Food-grade TiO₂ impairs intestinal mucus barrier via disrupting the gut microbiota-ILA-mucin sulfation axis: novel insights and dietary intervention strategies

Abstract

The use of titanium dioxide (TiO₂) as a food additive has persistently elicited concerns about its potential toxicity, primarily attributed to the significant presence of nanoparticles. Studies have recently demonstrated that nano- or micro-sized food-grade TiO₂ (fg-TiO₂) particles can disrupt gut microbial balance and weaken the intestinal barrier; however, the underlying mechanisms remain poorly understood. Furthermore, research on dietary interventions for repairing fg-TiO₂-induced intestinal injury is limited. This study delved into the role of the interactions among gut microbes, indole-3-lactic acid (ILA) and mucin sulfation in fg-TiO₂-induced intestinal mucosal barrier damage through multi-omics analysis, and revealed the protective effects of quercetin. Prolonged oral administration of fg-TiO₂ at doses pertinent to human exposure resulted in notable intestinal inflammation and mucosal barrier damage through diminishing mucin sulfation. Further analysis demonstrated that fg-TiO₂ caused significant gut microbiota dysbiosis and metabolite changes. It was found that Lactobacillus and its metabolite ILA, an aryl hydrocarbon receptor (AHR) agonist, was significantly downregulated following oral ingestion of fg-TiO₂, which decreased the activation of AHR and ultimately led to a loss in mucin sulfation in colon tissues. Notably, experiments involving fecal microbiota transplantation (FMT) and ILA supplementation indicated gut microbial shifts and the consequent decrease in colonic ILA levels were accountable for the detrimental effects of fg-TiO₂ on mucin sulfation and intestinal barrier integrity. Moreover, this study found that dietary intervention with quercetin could effectively reverse the damage to the intestinal mucosal barrier induced by fg-TiO₂ through targeting gut microbiota-ILA-mucin sulfation axis. This research uncovered the adverse impacts of fg-TiO₂ on gut homeostasis and indicates the potential of quercetin to combat the intestinal toxicity of fg-TiO₂. These findings enhanced our comprehension of the safety profile of fg-TiO₂ and proposed a nutritional approach to mitigate the health risks associated with fg-TiO₂ exposure.

Graphical Abstract

Supplementary Information

The online version contains supplementary material available at 10.1186/s12951-026-04679-6.

Background

Titanium dioxide (TiO₂) is extensively utilized as a food additive, owing to its high refractive index and brightness [1]. Over 3,500 food and beverage labels report the inclusion of food-grade TiO₂ (fg-TiO₂) [2]. Furthermore, it is employed in various personal care products and pharmaceuticals [3]. This widespread application has resulted in significant oral exposure to TiO₂. Statistical data shows that the daily TiO₂ exposure for children (infants to 9 years old) averages between 0.4 and 10.4 mg per kilogram of body weight, while for adults (18–64 years old), it averages between 0.6 and 6.8 mg/kg [4]. Approximately 36% of fg-TiO₂ particles are at the nanoscale, prompting concerns regarding their potential toxicity [5, 6]. Despite the European Union’s prohibition of its use in food in 2022 due to concerns about genotoxicity risks, both the United States and China continue to permit use of TiO₂ within regulated limits [1, 7]. Several investigations have revealed that fg-TiO₂ can provoke notable genotoxic effects and immune responses [7], and it may be involved in the progression of chronic diseases like obesity and atherosclerosis [8, 9]. Given that more than 99% of fg-TiO₂ particles are unable to traverse the intestinal barrier and are subsequently excreted in feces [10], the intestine and its resident gut microbiota represent the primary exposure sites for fg-TiO₂. Animal studies have indicated that fg-TiO₂ particles can diminish the expression of MUC2 in the intestine, decrease the thickness of the mucus layer, and consequently compromise the integrity of gut barrier [11, 12]. Moreover, numerous studies have shown that oral ingestion of fg-TiO₂ particles directly interacts with the gut microbiota, significantly altering its structure and composition [13]. Research has revealed oral intake of fg-TiO₂ particles causes a potential decline in the abundances of beneficial microbes, such as Lactobacilli and Bifidobacteria, alongside a rise in certain potentially pathogenic bacteria, such as Ruminococcus, thereby disrupting the microecological balance [8, 14]. However, the precise mechanisms through which fg-TiO₂ affects the intestinal mucosal barrier are not well defined, and the possible contribution of microbial dysbiosis has yet to be comprehensively elucidated.

