group: oral gavage of hyperoside for 3 days without infection; (2) ATB + Hyp + HPLA group: oral gavage of hyperoside and HPLA for 3 days without infection; (3) ATB + E. coli group: oral gavage of PBS for 3 days with infection; (4) ATB + Hyp + E. coli group: oral gavage of hyperoside for 3 days with infection; and (5) ATB + Hyp + HPLA + E. coli group: oral gavage of hyperoside and HPLA for 3 days with infection. The mice were euthanized on day 8, and the uterine and colon tissues were harvested for subsequent analysis.

Antibiotic treatment

SPF mice underwent an ATB treatment as previously described.²⁶ In brief, animals received sterile drinking water containing an antibiotic cocktail, which consisted of ampicillin (1 mg mL⁻¹), streptomycin (5 mg mL⁻¹), vancomycin (0.25 mg mL⁻¹), and colistin (1 mg mL⁻¹). The ATB treatment was maintained for a duration of 3 days and discontinued 24 hours prior to fecal microbiota transplantation (FMT) or HPLA treatment.

FMT experiments

Feces from the control, E. coli, and hyperoside + E. coli groups were collected to prepare the FMT inoculum and then were processed as previously described.²⁷ Briefly, fresh fecal samples were immediately homogenized under sterile conditions with an N₂ gas within 2 hours of collection. Subsequently, the homogenates were filtered through 0.25 mm stainless steel cell strainers and centrifuged at 6000g for 15 min to remove particulate impurities. Prior to FMT, viable bacteria in the suspension were quantified using optical microscopy with methylene blue staining.²⁸ Antibiotic-pretreated SPF mice were administered with 100 μL of a suspension containing 1 × 10⁸ bacteria or PBS via oral gavage to perform FMT. One day after FMT, the mice were challenged with E. coli infection, and the treatment groups were as follows: (1) FMT-control group: FMT of feces from mice treated with PBS; (2) FMT-E. coli group: FMT of feces from mice infected with E. coli; and (3) FMT-hyp + E. coli group: FMT of feces from hyperoside-treated mice with infection. Then, the uterine tissues and feces were collected at 24 hours post-infection for subsequent analysis.

Histological analysis

The tissue samples were fixed in 4% paraformaldehyde, sequentially dehydrated in an ethanol gradient, and paraffin-embedded. Subsequently, sections were prepared and stained with hematoxylin and eosin (H&E) for histological examination. Histopathological alterations were evaluated under an optical microscope (Olympus, Japan) by two independent, blinded investigators using a previously established scoring system.^29,30 Briefly, the uterine tissue was scored according to the degree of damage such as the integrity of the tissue structure and the number of infiltrating inflammatory cells, and the colon tissue was assessed based on the severity of inflammatory infiltration (graded 0–3) and the extent of crypt damage (graded 0–4). The final score for each section was calculated by multiplying the severity grade by the percentage of affected tissue area.

Immunofluorescence

Immunofluorescence staining was performed on both cells and tissue sections following optimized protocols. For cell staining, cells were seeded in six-well chamber slides at 30–50% confluency, and then washed with PBS and fixed with 4% paraformaldehyde for 15 minutes. After permeabilization with 0.2% Triton X-100 on ice for 15 min and PBS washing, the samples were blocked with 10% bovine serum albumin (BSA) for 30 min at room temperature. Similarly, the sections were repaired with EDTA buffer (pH = 8.0) and washed with PBS three times after dewaxing paraffin sections to water, and then the sections were blocked with 10% BSA at 37 °C for 30 min. Both cell and tissue samples were incubated overnight at 4 °C with primary antibodies and then for 2 hours at room temperature with secondary antibodies. Nuclei were counterstained with DAPI for 10 min in the dark before mounting with fluorescent media. Fluorescence imaging was performed using an Olympus microscope equipped with MicroPublisher imaging software (Q-imaging) for quantitative analysis of fluorescence intensity.

Extraction of total RNA and quantitative real-time PCR (qPCR)

Total RNA was isolated from tissues or cells using TRIzol Reagent (HYCEZMBIO, China). RNA concentration and purity were measured spectrophotometrically (FC-3100, Life Real Bio-Technology Co., Ltd, China), followed by reverse transcription into cDNA with a commercial kit (TransGen Biotech, China). qPCR was performed using an SYBR Green Master Mix (TransGen Biotech, China) according to the manufacturer's protocol. Gene expression levels were quantified via the 2^−ΔΔCt method, with TNF-α, IL-1β and IL-6 mRNA normalized to GAPDH as internal controls. Primer sequences are provided in Table 1.

Table 1 Primers used for RT-qPCR

Name	Sequence (5′ → 3′): forward and reverse	GenBank accession no.	Product size (bp)
TNF-α	GGGTGTTCATCCATTCTC	NM_001278601.1	187
	GGTCACTGTCCCAGCAT
IL-1β	TGCCACCTTTTGACAGTGATG	NM_008361.4	138
	TGATGTGCTGCTGCGAGATT
IL-6	TTCCATCCAGTTGCCTTCTTG	NM_031168.2	171
	CATTTCCACGATTTCCCAGAGA
β-Actin	CATCGTCCACCGCAAAT	NM_007393.5	118
	GCCATGCCAATCTCATCTC

---

Western blot analysis

Protein samples were obtained from tissues or cells through homogenization in RIPA lysis buffer (Biosharp, China) supplemented with protease inhibitors. Following centrifugation at 12 000g for 15 min at 4 °C, the supernatant protein concentration was quantified using a BCA assay (Beyotime Biotechnology, China) according to established protocols. Protein samples (50 μg per lane) underwent electrophoretic separation on 10% polyacrylamide gels under denaturing conditions before being electrotransferred onto PVDF membranes (Millipore, USA). Membrane blocking was performed for 2 hours at ambient temperature using 5% BSA in TBST. Immunoblotting was conducted with specific primary antibodies targeting TLR4 (Abcam, UK; 1 : 2000), NF-κB p65 (Abcam, UK; 1 : 2000), phosphorylated NF-κB p65 (Cell Signaling Technology, USA; 1 : 1000), and β-actin (Cell Signaling Technology, USA; 1 : 1000) as an internal reference, with overnight incubation at 4 °C. Following five washes with TBST, membranes were probed with HRP-conjugated secondary antibodies (Thermo Fisher Scientific, USA; 1 : 5000) for 60 min at room temperature. Protein–antibody complexes were visualized using enhanced chemiluminescence reagents (GE Healthcare, USA) and quantified through densitometric analysis using Image-Pro Plus 6.0 software (Media Cybernetics, USA). Relative protein expression levels were calculated by normalizing target protein band intensities to the corresponding β-actin signals.

Immunohistochemical staining

Uterus tissue sections were sequentially dewaxed and rehydrated, followed by antigen retrieval using citrate buffer. The sections were incubated overnight at 4 °C with primary antibodies against TNF-α, IL-1β and IL-6 (dilution 1 : 200). Subsequently, the samples were treated with Goat Anti-Rabbit IgG H&L secondary antibody at 37 °C for 30 min. Color development was performed using a diaminobenzidine kit (Servicebio, Wuhan, China). Finally, tissues were counterstained with hematoxylin, differentiated in 1% hydrochloric acid, dehydrated through graded ethanol and xylene series, mounted with neutral resin, and examined under light microscopy.

DNA extraction and 16S rRNA gene sequencing

Fecal bacterial genomic DNA was isolated using QIAamp DNA Isolation Kits (Qiagen, Germany), with the concentration quantified via Nanodrop spectrophotometry (Thermo Fisher Scientific) and the integrity verified through 1.2% agarose gel electrophoresis. The V3–V4 hypervariable region of bacterial 16S rRNA genes was amplified with barcoded primers 338F (5′-ACTCCTACGGGAGGCAGCAG-3′) and 806R (5′-GGACTACHVGGGTWTCTAAT-3′) under standardized PCR conditions. The resulting amplicons (∼450 bp) underwent purification using a dual-validation approach: size selection on 1.2% agarose gels followed by cleanup with either the QIAquick PCR Purification Kit (Qiagen) or Vazyme VAHTSTM DNA Clean Beads (Vazyme, China). Purified DNA was quantified via a PicoGreen fluorescence assay (BioTek FLx800 microplate reader), ensuring concentrations >25 ng μL⁻¹. The final library construction utilized the TruSeq Nano DNA LT Kit (Illumina), with paired-end sequencing (2 × 300 bp) performed on the Illumina MiSeq platform (600-cycle chemistry).