The intestinal flora exerts significant impact on the structure and function of the mucosal barrier through multiple mechanisms [15]. Firstly, specific microbial constituents can directly stimulate mucus secretion and modulate the post-translational modifications of mucins [16, 17]. Secondly, certain symbiotic bacteria are capable of fermenting dietary fiber to generate short-chain fatty acids (SCFAs) like butyric acid, which can directly target goblet cells, thereby augmenting MUC2 protein synthesis and reinforcing the mucosal integrity [18]. Furthermore, it was demonstrated that tryptophan metabolites produced by certain gut microbes, including indole-3-acetic acid (IAA) and indole-3-lactic acid (ILA), can activate aryl hydrocarbon receptors (AHR) [19]. This activation results in the upregulation of crucial enzymes essential for mucin sulfation, thereby promoting mucin sulfation and enhancing the function of the intestinal mucosal barrier [20]. Sulfated mucin represents a significant form of O-linked glycan modification, characterized by the addition of sulfate groups, which modulates its physicochemical properties and biological functions. This process is integral to mucosal barrier protection, pathogen defense, and the development of diseases such as ulcerative colitis, cystic fibrosis, and colon cancer [21, 22]. The sulfation of mucin enhances its resistance to degradation by microbial enzymes, thus preserving the integrity of the mucus layer [23]. Recent research reveals that enhanced mucin sulfation can ameliorate intestinal inflammation and mitigate obesity via modulating the gut microbiota [24]. Serving as a significant interface for the interactions between host and microbiota, the dynamic equilibrium of mucin sulfation directly influences the integrity of the intestinal barrier [20]. Some studies have explored the influence of fg-TiO₂ particles on intestinal flora and tryptophan metabolism [14, 25, 26]. However, the influence of these effects on mucus sulfation in fg-TiO₂-induced intestinal mucosal barrier damage remains unexplored.

The oral consumption of TiO₂ can substantially disrupt the homeostasis of intestinal microbes and induce damage to the gut barrier, potentially contributing to the onset and progression of various chronic diseases [13]. Consequently, identifying effective strategies to mitigate the intestinal damage induced by fg-TiO₂ holds considerable scientific and societal importance. Quercetin, a natural flavonoid constituting 60–75% of total dietary flavonoids, is present in many foods, including fruits, vegetables, and tea [27]. Quercetin and its derivatives possess the capacity to alter numerous cellular targets and signal pathways, demonstrating significant antioxidant and anti-inflammatory properties [28]. Recently, dietary supplementation with quercetin has been extensively studied and reported as a therapeutic strategy for various chronic diseases, chiefly due to its role in modulating gut flora and enhancing intestinal mucosal barrier function [29]. Some studies demonstrated that quercetin can augment the amounts of beneficial microorganisms like Lactobacillus and Akkermansia in the gut microbiota of mice, which elevated levels of the SCFAs and tryptophan metabolites, ultimately enhancing the function of the intestinal mucosal barrier [30, 31]. Furthermore, several studies have indicated that quercetin may initiate the AHR signal pathway, thus strengthening gut barrier function [31, 32]. Additionally, research has suggested that quercetin can significantly ameliorate gut barrier damage and intestinal inflammation caused by aged nanoplastic particles [33]. Nonetheless, there is a paucity of research on the potential of quercetin supplementation to mitigate gut microbiota dysbiosis and mucosal barrier compromise induced by fg-TiO₂. Moreover, there have been no reports investigating the role of the interaction between gut microbiota and mucin sulfation in this context.

This study focused on investigating how gut microbes and their metabolites contribute to mucosal barrier damage and inflammation triggered by fg-TiO₂ exposure, utilizing gut microbiota analysis and non-targeted metabolomics techniques. Additionally, the study also explored how quercetin protects against fg-TiO₂-induced intestinal injury and the mechanisms involved. Our findings provided substantial evidence supporting the critical role of the gut microbiota-ILA-AHR-mucin sulfation axis in modulating intestinal mucosal barrier damage by fg-TiO₂ and its dietary intervention with quercetin. The microbial tryptophan metabolite ILA was recognized as a pivotal metabolite with significant functional implications. ILA modulated colonic mucin sulfation levels by regulating the AHR signaling pathway, thereby influencing the detrimental or protective effects of fg-TiO₂ and quercetin on the intestinal mucosal barrier. This study not only enhanced the understanding of the mechanisms underlying fg-TiO₂-induced intestinal injury but also proposed an effective dietary strategy for mitigating intestinal damage associated with fg-TiO₂ exposure.

Experimental

Materials

Fg-TiO₂ (E171) particles (Jianghu Titanium White Product Co., Ltd., China), anatase crystal, with a purity of at least 99%. Quercetin (PS1301, Chengdu Push Bio-technology, China). Vancomycin (V301569, Aladdin, China). Neomycin (N109017, Aladdin, China). Ampicillin (A105483, Aladdin, China). Metronidazole (M109874, Aladdin, China). ILA (I157602, Aladdin, China). Hematoxylin & eosin (H&E) staining kit (G1005, Servicebio, China). Anti-MUC2 antibody (GB11344, Servicebio, China). Anti-PAPSS2 antibody (sc-100801, Santa Cruz, USA). HID-AB staining kit (G2070, Solarbio, China). PrimeScript RT reagent Kit (RR047A, Takara Biotechnology, Japan). TB Green^® Premix Ex Taq™ II (RR820A, Takara Biotechnology, Japan). Interleukin 1β (IL-1β) enzyme-linked immunosorbent assay (ELISA) kit (KT21178, MSKBIO, China). Interleukin 6 (IL-6) ELISA kit (KT99854, MSKBIO, China). Tumor necrosis factor α (TNF-α) ELISA kit (KT99985, MSKBIO, China). Lipopolysaccharide (LPS) ELISA kit (KT37561, MSKBIO, China). MUC2 ELISA kit (KT32461, MSKBIO, China). ILA ELISA kit (YX-23042 M, Sinobestbio, China). AHR ELISA kit (RJ17093, Renjiebio, China). FITC-labeled dextran (FD4, 60842-46-8, Sigma-Aldrich, USA).

Characterization of fg-TiO₂

The morphology of fg-TiO₂ was analyzed using transmission electron microscopy (TEM) (Hitachi, Japan), and the particle size distribution was quantified with Image J software (USA). The crystalline phase of fg-TiO₂ particles was identified using X-ray diffraction (XRD) (Bruker D8 Advance, Germany).

Animals

Six-week-old male C57BL/6J mice, free from specific pathogens, were obtained from Myhalic Biotechnology Co., Ltd. in Wuhan, China. They were kept in conditions with a 12-hour cycle of light and darkness, temperatures set at 22 ± 2 °C, and humidity at 55 ± 5%. All procedures complied with the ethical guidelines of the Institutional Animal Care and Use Committee at Myhalic Biotechnology Co., Ltd. in Wuhan, China (Approval No. HLK-20250109-003).

Experiment 1: Following one-week acclimatization, the mice were assigned randomly into three groups: a control group, a fg-TiO₂ group, and a fg-TiO₂Q group (6 mice per group). The control group received deionized water via oral gavage; the fg-TiO₂ group was given a daily dose of 40 mg/kg fg-TiO₂ particles via oral gavage over a period of 10 weeks; the fg-TiO₂Q group was administered 40 mg/kg fg-TiO₂ particles and 50 mg/kg quercetin each day via oral gavage for 10 weeks. Upon the conclusion of the experiments, the colons were entirely resected, and their lengths were subsequently measured.

Experiment 2: Mice in the ControlF, fg-TiO₂F, and fg-TiO₂QF groups (6 mice per group) received drinking water containing 1 g/L ampicillin, 1 g/L metronidazole, 1 g/L neomycin, and 0.5 g/L vancomycin for two weeks to deplete gut bacteria. Then ControlF, fg-TO₂F, and fg-TO₂QF mice received fecal supernatants via oral gavage from the control, fg-TiO₂, and fg-TO₂Q groups of mice respectively twice a week with a dosage of 100 µL per 10 g of body weight per mice. The fecal supernatants were prepared as described [34]. Briefly, the fresh fecal samples were mixed with ice-cold sterile anaerobic saline at a ratio of 100 mg of feces per 1 mL of saline. The mixture was allowed to steep in the cold saline for 5 min, then vigorously vortexed for 10 s. Subsequently, the resulting suspension was centrifuged (800 g, 4 °C) for 3 min to retrieve the supernatant for the following fecal microbiota transplantation (FMT). All procedures were performed in a laminar flow hood using sterile equipment and reagents. FMT was processed within 30 min of collection to maintain microbial viability.