Sequencing data analysis

To assess species diversity within each sample, alpha and beta diversity analyses were conducted using the q2-diversity plugin in QIIME2 (version 2020.02, https://www.r-project.org/). The alpha diversity level of each sample was evaluated on the basis of the distribution of ASV/OUT. Beta diversity analysis was performed based on Jaccard, Bray–Curtis, unweighted UniFrac, and weighted UniFrac distances, which were visualized through Principal Coordinates Analysis (PCoA). Statistical differences among groups were evaluated using permutational multivariate analysis of variance (PERMANOVA) as implemented in the “vegan” package of R (version 3.3.1). To identify taxa that were significantly enriched in relation to treatment conditions, linear discriminant analysis effect size (LEfSe) was employed with a logarithmic LDA score threshold set at 2.0. The functional potential of the gut microbiota was inferred through Phylogenetic Investigation of Communities by Reconstruction of Unobserved States (PICRUSt), which predicts gene family abundance based on 16S rRNA gene sequences. The amplicon sequence variant (ASV) table was converted into PICRUSt-compatible format and normalized according to the 16S rRNA gene copy number to mitigate potential biases in microbial abundance estimates. Functional predictions were subsequently mapped to the Kyoto Encyclopedia of Genes and Genomes (KEGG) database for pathway annotation.

Metabolomic analyses

Fecal metabolomic profiling was performed according to established protocols with modifications.³¹ 5 mg of lyophilized fecal specimens underwent cryopreservation thawing on ice prior to metabolite extraction using 120 μL of methanol supplemented with isotopic internal standards. Following mechanical homogenization, the samples were subjected to centrifugation at 18 000g and 4 °C for 20 min. Derivatized extracts were then diluted with 330 μL of ice-cold 50% aqueous methanol, incubated at −20 °C for 20 min, and centrifuged at 4000g for 30 min. The processed supernatants were combined with 10 μL of retention time locking standards in fresh microplates, alongside serially diluted calibration standards for quantitative analysis. Metabolite separation and detection were achieved using a Waters ACQUITY UPLC system coupled with a Xevo TQ-S tandem mass spectrometer. Chromatographic separation employed ACQUITY UPLC BEH C18 columns (VanGuard pre-column: 2.1 × 5 mm; analytical column: 2.1 × 100 mm, both 1.7 μm particle size) with a binary mobile phase system (A: 0.1% formic acid in water; B: acetonitrile/isopropanol [70 : 30, v/v]) at a constant flow rate of 0.40 mL min⁻¹. Mass spectrometric detection was performed in the multiple reaction monitoring mode with optimized parameters: an ion source temperature of 150 °C, a desolvation temperature of 550 °C, and a desolvation gas flow of 1000 L h⁻¹.

Raw mass spectral data were processed through an integrated bioinformatics pipeline. Initial data reduction including peak picking, alignment, and quantification was performed using TMBQ software (v1.0). Subsequent multivariate analyses, including principal component analysis (PCA), partial least squares-discriminant analysis (PLS-DA), and univariate statistical comparisons (Student's t-test, Mann–Whitney U test, and ANOVA), were conducted using the iMAP platform (v1.0). Microbial–metabolite associations were assessed using Spearman's rank-order correlation analysis.

Hydroxyphenyllactic acid analysis

Uterine tissue and serum levels of HPLA were measured using enzyme-linked immunosorbent assay (ELISA) kits (Meimian Technology, Jiangsu, China) according to the manufacturer's instructions. Briefly, standard solutions and samples were incubated in designated wells at 37 °C for 30 min. Subsequently, 100 μL of detection antibody was dispensed into each well and incubated under identical conditions. After washing, wells received 50 μL of avidin-conjugated horseradish peroxidase solution and were incubated at 37 °C for 30 min. Finally, the immunoreaction was developed by adding 100 μL of TMB substrate solution. The enzymatic reaction was halted using a stop solution, and absorbance at 450 nm was quantified using a microplate reader.

Bioinformatic analysis

Next, potential protein targets of HPLA (PubChem CID: 9378) were predicted using Swiss Target Prediction (https://www.swisstargetprediction.ch/). Subsequently, these predicted targets were functionally annotated via the GO function analysis conducted using the BINGO plugin and the David database (https://david.ncifcrf.gov/). Molecular docking was then performed between HPLA and the TLR4 crystal structure (PDB ID: 4G8A) using AutoDock Vina (version 1.2.0). The binding affinity was quantified by calculating the free energy of binding (ΔG, kcal mol⁻¹).

Cell cultures and treatments

The endometrial epithelial cell line BEND cells were obtained from the American Typical Culture Collection (ATCC) and maintained in Dulbecco's Modified Eagle's Medium (DMEM, 4.5 g L⁻¹ glucose, HyClone, USA) supplemented with 10% fetal bovine serum (FBS) under standard culture conditions with 37 °C and 5% CO₂. Upon reaching 80–90% confluency, the BEND cells were challenged with either E. coli suspension alone at a multiplicity of infection (MOI) of 100 or combinatorial treatments.³¹ Following specified interventions, the cells were harvested for subsequent analyses.

Statistical analysis

All quantitative data were analyzed and visualized using GraphPad Prism 10 (GraphPad Software, USA), with the results expressed as mean ± standard error of the mean (SEM). The statistical differences between groups were determined using Student's t-test, and one-way analysis of variance (ANOVA) for various treatment groups. Statistical significance was established at p < 0.05 for all analyses.

Results

Hyperoside ameliorates E. coli-induced endometritis and associated intestinal inflammation

To investigate the therapeutic effect of hyperoside on endometritis, a mouse model was established via intrauterine injection of E. coli, followed by 3-day oral administration of hyperoside (Fig. 1A). Compared with the control group, the uteri in the E. coli group exhibited significant congestion and swelling, which were observably alleviated after hyperoside treatment (Fig. 1B). Histopathological analysis revealed that the E. coli group displayed abnormal endometrial epithelial hyperplasia, glandular hyperplasia with obscured luminal structure, loss of boundary definition in the lamina propria, and extensive inflammatory cells and erythrocyte infiltration. Consequently, pathological scores were significantly elevated versus the control group. Hyperoside administration substantially attenuated these histopathological changes, resulting in significantly reduced pathological scores and restoration of near-normal tissue architecture (Fig. 1C and D). Furthermore, hyperoside significantly upregulated the expression of uterine tight junction proteins occludin and ZO-1 (Fig. 1E–G), while downregulating the levels of pro-inflammatory cytokines, including TNF-α, IL-1β, and IL-6 (Fig. 1H–J). To further evaluate the effect of hyperoside on endometritis, we subsequently examined the activation of the TLR4/NF-κB pathway, which have been extensively implicated in the pathogenesis of endometritis.^1,32 Indeed, we found significantly elevated levels of TLR4 and NF-κB in the E. coli-infected group, which was effectively suppressed by hyperoside treatment (Fig. 1K–M). Notably, E. coli infection also induced colonic inflammation characterized by shortened colon length and inflammatory infiltration when compared with the control group. Nevertheless, hyperoside improved colonic growth inhibition and histopathological changes in the colon (Fig. 1N–Q). Taken together, our data suggest that hyperoside exerts dual therapeutic effects by mitigating both E. coli-triggered endometritis and associated intestinal inflammation.

Fig. 1 Hyperoside attenuated E. coli-induced endometritis and intestinal inflammation in mice. (A) Illustration of the in vivo mouse experiment. The mice were divided into four groups and injected with 10⁸ CFU of E. coli in the uterus to induce endometritis and then treated with hyperoside. Created with BioGDP.com.⁶⁴ (B) Uterine tissue morphology. (C and D) H&E staining of the mice uterine tissue samples and histological scores of uterine tissues (n = 3 per group). (E–G) The expression levels of tight junction proteins occludin and ZO-1 were measured by immunofluorescence. Occludin protein was labeled with a green fluorophore, ZO-1 protein was labeled with a red fluorophore, and the cell nucleus was labeled with a blue fluorophore (n = 5 per group). (H–J) The expression levels of TNF-α, IL-1β and IL-6 were measured by RT-qPCR (n = 3 per group). (K–M) The expression levels of TLR4 and p-p65 were detected by western blotting (n = 3 per group). β-Actin was used as a control. (N) Colon morphology. (O) Colon length (n = 4 per group). (P and Q) H&E staining of the mice colon tissue samples and histological scores of colonic tissues (n = 3 per group). Data are presented as the mean ± SEM of three independent experiments. Statistical significance was determined using one-way ANOVA. *p < 0.05, **p < 0.01, and ***p < 0.001.