Experiment 3: Mice were allocated into three groups: a control group, a fg-TiO₂ group, and a fg-TiO₂+ILA group with 6 mice per group. The control group received deionized water via oral gavage; the fg-TiO₂ group was given 40 mg/kg fg-TiO₂ daily via oral gavage for 10 weeks; the fg-TiO₂+ILA group was administered 40 mg/kg fg-TiO₂ particles and 10 mg/kg ILA via oral gavage for 10 weeks.

Histopathology

To evaluate inflammation, distal colon tissues were preserved in 4% paraformaldehyde for 12 h, embedded in paraffin, and stained with H&E. Immunofluorescence staining was utilized to observe the MUC2 protein content in the distal colon using an anti-MUC2 antibody (1:500). Additionally, the staining of colonic sections with an HID-AB kit resulted in the sulfated acidic mucins being visualized in varying shades from purple-brown to brown-black. Immunohistochemical staining was conducted to assess the protein levels of PAPSS2 in colonic tissues using an anti-PAPSS2 antibody (1:500). Imaging was performed with optical and fluorescent microscopes (NIKON, Japan), and quantitative analysis was carried out using ImageJ software (USA).

Quantitative real-time PCR (qRT-PCR)

The RNA extracted utilizing the RNAiso Plus reagent was subsequently quantified using a NanoDrop 2000 spectrophotometer (Thermo Scientific, USA). Complementary DNA (cDNA) synthesis was conducted in accordance with the PrimeScript RT reagent Kit protocol. The assay was performed via qRT-PCR with the TB Premix Ex TaqTM II on the 7500 Real-Time PCR System (Applied Biosystems, USA). In this study, β-actin serves as a reference gene for normalizing gene expression levels. The primer sequences used are provided in Table S1.

ELISA

Colonic tissues and fecal samples were weighed and homogenized in pre-cooled phosphate-buffered saline (PBS) with a tissue grinder (Servicebio, China). The concentrations of fecal LPS and ILA were determined using ELISA kits. The colonic MUC2 and AHR content were also examined with ELISA kits. Additionally, serum concentrations of LPS, IL-1β, IL-6, and TNF-α were also measured with ELISA kits.

Intestinal permeability test

The test was performed in accordance with established protocols [8]. In summary, after a 3-hour fast, every mouse received 15 mg of FITC-Dextran orally. Four hours following administration, serum samples were collected to evaluate FITC intensity and quantify FITC-dextran levels. This quantification was achieved by referencing a standard curve generated through the dissolution of FITC-dextran in serum.

Microbiota analysis

Microbial DNA was extracted using the Omega Bio-tek Stool DNA Kit (Omega, USA) and quantified with a NanoDrop 2000 spectrophotometer (Thermo Scientific, USA). The V4 region of 16 S rDNA was amplified and evaluated with the AxyPrep DNA gel extraction kit. It was then quantified and diluted for sequencing on the MiSeq per established protocol. Purified amplicons underwent sequencing on the Illumina MiSeq PE250 platform to obtain paired-end reads. Standard parameters were applied to eliminate reads with low quality, and overlapped paired-end reads were merged into consensus sequences in FLASH (version 1.2.11). Using the UPARSE algorithm, operational taxonomic units (OTUs) were grouped at a 97% similarity threshold. Subsequently, a profiling table was created by consolidating tags to evaluate the relative abundance of taxa at different taxonomic levels. For α and β diversity analyses, the software QIIME 2.0 was employed. Additionally, Principal coordinates analysis (PCoA) utilizing Weighted Unifrac distances was executed using R (version 3.1.1). Using the ImageGP online platform, linear discriminant analysis of effect size (LEfSe) was conducted.

Metabolomics

After gradually thawing on the ice, fecal samples with an exact mass of 25 mg were weighted. Subsequently, 800 µL of an extraction solution (methanol, acetonitrile, and water in a 2:2:1 ratio) and 10 µL of an internal standard were incorporated. After homogenization, the samples were incubated in a 4 °C water bath for 10 min, after which they were placed in a freezer at -20 °C for an hour. 600 µL supernatant was collected by centrifuging at 4 °C and 25,000 g for 15 min, followed by desiccation using a freeze vacuum concentrator. To reconstitute the dried sample, 600 µL of a solution made of methanol and ultrapure water in a 9:1 ratio was added. Samples were incubated in a 4 °C water bath for 10 min and then centrifuged at 4 °C and 25,000 g for 15 min. The collected supernatant underwent liquid chromatography-mass spectrometry (LC-MS) analysis with a Q Exactive mass spectrometer connected to a Waters UPLC I-Class Plus system. Chromatographic separations followed a previous method [35]. Both positive and negative ion modes were used to collect mass spectrometry data, which were processed using Compound Discoverer 3.3, integrating with mzCloud, BMDB, and ChemSpider databases to produce a dataset of metabolite peak areas along with identification outcomes.

Statistical analysis

The data are presented as the mean ± standard error of the mean (s.e.m). Comparisons of two groups utilized a two-tailed Student’s t-test, while a one-way analysis of variance (ANOVA) followed by Tukey’s post hoc test was used for multiple group comparisons. Data visualization and statistical analyses were conducted using GraphPad Prism (version 9.0).

Results

Evaluation of intestinal inflammation with fg-TiO₂ and quercetin treatment

In this study, fg-TiO₂ exhibited an average particle size of 110.9 ± 35.1 nm, as shown by TEM analysis (Fig. S1A). The particles varied in size from 37.1 nm to 262.4 nm, with about 38.2% having at least one dimension under 100 nm (Fig. S1B). XRD analysis indicated that the fg-TiO₂ particles utilized in this study was present in the anatase crystal form (Fig. S1C). These findings aligned with prior researches concerning the characterization of TiO₂ particles sourced from food [6, 36].

The intestine serves as the primary target organ for oral exposure to fg-TiO₂. Consequently, we initially assessed the impact of oral intake of fg-TiO₂ particles and quercetin supplementation at doses equivalent to those consumed by humans on the colon tissue. The findings indicated that sustained exposure to 40 mg/kg of fg-TiO₂ particles per day over a period of 10 weeks did not notably influence the body weight, food intake, and water intake of mice (Fig. S2). However, it did result in a significant reduction in colon length, an effect that was mitigated by quercetin treatment in the mice (Fig. S3). The histological analysis of distal colon tissues, as indicated by H&E staining, demonstrated that exposure to fg-TiO₂ resulted in pronounced inflammatory responses associated with intestinal damage (Fig. 1A). Conversely, administration of quercetin effectively ameliorated the colonic inflammatory responses induced by fg-TiO₂ (Fig. 1A). Subsequent qRT-PCR analysis revealed that treatment with fg-TiO₂ led to a significant upregulation in the mRNA expression levels of the inflammatory cytokines Il-1β, Il-6, and Tnf-α in colon tissues, with increases of 70.7%, 81.2%, and 64.1%, respectively (Fig. 1B-1D). Following quercetin administration, the Il-1β, Il-6, and Tnf-α mRNA expression levels were restored to levels similar to those in the control group (Fig. 1B-1D). Further ELISA results indicated that fg-TiO₂ induced an obvious 34.6% rise in the fecal LPS content in mice, while quercetin resulted in a notable 19.2% reduction in fecal LPS levels in the fg-TiO₂-treated mice (Fig. 1E). This alteration in fecal LPS concentration might be a critical factor contributing to intestinal inflammation. The quantitative assessments of serum LPS content demonstrated that fg-TiO₂ exposure significantly elevated serum LPS levels in mice, while it was effectively mitigated by the administration of quercetin (Fig. 1F). The marked increase in serum concentration of IL-1β, IL-6, and TNF-α following exposure to fg-TiO₂ suggested that fg-TiO₂ induced low-grade systemic inflammation (Fig. 1G-1I). However, treatment with quercetin ameliorated this condition (Fig. 1G-1I). Collectively, these findings indicated that prolonged consumption of fg-TiO₂ caused both intestinal and systemic inflammation, whereas supplementation with the dietary flavonoid quercetin could effectively mitigate the inflammation induced by fg-TiO₂.

Fig. 1.