Hyperoside-modulated gut microbiota confer protection against E. coli-induced endometritis

To elucidate the gut microbiota-dependent mechanism underlying the protective effect of hyperoside against E. coli-induced endometritis, we employed FMT in antibiotic-pretreated mice (Fig. 2A). Recipients receiving microbiota from E. coli-infected donors developed characteristic uterine pathology including pronounced congestion and edema, while those colonized with hyperoside-treated donor microbiota preserved the normal uterine morphology (Fig. 2B). Quantitative histopathological evaluation demonstrated that the FMT-Hyp + E. coli group maintained intact endometrial epithelium with significantly attenuated inflammatory cell infiltration compared to the FMT-E. coli group, as evidenced by reduced pathological scores (Fig. 2C and D). Immunofluorescence staining quantitatively confirmed the restoration of epithelial barrier function in the FMT-Hyp + E. coli group, showing markedly enhanced expression of tight junction proteins occludin and ZO-1 compared with the FMT-E. coli group (Fig. 2E–G). Furthermore, the hyperoside-modulated gut microbiota significantly downregulated the production of uterine proinflammatory cytokines (Fig. 2H–J). Notably, TLR4/NF-κB signaling pathway activation was substantially inhibited in mice receiving hyperoside-treated donor microbiota (Fig. 2K–M). These findings established that the therapeutic efficacy of orally administered hyperoside against E. coli-induced endometritis was mediated through functional modulation of the gut microbiota.

Fig. 2 Fecal microbiota transplantation (FMT) from hyperoside-treated donors alleviated E. coli-induced endometritis in recipient mice. (A) Diagram of FMT experiments. The mice that were pre-treated with an antibiotic solution (ATB) for 3 days underwent FMT. One day after FMT, the mice were infected with 10⁸ CFU of E. coli. Created with BioGDP.com.⁶⁴ (B) Uterine tissue morphology. (C and D) H&E staining of the mice uterine tissue samples and histological scores of uterine tissues (n = 3 per group). (E–G) The expression levels of tight junction proteins occludin and ZO-1 were measured by immunofluorescence. Occludin protein was labeled with a green fluorophore, ZO-1 protein was labeled with a red fluorophore, and the cell nucleus was labeled with a blue fluorophore (n = 4 per group). (H–J) The expression levels of TNF-α, IL-1β and IL-6 were measured by RT-qPCR (n = 3 per group). (K–M) The expression levels of TLR4 and p-p65 were detected by western blotting (n = 3 per group). β-Actin was used as a control. (N) Relative abundances of fecal bacterial at the genus level. (O) Differentially abundant taxa of fecal microbiota among FMT groups was analyzed by LEfSe. LDA score ≥ 2. Data are presented as the mean ± SEM of three independent experiments (n = 4 per group). Statistical significance was determined using one-way ANOVA. *p < 0.05, **p < 0.01, and ***p < 0.001.

Hyperoside alters the composition of the gut microbiota in mice suffering from E. coli-induced endometritis

The above findings confirmed the essential role of the gut microbiota in the protective effect of hyperoside against E. coli-induced endometritis. However, the specific gut microbiota responsible for the protective effect remain undefined. To identify the critical gut microbiota involved in the anti-inflammatory effect of hyperoside, 16S rRNA sequencing was performed to analyze the fecal microbial composition. Alpha diversity indices, including Shannon and Chao1, were significantly lower in the control, Hyp, and Hyp + E. coli groups compared to the E. coli group, indicating a substantial change in species richness and diversity following E. coli infection (Fig. 3A and B). Beta diversity analysis, assessed via PCoA of weighted UniFrac distances, Principal Component Analysis (PCA), and Non-metric Multidimensional Scaling (NMDS), revealed distinct clustering and significant dispersion of data points among the control, Hyp, E. coli, and Hyp + E. coli groups, demonstrating pronounced differences in fecal microbiota composition (Fig. 3C and Fig. S2A, B). At the phylum level, the E. coli group exhibited decreased abundance of Firmicutes, with increased abundance of Proteobacteria compared with other groups (Fig. S2C). At the family level, hyperoside treatment markedly increased the abundance of Lactobacillaceae while decreasing Lachnospiraceae abundance compared to the E. coli group (Fig. S2D). Genus-level analysis revealed that hyperoside treatment significantly upregulated Lactobacillus and Prevotella when compared with the E. coli group, while reducing Mucispirillum abundance compared with the E. coli group (Fig. 3D and E). LEfSe analysis identified differentially abundant bacterial taxa, and the results showed that the relative abundance of Mucispirillum significantly increased after E. coli infection, while hyperoside treatment increased the abundance of Lactobacillus (Fig. 3F–H). KEGG pathway enrichment analysis of the differentially abundant microbiota revealed enrichment in biological pathways including cell growth and death, infectious diseases, and carbohydrate metabolism (Fig. S3). To investigate potential associations between the gut microbiota and endometritis-related inflammatory phenotypes, we performed Spearman correlation analysis between pro-inflammatory cytokines, inflammation-related proteins, tight junction proteins, and gut microbiota. The results revealed significant negative correlations (p < 0.05) between bacteria such as Lactobacillales, Blautia, and Corynebacterium and the pro-inflammatory factors. Conversely, Blautia, Corynebacterium, and Streptococcus showed significant positive correlations (p < 0.05) with occludin and ZO-1 (Fig. 3I and Table S1). Taken together, these data demonstrated that hyperoside treatment triggered changes in the gut microbiota of mice with E. coli-induced endometritis, and the protective effect of hyperoside may be attributed to an increase in Lactobacillus and a concurrent decrease in Mucispirillum.

Fig. 3 Oral administration of hyperoside altered the composition of the gut microbiota in mice. (A) Microbial community diversity (measured using the Shannon index). (B) Microbial community abundance (measured using the Chao1 index). (C) Principal coordinates analysis (PCoA) based on the weighted UniFrac distance matrix. (D) Relative abundances of fecal bacteria at the genus level. (E) Heat map of microbiota at the genus level in fecal samples. Color indicates the relative microbiota abundances in the group samples; the corresponding relationship between the color gradient and the value is shown in the gradient color block. (F) Differentially abundant taxa of fecal microbiota from different treatment groups were analyzed by LEfSe. LDA score ≥ 2. (G) The abundance of the genus Mucispirillum. (H) The abundance of the genus Lactobacillus. (I) Heatmap of Spearman's correlation between gut microbiota abundance and phenotype/inflammatory factors. *p < 0.05 and **p < 0.01.

To further explore whether hyperoside protects against E. coli infection by maintaining the balance of the gut microbiota, we next analyzed the composition of fecal microbial community in the FMT experiment using 16S rRNA gene sequencing. PCoA and NMDS analyses indicated that the gut microbiota composition in the FMT-E. coli group was significantly distinct from the other groups (Fig. S4A and B). At the phylum level, Bacteroidetes, Proteobacteria, and Firmicutes were the predominant phyla across groups. At the family level, Enterobacteriaceae was more enriched in the FMT-Hyp + E. coli group (Fig. S4C and D). In addition, Bacteroides and Lactobacillus were the dominant genera, with Clostridium showing significant enrichment specifically in the FMT-Hyp + E. coli group compared with the FMT-E. coli group (Fig. 2N and O). All the above results signified that FMT from hyperoside-treated donor mice alleviated endometritis, potentially mediated through the modulation of the gut microbiota.