Effects of fg-TiO₂ and quercetin on intestinal inflammation. (A) Representative H&E staining images of distal colon tissues. Scale bar: 100 μm. (B-D) Gene expression levels of Il-1β, Il-6, and Tnf-α in colon tissues. (E) Fecal LPS concentration. (F) Serum LPS concentration. (G-I) Serum concentration of IL-1β, IL-6, and TNF-α. Data are shown as mean ± s.e.m., n = 6 in each group, *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001, n.s., non-significant

Impact of fg-TiO₂ and quercetin on gut mucosal barrier function and mucin sulfation

Subsequently, we assessed the impact of fg-TiO₂ and quercetin on the intestinal mucosal barrier. The findings from MUC2 immunofluorescence staining revealed that fg-TiO₂ exposure decreased the colonic MUC2 content, whereas quercetin treatment significantly elevated MUC2 levels in mice (Fig. 2A). Statistical analysis of crypt depth demonstrated a significant reduction in colonic crypt depth following fg-TiO₂ exposure, while quercetin treatment facilitated the restoration of crypt depth (Fig. 2B). The qRT-PCR and ELISA analyses of MUC2 expression in colon tissues suggested that quercetin mitigated the reduction in MUC2 content induced by fg-TiO₂ particles (Fig. 2C and 2D). Notably, the results from the intestinal permeability test indicated that fg-TiO₂ treatment resulted in a 69.1% increase in serum FITC-Dextran content, which was effectively reduced by quercetin supplementation (Fig. 2E). Collectively, these results suggested that long-term consumption of fg-TiO₂ induced obvious damage to the intestinal mucosal barrier, whereas quercetin supplementation could significantly ameliorate this damage.

Fig. 2.

Effects of fg-TiO₂ and quercetin on intestinal mucosal barrier function and mucin sulfation. (A) Representative MUC2 immunofluorescence staining images of distal colon tissues. Scale bar: 100 μm. (B) Crypt depth. (C) Muc2 mRNA expression levels in colon tissues. (D) Colonic MUC2 content. (E) Intestinal permeability determined by serum FITC-dextran levels. (F) Representative images of distal colonic sections stained with HID-AB. Scale bar: 200 μm (up) and 100 μm (down). (G) Sulfomucin positive cell number per crypt. (H-J) Papss2, Slc35b3, and GlcNAc6st2 mRNA expression levels in colon tissues. (K, L) Representative PAPSS2 immunohistochemical staining images of distal colon tissues and statistics. Data are presented as mean ± s.e.m., n = 6 in each group in B-E and G-J, n = 5 in each group in L, *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001

Sulfated polysaccharides constitute a crucial component of mucin. Compromise in mucin sulfation has been documented to result in substantial intestinal barrier damage and inflammation [20]. Hence, we hypothesized that the mucosal barrier damage induced by fg-TiO₂ exposure might be associated with loss in mucin sulfation. To verify this hypothesis, we employed HID-AB staining, a technique that facilitated the visualization of the sulfomucin subtype of mucins. The findings revealed a significant reduction in sulfated mucin within the colonic tissues of mice exposed to fg-TiO₂ compared to control mice (Fig. 2F). Conversely, supplementation with quercetin led to an increase in sulfated mucin content in colonic tissues (Fig. 2F). Statistical analysis indicated a 61.1% decrease in the number of sulfomucin-positive cells per crypt following fg-TiO₂ exposure, whereas mice supplemented with quercetin exhibited an 89.3% increase in the sulfomucin-positive cell number per crypt compared to the only fg-TiO₂-treated mice (Fig. 2G). Furthermore, we analyzed in detail the gene expression levels of key enzymes involved in the mucin sulfation process, specifically Papss2, Slc35b3, and GlcNAc6st. Our findings revealed a significant reduction of Papss2, Slc35b3, and GlcNAc6st mRNA expression levels in the colon tissues of mice treated with fg-TiO₂, as opposed to the control mice (Fig. 2H-2J). Notably, administration with quercetin effectively restored the mRNA expression levels of these enzymes in mice exposed to fg-TiO₂ (Fig. 2H-2J). PAPSS2, recognized as the rate-limiting enzyme for mucin sulfation [20], exhibited a marked decrease of 48.5% in expression as determined by immunohistochemical staining following fg-TiO₂ exposure (Fig. 2K and 2L). However, quercetin supplementation on the basis of fg-TiO₂ exposure resulted in a 59.6% increase in PAPSS2 expression levels (Fig. 2K and 2L). These findings indicated that fg-TiO₂ exposure compromised the function of the gut mucosal barrier through partially diminishing mucin sulfation. Conversely, supplementation with quercetin appeared to mitigate this adverse effect by facilitating the restoration of sulfated mucin levels.

Alterations of gut microbiota and fecal metabolites under fg-TiO₂ and quercetin exposure

Intestinal flora is closely linked to the health of the mucosal barrier and the sulfation process of mucins [20, 37]. To examine the impact of fg-TiO₂ and quercetin treatment on gut microbes and its metabolites, we conducted microbiome and untargeted metabolomics analyses. The findings revealed that the α diversity of intestinal microbes was minimally influenced by both fg-TiO₂ and quercetin (Fig. 3A, S4). However, PCoA demonstrated that the clustering pattern of the microbial community in the fg-TiO₂ group was significantly different from that of the control group, whereas the fg-TiO₂Q group clustered similarly to the control group (Fig. 3B). This suggested that fg-TiO₂ altered the composition of the intestinal flora, while quercetin aid in restoring the fg-TiO₂-induced imbalance in the gut microbiota. This observation was further corroborated by horizontal structural analysis at both the phylum and genus levels (Fig. S5). Subsequent statistical analysis indicated that the relative proportion of Firmicutes boosted by 131.3% in the fg-TiO₂ group relative to the control group, whereas it declined by 44.2% in the fg-TiO₂Q group relative to the fg-TiO₂ group (Fig. 3C). Exposure to fg-TiO₂ resulted in a significant reduction in the relative abundance of Bacteroidetes, decreasing from 73.8% ± 11.5% to 56.3% ± 12.1%. Conversely, treatment with quercetin resulted in a notable rise in the relative proportion of Bacteroidetes, reaching 72.2% ± 6.1% (Fig. 3C). Additionally, the Firmicutes-to-Bacteroidetes ratio in the fg-TiO₂ group increased by 2.0-fold in comparison to the control group, whereas this ratio reduced by 1.4-fold in the fg-TiO₂Q group relative to the fg-TiO₂ group (Fig. 3D). Further analysis using LEfSe highlighted specific variations in the bacterial community structure among the control, fg-TiO₂, and fg-TiO₂Q groups. Compared to the control group, fg-TiO₂ exposure significantly enriched the families Ruminococcaceae, Veillonellaceae, and Desulfovibriaceae, while reducing the relative abundance of Lactobacillaceae (Fig. 3E). Supplementation with quercetin in mice exposed to fg-TiO₂ led to significant reductions in the families Deferribacteraceae, Veillonellaceae, and Desulfovibriaceae, whereas Lactobacilaceae was enriched by quercetin treatment (Fig. 3F). Genus-level statistical analyses demonstrated that fg-TiO₂ exposure caused an increase in the abundances of Oscillospira, Coprococcus, and Butyricimonas, while it resulted in a decrease in the content of Helicobacter, Rikenella, and Lactobacillus (Fig. 3G). In contrast, the fg-TiO₂ group treated with quercetin displayed an elevated relative abundances of microorganisms including Helicobacter, Rikenella, and Lactobacillus, whereas a reduction was observed in Prevotella, Mucispirillum, Oscillospira, Coprococcus, and Butyricimonas (Fig. 3G). Furthermore, quercetin treatment enhanced the presence of Lactobacillus_reuteri and reduced the abundance of Mucispirillum schaedleri in mice exposed to fg-TiO₂ (Fig. S6). The above results suggested that fg-TiO₂ induced notable gut microbiota dysbiosis, while quercetin was beneficial to ameliorating this imbalance.

Fig. 3.

Effects of fg-TiO₂ and quercetin on gut microbiota and metabolites. (A) Chao index. (B) PCoA based on Weighted Unifrac distances. (C) Relative abundances of Bacteroidetes and Firmicutes. (D) Ratio of Firmicutes to Bacteroidetes. (E) LEfSe of gut microbiota between Control and fg-TiO₂ groups. (F) LEfSe of gut microbiota between fg-TiO₂ and fg-TiO₂Q groups. (G) Relative abundances of gut microbes at genus level. (H) Volcano plot of fecal metabolites between Control and fg-TiO₂ groups. (I) Volcano plot of fecal metabolites between fg-TiO₂ and fg-TiO₂Q groups. (J) Relative expression levels of ILA in feces. (K) Fecal concentration of ILA. (L) Ahr mRNA expression levels in colon tissues. (M) Colonic AHR content. Data are presented as mean ± s.e.m., n = 6 in per group, *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001, n.s., non-significant