Gut microbiota-derived HPLA exhibits a close association with the protection of hyperoside against E. coli-induced endometritis

Accumulating evidence indicated that gut microbiota-derived metabolites could combat exogenous pathogens through host–microbiota interactions.^14,33 Therefore, targeted metabolomics based on UPLC-MS/MS was performed to identify potential gut microbiota-derived metabolites against E. coli-induced endometritis. Metabolite classification revealed that amino acids accounted for 56.17% of identified compounds, followed by carbohydrates (13.92%), short-chain fatty acids (SCFAs, 11.36%), organic acids (10.75%), and fatty acids (5.12%) (Fig. S5A). PLS-DA demonstrated significant compositional differences among groups (Component 1: p = 0.023; Component 2: p = 0.031; Fig. 4A and Fig. S5B). The volcano plots showed the differentially expressed metabolites, with 1 upregulated and 8 downregulated metabolites in the E. coli group compared with the control group (Fig. S6A), and 1 significantly upregulated metabolite (hydroxyphenyllactic acid, HPLA) and 4 downregulated metabolites were identified in the Hyp + E. coli group compared with the E. coli group (Fig. S6B). 14 potential biomarkers were identified by screening metabolites with different levels of content and visualized in a clustered heatmap (Fig. 4B). Phenylpropanoic acids constituted the most abundant class of biomarkers, and the concentration of HPLA in the Hyp + E. coli group was significantly higher than that in the E. coli group (Fig. 4B and C). Spearman correlation analysis revealed significant associations between differentially enriched bacteria and metabolites. HPLA exhibited a strong positive correlation with Lactobacillus (R = 0.58 and p = 0.02) and an observable negative correlation with Mucispirillum (R = −0.53 and p = 0.04) (Fig. 4D). To verify whether the effect of hyperoside on combating bacterial endometritis was mediated by gut microbiota-derived HPLA, we next quantified HPLA levels in uterine tissue and serum. As shown in Fig. 4E and F, HPLA was significantly reduced in both uterine tissue and serum of the E. coli group compared to the control group, while hyperoside treatment substantially restored HPLA levels. Collectively, these findings suggested that hyperoside treatment in E. coli-induced endometritis may reshape commensal microbiota structure, enhance HPLA synthesis, and subsequently elevate circulating HPLA levels, which likely contributed to the attenuation of uterine inflammation.

Fig. 4 Changes in the fecal metabolomic profiles among different treatment groups. (A) The fecal metabolomic profiles were clustered using PLS-DA. Data are presented as the mean ± SEM. p-Values were determined using the nonparametric Kruskal–Wallis test. (B) Relative abundances of metabolites were clustered using a UPGMA dendrogram and shown in a heatmap. The metabolite variation is shown using the Z_score. (C) The concentration of fecal hydroxyphenyllactic acid (HPLA) was displayed as a box and dot plot. Data are presented as the mean ± SEM. p-Values were determined using the nonparametric Kruskal–Wallis test. (D) Heatmap of Spearman correlation between the gut microbiota and metabolites. The relevance between differentially enriched bacteria and metabolites from mice with different treatments was interpreted using Spearman correlation analysis. The colors range from blue (negative correlation) to red (positive correlation). (E) The concentration of HPLA was measured in the uterus. (F) The concentration of HPLA was measured in serum. Data are presented as the mean ± SEM of three independent experiments. Statistical significance was determined using one-way ANOVA. *p < 0.05, **p < 0.01, and ***p < 0.001.

Oral administration of HPLA ameliorates endometritis induced by E. coli

To investigate whether gut microbiota-derived HPLA alleviated inflammatory injury of the uterus, a murine endometritis model was established via intrauterine inoculation of E. coli, followed by continuous oral administration of HPLA for 3 days (Fig. 5A). Macroscopic examination revealed that HPLA treatment substantially ameliorated E. coli-induced uterine hyperemia and edema (Fig. 5B). Subsequent histopathological analysis demonstrated severe pathological alterations in the E. coli group compared with the control group, including extensive endometrial epithelial denudation and dense inflammatory cell infiltration within the endometrial stroma and glands. Remarkably, oral gavage of HPLA significantly attenuated these histopathological changes (Fig. 5C and D). To further evaluate the anti-inflammatory effect of HPLA, the levels of uterine tight junction proteins and pro-inflammatory cytokines were assessed. Immunofluorescence analysis indicated markedly elevated expression of both occludin and ZO-1 following HPLA treatment, suggesting restoration of endometrial barrier integrity (Fig. 5E–G). Concurrently, the delivery of HPLA obviously suppressed the expression of pro-inflammatory mediators TNF-α, IL-1β, and IL-6 compared with the E. coli group (Fig. 5H–K). Further investigation showed that HPLA intervention potently inhibited the TLR4/NF-κB signaling pathway. Specifically, HPLA-treated mice exhibited dramatically reduced levels of both TLR4 and phosphorylated NF-κB p65 proteins compared with the E. coli-infected group (Fig. 5L–O). These findings demonstrated that HPLA can attenuate E. coli-induced endometritis through suppression of the TLR4/NF-κB pathway.

Fig. 5 Administration of HPLA alleviated E. coli-induced endometritis in mice. Oral administration of hyperoside altered the composition of the gut microbiota in mice. (A) Experimental diagram of the mouse infection model. The mice were divided into four groups, injected with 10⁸ CFU of E. coli in the uterus to induce endometritis, and then treated with HPLA. Created with BioGDP.com.⁶⁴ (B) Uterine tissue morphology. (C and D) H&E staining of the mice uterine tissue samples and histological scores of uterine tissues (n = 3 per group). (E–G) The expression levels of tight junction proteins occludin and ZO-1 were measured by immunofluorescence. Occludin protein was labeled with a green fluorophore, ZO-1 protein was labeled with a red fluorophore, and the cell nucleus was labeled with a blue fluorophore (n = 3 per group). (H–K) The expression levels of TNF-α, IL-1β and IL-6 were measured by immunohistochemical staining (n = 3 per group). (L–O) The expression levels of TLR4 and p-p65 were detected by western blotting. β-Actin was used as a control. Data are presented as the mean ± SEM of three independent experiments (n = 3 per group). Statistical significance was determined using one-way ANOVA. *p < 0.05, **p < 0.01, and ***p < 0.001.

The therapeutic efficacy of hyperoside against E. coli-induced endometritis is dependent on gut microbiota-derived HPLA

To further validate that the therapeutic effect against E. coli-induced endometritis was mediated by gut microbiota-derived HPLA enrichment, we administered ATB to eliminate the gut microbiota, followed by oral treatment with hyperoside with or without HPLA supplementation. This strategy enabled the evaluation of whether HPLA supplementation could restore the therapeutic effect of hyperoside on endometritis after antibiotic-induced gut microbiota depletion (Fig. 6A). Gross examination of uterine tissues revealed significant swelling in both the ATB + E. coli and ATB + Hyp + E. coli groups. In contrast, uterine tissues from the ATB + Hyp + HPLA + E. coli group appeared nearly normal (Fig. 6B). Histopathological analysis confirmed more severe uterine tissue damage in the ATB + E. coli and ATB + Hyp + E. coli groups, characterized by pronounced hyperemia and increased immune cell infiltration. Conversely, tissue damage was markedly alleviated in the ATB + Hyp + HPLA + E. coli group, evidenced by an observably reduced histological score (Fig. 6C and D). Consistent with the results, HPLA treatment reversed the decrease in the expression of the tight junction proteins occludin and ZO-1 observed in the ATB + Hyp + E. coli group (Fig. 6E–G). Compared with the ATB + Hyp + E. coli group, the ATB + Hyp + HPLA + E. coli group exhibited significantly reduced levels of pro-inflammatory cytokines, including TNF-α, IL-1β, and IL-6 (Fig. 6H–J). Furthermore, the ATB + Hyp + HPLA + E. coli group also showed significant suppression of the activation of the TLR4/NF-κB inflammatory signaling pathway (Fig. 6K and M). These findings suggested that the therapeutic effect of hyperoside against E. coli-induced endometritis was not directly attributable to its action on the uterus, but rather stemmed from the role in enriching gut microbiota-derived HPLA.

Fig. 6 Hyperoside, rather than directly exerting anti-inflammatory effects, regulated protective effects against E. coli-induced endometritis through gut microbiota-derived HPLA. (A) Illustration of the in vivo mouse experiment. The mice were pre-treated with an antibiotic solution (ATB) for 3 days to deplete the gut microbiota, and then injected with 10⁸ CFU of E. coli in the uterus to induce endometritis. Subsequently, hyperoside was administered orally. Meanwhile, HPLA was administered for 3 days. Created with BioGDP.com.⁶⁴ (B) Uterine tissue morphology. (C and D) H&E staining of the mice uterine tissue samples and histological scores of uterine tissues (n = 3 per group). (E–G) The expression levels of tight junction proteins occludin and ZO-1 were measured by immunofluorescence. Occludin protein was labeled with a green fluorophore, ZO-1 protein was labeled with a red fluorophore, and the cell nucleus was labeled with a blue fluorophore (n = 3 per group). (H–J) The expression levels of TNF-α, IL-1β and IL-6 were measured by RT-qPCR (n = 3 per group). (K–M) The expression levels of TLR4 and p-p65 were detected by western blotting (n = 3 per group). β-Actin was used as a control. Data are presented as the mean ± SEM of three independent experiments. Statistical significance was determined using one-way ANOVA. “ns” represents not significant. *p < 0.05, **p < 0.01, and ***p < 0.001.

HPLA directly targets TLR4 protein and ameliorates E. coli-induced endometritis by inhibiting the expression of TLR4

We next investigated the modulatory effect of HPLA on the activation of TLR4 and the downstream NF-κB pathway in vitro. Immunofluorescence analysis revealed that HPLA significantly reduced the expression of TLR4 protein and inhibited nuclear translocation of p65 in endometrial epithelial cells (Fig. 7A–D). Western blot measurement further confirmed that HPLA treatment restrained the upregulation of TLR4 and phosphorylated p65 protein levels caused by E. coli infection, which was consistent with in vivo findings (Fig. 7E–G). To elucidate the molecular mechanism underlying the anti-endometritis activity of HPLA, bioinformatic screening using Swiss Target Prediction identified 100 potential targets of HPLA. GO enrichment analysis of the top 20 significantly enriched biological processes involved in the intracellular receptor signaling pathway, response to lipopolysaccharide, and inflammatory response (Fig. 7H). Molecular docking demonstrated that HPLA binds within the active pocket of TLR4 with a binding energy of −6.8 kcal mol⁻¹ (Fig. 7I), which is significantly lower than the −4.35 kcal mol⁻¹ reported for vitexin, a LPS antagonist candidate, suggesting the superior binding affinity of HPLA to TLR4.³⁴ The cellular thermal shift assay (CETSA) indicated that HPLA markedly enhanced TLR4 thermal stability, which suggested the direct interaction of HPLA with TLR4 (Fig. 7J and K). Furthermore, a translational inhibitor, cycloheximide (CHX), was applied to treat cells to investigate whether HPLA can modulate the stability of TLR4 protein. The results from Fig. 7L and M revealed that HPLA co-treatment significantly accelerated time-dependent degradation of TLR4 protein. These aforementioned results demonstrated that HPLA suppressed E. coli-induced inflammation in endometrial epithelial cells by modulating the translation and stability of TLR4 protein.

Fig. 7 HPLA can mitigate E. coli-induced inflammation in endometrial epithelium cells through binding to TLR4 and affecting its translation. (A and B) The expression levels of TLR4 protein were measured by immunofluorescence. TLR4 protein was labeled with a red fluorophore, and the cell nucleus was labeled with a blue fluorophore. (C and D) The expression levels of p-p65 protein were measured by immunofluorescence. p-p65 Protein was labeled with a red fluorophore, and the cell nucleus was labeled with a blue fluorophore. (E–G) The expression levels of TLR4 and p-p65 were detected by western blotting. β-Actin was used as a control. (H) GO enrichment analysis of the top 20 significantly enriched biological processes of targets related to HPLA. (I) Molecular docking of HPLA molecules and TLR4 protein using AutoDock Vina. The spatial binding model demonstrates HPLA binding within the TLR4 binding pocket. (J and K) CETSA quantitative analysis showing the thermal stability of TLR4 protein with HPLA treatment (10 μM) or without treatment. (L and M) BEND cells were treated with CHX, with or without HPLA treatment. Protein expression of TLR4 was analyzed by western blotting. Data are presented as the mean ± SEM of three independent experiments. Statistical significance was determined using one-way ANOVA. “ns” represents not significant. *p < 0.05, **p < 0.01, and ***p < 0.001.

Discussion

Endometritis, a prevalent disorder of the female reproductive system, is closely associated with the infection of exogenous pathogens such as E. coli.^35,36 The antibiotic therapy remains the primary clinical intervention, while the emergence of resistance due to its long-term use poses a significant therapeutic challenge.³⁷ Consequently, the development of safe, effective, and resistance-sparing alternative treatments is of paramount scientific importance. Growing evidence has implicated the gut microbiota and its metabolites in the pathogenesis of various diseases, including endometritis.³⁸ However, the specific mechanisms by which gut microbiota derived metabolites influence endometritis remain unclear. Our study explored the therapeutic potential of hyperoside, a naturally occurring flavonol glycoside ubiquitous in the diet (e.g., fruits, vegetables and edible plants), against E. coli-induced endometritis. We focused on its role as a nutraceutical strategy by elucidating the underlying mechanisms involving the gut microbiota and its metabolites. In the current study, we initially confirmed the significant therapeutic effect of hyperoside on murine endometritis, which was closely related to the inhibition of the TLR4/NF-κB pathway, leading to the reduction in the production of pro-inflammatory cytokines, which was consistent with the previous report by Xie et al.³⁹ Nevertheless, the precise mechanism by which hyperoside regulates the TLR4/NF-κB pathway to alleviate endometritis still requires further investigation.

Given that hyperoside concurrently mitigated both E. coli-induced endometritis and associated intestinal inflammation, we subsequently focused on gut microbiota homeostasis. Accumulating evidence indicates that gut microbiota dysbiosis, by perturbing the microbial metabolome, contributes to diverse pathological processes, including diseases of distal organs.^40–42 Research on established axes like the gut–brain, gut–lung, and gut–liver axes highlights mechanisms of remote organ regulation by gut microbes via metabolites and signaling molecules.^43–45 Nevertheless, the potential existence of a similar axis connecting the gut and uterus had not been systematically investigated. FMT experiments revealed that recipient mice receiving gut microbiota from hyperoside-treated donors exhibited significantly attenuated endometrial damage following E. coli challenge compared with those receiving microbiota from the control or E. coli-infected donors. These results provide direct and compelling evidence that gut microbiota remodeling is essential for the protective effects of hyperoside on endometritis. Further 16S rRNA gene sequencing delineated the specific impact of hyperoside intervention on the composition of the gut microbiota. Mice with E. coli-induced endometritis displayed a noteworthy decrease in the relative abundance of beneficial bacteria, such as Lactobacillus and Prevotella, alongside a reduction in Mucispirillum. Critically, hyperoside treatment effectively reversed these alterations, markedly increasing Lactobacillus and Prevotella abundance. This observation is consistent with previous studies showing diminished abundance of Lactobacillus in the gut during endometritis and the documented therapeutic role of Prevotella in mitigating microbiota dysbiosis-associated uterine injury.^46–48 It has been demonstrated that probiotic strains of human-derived Lactobacillus alleviate intestinal inflammation through immunomodulatory effects in a strain-dependent manner.⁴⁹ Similarly, human gut-derived Prevotella histicola reveals immunomodulatory properties, as evidenced by its suppression of inflammatory arthritis in humanized mice.⁵⁰ Therefore, we propose that hyperoside alleviates uterine inflammation by enriching beneficial gut bacteria.

To identify key signaling molecules mediating the “gut–uterus” crosstalk, we performed targeted metabolomic analysis on murine fecal samples. HPLA was identified as a potential biomarker strongly correlated with disease status and therapeutic response. Spearman correlation analysis integrating microbiome and metabolome data revealed a significant positive correlation between HPLA levels and Lactobacillus abundance, and a significant negative correlation with Mucispirillum abundance, which suggests that HPLA production is intimately related to the enrichment of beneficial bacteria mediated by hyperoside. Supporting a potential role in host defense, prior studies have indicated that HPLA may confer resistance against exogenous pathogens.^51–53 More importantly, the functional validation ulteriorly demonstrated that exogenous administration of HPLA effectively ameliorated endometrial inflammation. Crucially, HPLA levels were significantly reduced in both serum and uterine tissues of E. coli-infected mice compared with the control group, while hyperoside treatment restored the HPLA concentrations in both compartments. Host–microbiota interactions are frequently mediated by metabolites.^54,55 Our research definitively established that the key effector substance for directly alleviating E. coli-induced endometritis was gut microbiota-derived HPLA, rather than hyperoside. It has also been reported that increased gut-derived HPLA alleviates renal cell injury by upregulating an apoptosis repressor with a caspase recruitment domain to protect mice from sepsis-induced acute kidney injury.⁵⁶ The series of findings clearly demonstrates that gut microbiota derived HPLA enters systemic circulation and reaches the uterine tissue to exert local anti-inflammatory effects.

We next sought to determine how HPLA acts within the uterine microenvironment. TLR4 was identified as the primary target mediating the anti-inflammatory effects of HPLA. As a vital innate immune receptor, TLR4 plays a central role in initiating the inflammatory cascade of endometritis, and its inhibitors have proven effective in reducing endometritis.^57,58 The NF-κB signaling pathway, downstream of TLR4, is a well-established core hub for uterine inflammatory responses and a promising therapeutic target.^59,60 Molecular docking simulations revealed that HPLA stably bound within the active pocket of the TLR4 protein, which provides a structural basis for directly inhibiting the effect of HPLA on TLR4 and subsequent phosphorylation of NF-κB, thereby disrupting the inflammatory cascade.

Although there are known differences between mice and humans in immune regulation and the gut microbiota, the core mechanism of anti-inflammatory action of TLR4/NF-κB revealed in this study is a highly conserved inflammatory regulatory pathway among different mammals.⁶¹ Furthermore, the key beneficial bacterial genera highlighted in our study, particularly Lactobacillus, which is closely associated with HPLA production, represent dominant and functionally integral components of the human gut microbiota.⁶² It has been reported that HPLA is a well-documented microbial metabolite in humans, and its systemic levels are closely associated with host health status.⁶³ Collectively, the high degree of conservation across species in both the HPLA-mediated anti-inflammatory mechanism and its underlying microbial drivers provides a solid foundation for future translational research targeting the gut–uterus axis in the treatment of endometritis.

In summary, our study is the first to systematically elucidate the core mechanism by which hyperoside alleviates E. coli-induced endometritis via the gut–uterus axis. Dietary supplementation with hyperoside remodels the gut microbiota structure, significantly enriching beneficial genera, such as Lactobacillus. These bacteria promote the production of their pivotal metabolite, HPLA. HPLA enters the circulatory system and reaches the uterine tissue. Within the uterus, HPLA directly targets and binds TLR4, effectively inhibiting the TLR4/NF-κB signaling pathway, ultimately mitigating inflammatory damage and tissue destruction (Fig. 8). Our present study not only elucidates a detailed and novel molecular mechanism for the therapeutic efficacy of hyperoside against endometritis but also, more significantly, provides compelling evidence for the existence and pathophysiological relevance of the gut–uterus axis.

Fig. 8 Schematic diagram of the effects of hyperoside on regulating E. coli-induced endometritis. Created with BioGDP.com.⁶⁴

Author contributions

K. J. and J. Y. designed and supervised the study. K. J., J. Y., Y. C., and A. X. performed the experiments and drafted the manuscript. J. Y. and A. X. analyzed the data. J. L., F. W., and X. L. collected the tissue samples. K. J., J. Y., J. B., and B. X. revised the manuscript. All authors read and approved the final manuscript.

Conflicts of interest

There are no conflicts to declare.

Abbreviations

E. coli	Escherichia coli
HPLA	Hydroxyphenyllactic acid
LPS	Lipopolysaccharide
TLR4	Toll-like receptor 4
ATB	Antibiotic cocktail
FMT	Fecal microbiota transplantation
H&E	Hematoxylin and eosin
DMEM	Dulbecco's modified Eagle's medium
FBS	Fetal bovine serum
PCoA	Principal coordinates analysis
PCA	Principal component analysis
NMDS	Non-metric multidimensional scaling
LEfSe	Linear discriminant analysis of effect size
CHX	Cycloheximide

Data availability

16S rRNA gene data have been deposited in the National Center for Biotechnology Information (NCBI) Sequence Read Archive (SRA) database under accession numbers PRJNA1297326 and PRJNA1297348. Metabolomics raw data have been deposited in Mendeley Data (https://data.mendeley.com/datasets/w67tnzmwgy/1). The data that support the findings of this study are available from the corresponding author upon reasonable request.

See DOI: https://doi.org/10.1039/d5fo04275e.

Acknowledgements

This work was supported by the National Natural Science Foundation of China (grant no. 32560879 and 32202885) and the Yunnan Fundamental Research Project (grant no. 202301AT070492).

References

1. K. Jiang, J. Yang, C. Song, F. He, L. Yang and X. Li, Enforced expression of miR-92b blunts E. coli lipopolysaccharide-mediated inflammatory injury by activating the PI3K/AKT/β-catenin pathway via targeting PTEN, Int. J. Biol. Sci., 2021, 17, 1289–1301.
2. N. Singh and A. Sethi, Endometritis - Diagnosis,Treatment and its impact on fertility - A Scoping Review, JBRA Assist. Reprod., 2022, 26, 538–546.
3. I. M. Sheldon, G. S. Lewis, S. LeBlanc and R. O. Gilbert, Defining postpartum uterine disease in cattle, Theriogenology, 2006, 65, 1516–1530.
4. K. Jiang, J. Cai, Q. Jiang, J. J. Loor, G. Deng, X. Li and J. Yang, Interferon-tau protects bovine endometrial epithelial cells against inflammatory injury by regulating the PI3K/AKT/β-catenin/FoxO1 signaling axis, J. Dairy Sci., 2024, 107, 555–572.
5. C. Piras, Y. Guo, A. Soggiu, M. Chanrot, V. Greco, A. Urbani, G. Charpigny, L. Bonizzi, P. Roncada and P. Humblot, Changes in protein expression profiles in bovine endometrial epithelial cells exposed to E. coli LPS challenge, Mol. BioSyst., 2017, 13, 392–405.
6. K. Jiang, J. Yang, G. Xue, A. Dai and H. Wu, Fisetin Ameliorates the Inflammation and Oxidative Stress in Lipopolysaccharide-Induced Endometritis, J. Inflammation Res., 2021, 14, 2963–2978.
7. A. Ciesielska, M. Matyjek and K. Kwiatkowska, TLR4 and CD14 trafficking and its influence on LPS-induced pro-inflammatory signaling, Cell. Mol. Life Sci., 2021, 78, 1233–1261.
8. W. Cao, W. Zhang, J. Liu, Y. Wang, X. Peng, D. Lu, R. Qi, Y. Wang and H. Wang, Paeoniflorin improves survival in LPS-challenged mice through the suppression of TNF-α and IL-1β release and augmentation of IL-10 production, Int. Immunopharmacol., 2011, 11, 172–178.
9. Y. Belkaid and T. W. Hand, Role of the microbiota in immunity and inflammation, Cell, 2014, 157, 121–141.
10. C. Chen, J. Liao, Y. Xia, X. Liu, R. Jones, J. Haran, B. McCormick, T. R. Sampson, A. Alam and K. Ye, Gut microbiota regulate Alzheimer's disease pathologies and cognitive disorders via PUFA-associated neuroinflammation, Gut, 2022, 71, 2233–2252.
11. P. Zhang, L. Zheng, Y. Duan, Y. Gao, H. Gao, D. Mao and Y. Luo, Gut microbiota exaggerates triclosan-induced liver injury via gut-liver axis, J. Hazard. Mater., 2022, 421, 126707.
12. Q. Zhang, J. Hu, J. W. Feng, X. T. Hu, T. Wang, W. X. Gong, K. Huang, Y. X. Guo, Z. Zou, X. Lin, R. Zhou, Y. Q. Yuan, A. D. Zhang, H. Wei, G. Cao, C. Liu, L. L. Chen and M. L. Jin, Influenza infection elicits an expansion of gut population of endogenous Bifidobacterium animalis which protects mice against infection, Genome Biol., 2020, 21, 99.
13. X. Hu, R. Mu, M. Xu, X. Yuan, P. Jiang, J. Guo, Y. Cao, N. Zhang and Y. Fu, Gut microbiota mediate the protective effects on endometritis induced by Staphylococcus aureus in mice, Food Funct., 2020, 11, 3695–3705.
14. Y. He, J. Cai, X. Xie, X. Zhang, L. Qu, J. Liu and Y. Cao, Dimethyl Itaconate Alleviates Escherichia coli-Induced Endometritis Through the Guanosine-CXCL14 Axis via Increasing the Abundance of norank_f_Muribaculaceae, Adv. Sci., 2025, 12, e2414792.
15. B. Fang, L. Lu, M. Zhao, X. Luo, F. Jia, F. Feng and J. Wang, Mulberry (Fructus mori) extract alleviates hyperuricemia by regulating urate transporters and modulating the gut microbiota, Food Funct., 2024, 15, 12169–12179.
16. J. Xia, Y. Wan, J. J. Wu, Y. Yang, J. F. Xu, L. Zhang, D. Liu, L. Chen, F. Tang, H. Ao and C. Peng, Therapeutic potential of dietary flavonoid hyperoside against non-communicable diseases: targeting underlying properties of diseases, Crit. Rev. Food Sci. Nutr., 2024, 64, 1340–1370.
17. E. Jang, Hyperoside as a Potential Natural Product Targeting Oxidative Stress in Liver Diseases, Antioxidants, 2022, 11, 1437.
18. A. Jakubczyk, K. Kiersnowska, B. Ömeroğlu, U. Gawlik-Dziki, K. Tutaj, K. Rybczyńska-Tkaczyk, M. Szydłowska-Tutaj, U. Złotek and B. Baraniak, The Influence of Hypericum perforatum L. Addition to Wheat Cookies on Their Antioxidant, Anti-Metabolic Syndrome, and Antimicrobial Properties, Foods, 2021, 10, 1379.
19. Y. Yang, J. J. Huang, G. S. Zhu and W. Hu, Hyperoside attenuates osteoarthritis progression by targeting PI3K/Akt/NF-κB signaling pathway: In vitro and in vivo studies, J. Funct. Foods, 2022, 95, 10.
20. S. J. Kim, J. Y. Um and J. Y. Lee, Anti-inflammatory activity of hyperoside through the suppression of nuclear factor-κB activation in mouse peritoneal macrophages, Am. J. Chin. Med., 2011, 39, 171–181.
21. H. H. Fan, L. B. Zhu, T. Li, H. Zhu, Y. N. Wang, X. L. Ren, B. L. Hu, C. P. Huang, J. H. Zhu and X. Zhang, Hyperoside inhibits lipopolysaccharide-induced inflammatory responses in microglial cells via p38 and NFκB pathways, Int. Immunopharmacol., 2017, 50, 14–21.
22. T. Zhang, L. Zhao and R. Kuang, Impact of Hyperoside on the Blood-Brain Barrier in Rats with Bacterial Meningitis through the MicroRNA-155/Brain-Derived Neurotrophic Factor Pathway, Neuroendocrinology, 2025, 115, 704–718.
23. K. Wagener, T. Grunert, I. Prunner, M. Ehling-Schulz and M. Drillich, Dynamics of uterine infections with Escherichia coli, Streptococcus uberis and Trueperella pyogenes in post-partum dairy cows and their association with clinical endometritis, Vet. J., 2014, 202, 527–532.
24. A. Song, R. Cheng, J. Jiang, H. Qu, Z. Wu, F. Qian, S. Shen, L. Zhang, Z. Wang, W. Zhao and Y. Lou, Antidepressant-like effects of hyperoside on chronic stress-induced depressive-like behaviors in mice: Gut microbiota and short-chain fatty acids, J. Affective Disord., 2024, 354, 356–367.
25. C. Zhao, L. Bao, M. Qiu, L. Feng, L. Chen, Z. Liu, S. Duan, Y. Zhao, K. Wu, N. Zhang, X. Hu and Y. Fu, Dietary Tryptophan-Mediated Aryl Hydrocarbon Receptor Activation by the Gut Microbiota Alleviates Escherichia coli-Induced Endometritis in Mice, Microbiol. Spectrum, 2022, 10, e0081122.
26. B. Routy, E. Le Chatelier, L. Derosa, C. P. M. Duong, M. T. Alou, R. Daillère, A. Fluckiger, M. Messaoudene, C. Rauber, M. P. Roberti, M. Fidelle, C. Flament, V. Poirier-Colame, P. Opolon, C. Klein, K. Iribarren, L. Mondragón, N. Jacquelot, B. Qu, G. Ferrere, C. Clémenson, L. Mezquita, J. R. Masip, C. Naltet, S. Brosseau, C. Kaderbhai, C. Richard, H. Rizvi, F. Levenez, N. Galleron, B. Quinquis, N. Pons, B. Ryffel, V. Minard-Colin, P. Gonin, J. C. Soria, E. Deutsch, Y. Loriot, F. Ghiringhelli, G. Zalcman, F. Goldwasser, B. Escudier, M. D. Hellmann, A. Eggermont, D. Raoult, L. Albiges, G. Kroemer and L. Zitvogel, Gut microbiome influences efficacy of PD-1-based immunotherapy against epithelial tumors, Science, 2018, 359, 91–97.
27. J. Yang, S. Liu, Q. Zhao, X. Li and K. Jiang, Gut microbiota-related metabolite alpha-linolenic acid mitigates intestinal inflammation induced by oral infection with Toxoplasma gondii, Microbiome, 2023, 11, 273.
28. J. Hu, L. Ma, Y. Nie, J. Chen, W. Zheng, X. Wang, C. Xie, Z. Zheng, Z. Wang, T. Yang, M. Shi, L. Chen, Q. Hou, Y. Niu, X. Xu, Y. Zhu, Y. Zhang, H. Wei and X. Yan, A Microbiota-Derived Bacteriocin Targets the Host to Confer Diarrhea Resistance in Early-Weaned Piglets, Cell Host Microbe, 2018, 24, 817–832.
29. H. Bao, J. Cong, Q. Qu, S. He, D. Zhao, H. Zhao, S. Yin and D. Ma, Rosiglitazone alleviates LPS-induced endometritis via suppression of TLR4-mediated NF-κB activation, PLoS One, 2024, 19, e0280372.
30. L. A. Dieleman, M. J. Palmen, H. Akol, E. Bloemena, A. S. Peña, S. G. Meuwissen and E. P. Van Rees, Chronic experimental colitis induced by dextran sulphate sodium (DSS) is characterized by Th1 and Th2 cytokines, Clin. Exp. Immunol., 1998, 114, 385–391.
31. P. Wang, J. Zhang, Y. Chen, H. Zhong, H. Wang, J. Li, G. Zhu, P. Xia, L. Cui, J. Li, J. Dong, Q. Gao and X. Meng, Colibactin in avian pathogenic Escherichia coli contributes to the development of meningitis in a mouse model, Virulence, 2021, 12, 2382–2399.
32. H. Wu, G. Zhao, K. Jiang, C. Li, C. Qiu and G. Deng, Engeletin Alleviates Lipopolysaccharide-Induced Endometritis in Mice by Inhibiting TLR4-mediated NF-κB Activation, J. Agric. Food Chem., 2016, 64, 6171–6178.
33. A. Lavelle and H. Sokol, Gut microbiota-derived metabolites as key actors in inflammatory bowel disease, Nat. Rev. Gastroenterol. Hepatol., 2020, 17, 223–237.
34. M. A. F. Yahaya, A. R. A. Bakar, J. Stanslas, N. Nordin, M. Zainol and M. Z. Mehat, Insights from molecular docking and molecular dynamics on the potential of vitexin as an antagonist candidate against lipopolysaccharide (LPS) for microglial activation in neuroinflammation, BMC Biotechnol., 2021, 21, 38.
35. M. Talib, M. A. Nabeel, S. U. Haq, M. S. Waqas, H. Jamil, A. I. Aqib, A. Muneer, D. Fouad and F. S. Ataya, Recent Trends in S. aureus and E. coli-Based Endometritis, and the Therapeutic Evaluation of Sodium Alginate-Based Antibiotics and Nanoparticles, Gels, 2023, 9, 955.
36. L. Cao, S. Gao, J. Liu, J. Wang and R. Qin, Selenomethionine protects against Escherichia coli-induced endometritis by inhibiting inflammation and necroptosis via regulating the PPAR-γ/NF-κB pathway, Chem.-Biol. Interact., 2023, 379, 110532.
37. A. D. Mackeen, R. E. Packard, E. Ota and L. Speer, Antibiotic regimens for postpartum endometritis, Cochrane Database Syst. Rev., 2015, 2015, CD001067.
38. K. Wang, K. Wang, J. Wang, F. Yu and C. Ye, Protective Effect of Clostridium butyricum on Escherichia coli-Induced Endometritis in Mice via Ameliorating Endometrial Barrier and Inhibiting Inflammatory Response, Microbiol. Spectrum, 2022, 10, e0328622.
39. T. Xie, J. Yuan, L. Mei, P. Li and R. Pan, Hyperoside ameliorates TNF-α-induced inflammation, ECM degradation and ER stress-mediated apoptosis via the SIRT1/NF-κB and Nrf2/ARE signaling pathways in vitro, Mol. Med. Rep., 2022, 26, 260.
40. J. Wang, N. Zhu, X. Su, Y. Gao and R. Yang, Gut-Microbiota-Derived Metabolites Maintain Gut and Systemic Immune Homeostasis, Cells, 2023, 12, 793.
41. A. Agus, K. Clément and H. Sokol, Gut microbiota-derived metabolites as central regulators in metabolic disorders, Gut, 2021, 70, 1174–1182.
42. C. Di Salvo, V. D'Antongiovanni, L. Benvenuti, A. d'Amati, C. Ippolito, C. Segnani, C. Pierucci, G. Bellini, T. Annese, D. Virgintino, R. Colucci, L. Antonioli, M. Fornai, M. Errede, N. Bernardini and C. Pellegrini, Lactiplantibacillus plantarum HEAL9 attenuates cognitive impairment and progression of Alzheimer's disease and related bowel symptoms in SAMP8 mice by modulating microbiota-gut-inflammasome-brain axis, Food Funct., 2024, 15, 10323–10338.
43. J. F. Cryan and T. G. Dinan, Mind-altering microorganisms: the impact of the gut microbiota on brain and behaviour, Nat. Rev. Neurosci., 2012, 13, 701–712.
44. H. Tilg, T. E. Adolph and M. Trauner, Gut-liver axis: Pathophysiological concepts and clinical implications, Cell Metab., 2022, 34, 1700–1718.
45. A. T. Dang and B. J. Marsland, Microbes, metabolites, and the gut-lung axis, Mucosal Immunol., 2019, 12, 843–850.
46. G. Xue, Z. Zheng, X. Liang, Y. Zheng and H. Wu, Uterine Tissue Metabonomics Combined with 16S rRNA Gene Sequencing To Analyze the Changes of Gut Microbiota in Mice with Endometritis and the Intervention Effect of Tau Interferon, Microbiol. Spectrum, 2023, 11, e0040923.
47. S. D. George, O. T. Van Gerwen, C. Dong, L. G. V. Sousa, N. Cerca, J. H. Elnaggar, C. M. Taylor and C. A. Muzny, The Role of Prevotella Species in Female Genital Tract Infections, Pathogens, 2024, 13, 364.
48. Z. E. Ilhan, P. Łaniewski, A. Tonachio and M. M. Herbst-Kralovetz, Members of Prevotella Genus Distinctively Modulate Innate Immune and Barrier Functions in a Human Three-Dimensional Endometrial Epithelial Cell Model, J. Infect. Dis., 2020, 222, 2082–2092.
49. Y. Liu, N. Y. Fatheree, N. Mangalat and J. M. Rhoads, Human-derived probiotic Lactobacillus reuteri strains differentially reduce intestinal inflammation, Am. J. Physiol.: Gastrointest. Liver Physiol., 2010, 299, G1087–G1096.
50. E. V. Marietta, J. A. Murray, D. H. Luckey, P. R. Jeraldo, A. Lamba, R. Patel, H. S. Luthra, A. Mangalam and V. Taneja, Suppression of inflammatory arthritis by human gut–derived Prevotella histicola in humanized mice, Arthritis Rheumatol., 2016, 68, 2878–2888.
51. E. Sorokina, A. Pautova, O. Fatuev, V. Zakharchenko, A. Onufrievich, A. Grechko, N. Beloborodova and E. Chernevskaya, Promising Markers of Inflammatory and Gut Dysbiosis in Patients with Post-COVID-19 Syndrome, J. Pers. Med., 2023, 13, 971.
52. Y. Cui, J. Matias, X. Tang, B. Cibichakravarthy, K. DePonte, M. J. Wu and E. Fikrig, Metabolomic changes associated with acquired resistance to Ixodes scapularis, Ticks Tick-Borne Dis., 2024, 15, 102279.
53. A. A. Levchuk, R. A. Faron, S. A. Khrustalev and M. O. Raushenbakh, Effect of the carcinogenic tyrosin metabolite, p-hydroxyphenyllactic acid, on ascorbic acid levels in mouse organs and blood, Bull Exp. Biol. Med., 1986, 102, 1417–1418.
54. J. Cai, L. Sun and F. J. Gonzalez, Gut microbiota-derived bile acids in intestinal immunity, inflammation, and tumorigenesis, Cell Host Microbe, 2022, 30, 289–300.
55. X. Su, Y. Gao and R. Yang, Gut Microbiota-Derived Tryptophan Metabolites Maintain Gut and Systemic Homeostasis, Cells, 2022, 11, 2296.
56. S. An, Y. Yao, J. Wu, H. Hu, J. Wu, M. Sun, J. Li, Y. Zhang, L. Li, W. Qiu, Y. Li, Z. Deng, H. Fang, S. Gong, Q. Huang, Z. Chen and Z. Zeng, Gut-derived 4-hydroxyphenylacetic acid attenuates sepsis-induced acute kidney injury by upregulating ARC to inhibit necroptosis, Biochim. Biophys. Acta, Mol. Basis Dis., 2024, 1870, 166876.
57. S. Xiong, C. Xu, C. Yang, H. Luo, J. Xie, B. Xia, Z. Zhang, Y. Liao, C. Li, Y. Li and L. Lin, FuKe QianJin capsule alleviates endometritis via inhibiting inflammation and pyroptosis through modulating TLR4/NF-κB /NLRP3 pathway, J. Ethnopharmacol., 2025, 337, 118962.
58. H. Wu, A. Dai, X. Chen, X. Yang, X. Li, C. Huang, K. Jiang and G. Deng, Leonurine ameliorates the inflammatory responses in lipopolysaccharide-induced endometritis, Int. Immunopharmacol., 2018, 61, 156–161.
59. A. C. Feng, B. J. Thomas, P. K. Purbey, F. M. de Melo, X. Liu, A. E. Daly, F. Sun, J. H. Lo, L. Cheng, M. F. Carey, P. O. Scumpia and S. T. Smale, The transcription factor NF-κB orchestrates nucleosome remodeling during the primary response to Toll-like receptor 4 signaling, Immunity, 2024, 57, 462–477.
60. Y. Chen, Y. F. Zhao, J. Yang, H. Y. Jing, W. Liang, M. Y. Chen, M. Yang, Y. Wang and M. Y. Guo, Selenium alleviates lipopolysaccharide-induced endometritis via regulating the recruitment of TLR4 into lipid rafts in mice, Food Funct., 2020, 11, 200–210.
61. T. Yokoyama, A. Komori, M. Nakamura, Y. Takii, T. Kamihira, S. Shimoda, T. Mori, S. Fujiwara, M. Koyabu, K. Taniguchi, H. Fujioka, K. Migita, H. Yatsuhashi and H. Ishibashi, Human intrahepatic biliary epithelial cells function in innate immunity by producing IL–6 and IL–8 via the TLR4−NF–κB and–MAPK signaling pathways, Liver Int., 2006, 26, 467–476.
62. F. Turroni, M. Ventura, L. F. Buttó, S. Duranti, P. W. O’Toole, M. O. Motherway and D. van Sinderen, Molecular dialogue between the human gut microbiota and the host: a Lactobacillus and Bifidobacterium perspective, Cell Mol. Life Sci., 2014, 71, 183–203.
63. P. D. Sobolev, N. A. Burnakova, N. V. Beloborodova, A. I. Revelsky and A. K. Pautova, Analysis of 4-hydroxyphenyllactic acid and other diagnostically important metabolites of α-amino acids in human blood serum using a validated and sensitive ultra-high-pressure liquid chromatography-tandem mass spectrometry method, Metabolites, 2023, 13, 1128.
64. S. Jiang, H. Li, L. Zhang, W. Mu, Y. Zhang, T. Chen, J. Wu, H. Tang, S. Zheng, Y. Liu, Y. Wu, X. Luo, Y. Xie and J. Ren, Generic Diagramming Platform (GDP): a comprehensive database of high-quality biomedical graphics, Nucleic Acids Res., 2025, 53, D1670–D1676.

---

Footnote

† These authors have contributed equally to this work.

---