Next, the impact of fg-TiO₂ and quercetin on fecal metabolites was assessed with non-targeted metabolomics. The results from the volcano plot analysis indicated that fg-TiO₂ significantly elevated the levels of 223 metabolites while reducing the levels of 37 metabolites (Fig. 3H). In contrast with the fg-TiO₂ group, the fg-TiO₂Q group exhibited an increase in 198 metabolites and a decrease in 248 metabolites in fecal samples (Fig. 3I). Spearman correlation analysis revealed that among the top 20 metabolites altered by fg-TiO₂ and quercetin, ILA emerged as the metabolite most strongly associated with intestinal barrier integrity and inflammatory markers (Fig. S7). The concentration of ILA in the fg-TiO₂-exposed group was 2.0 times lower than that in the control group, whereas quercetin treatment led to a significant 7.6-fold increase in ILA levels (Fig. 3J). The absolute quantification of ILA concentrations via ELISA corroborated this finding, demonstrating a significant reduction in ILA content in the feces of the fg-TiO₂ group compared to the control group, with a marked increase following quercetin supplementation (Fig. 3K). It has been reported that ILA activated the AHR to safeguard gut barrier integrity [38, 39]. Therefore, AHR expression levels in colonic tissues were assessed, revealing that fg-TiO₂ consumption caused lower mRNA and protein expression levels of AHR in colon tissues (Fig. 3L and 3M). In contrast, treatment with quercetin reinstated the colonic expression levels of AHR in fg-TiO₂-exposed mice (Fig. 3L and 3M). Collectively, these results revealed that the reduction in intestinal ILA content and the inhibition of AHR induced by fg-TiO₂ can be effectively counteracted by quercetin treatment.

FMT duplicated the impact of fg-TiO₂ and quercetin on gut microbes and ILA levels

To clarify the involvement of gut microorganisms in the impact of fg-TiO₂ and quercetin on intestinal barrier function and mucin sulfation, we conducted FMT experiments subsequent to verifying that a two-week administration of antibiotics (ABX) via drinking water resulted in the depletion of over 93% of the gut bacterial taxa (Fig. S8). Following the determination that FMT exerted a negligible effect on body weight, food intake, and water intake (Fig. S9), we proceeded to conduct a microbiological analysis on the mice subjected to FMT. There were no noticeable variations in α diversity among the mice receiving FMT from the control, fg-TiO₂, and fg-TiO₂Q groups (Fig. 4A, S10). However, PCoA demonstrated that the microbial clustering pattern of fg-TiO₂F mice was distinct from that of ControlF mice (Fig. 4B). In contrast, the microbial clustering pattern of fg-TiO₂QF mice was more similar to that of the ControlF group (Fig. 4B). At the phylum level, the fg-TiO₂F group had a notably increased relative proportion of Firmicutes and a notably decreased relative proportion of Bacteroidetes compared to the ControlF group (Fig. 4C, S11A). Conversely, the fg-TiO₂QF group demonstrated a decline in the presence of Firmicutes and a rise in the presence of Bacteroidetes relative to the fg-TiO₂F group (Fig. 4C, S11A). The ratio of Firmicutes to Bacteroidetes was 1.3 times higher in the fg-TiO₂F group than that in the ControlF group, but in the fg-TiO₂QF group, it was similar to the ratio in the ControlF group (Fig. 4D). LEfSe analysis indicated that, relative to the ControlF group, the fg-TiO₂F group had considerably higher levels of Deferribacteraceae and Lachnospiraceae, with a significant drop in Lactobacillaceae (Fig. 4E). In comparison to the fg-TiO₂F group, the fg-TiO₂QF group showed a substantial increase in the relative abundance of the family Lactobacillaceae, alongside a notable decline in the relative abundances of Odoribacteraceae, Rikenellaceae, Deferribacteraceae, Lachnospiraceae, and Desulfovibrionaceae (Fig. 4F). At the genus level, the fg-TiO₂F group demonstrated lower relative proportions of Prevotella and Lactobacillus, while exhibiting higher relative abundance of Helicobacter compared to the ControlF group (Fig. 4G, S11B). Instead, the fg-TiO₂QF group showed elevated relative content of Prevotella, Lactobacillus, Flexispira, and Butyricimonas, whereas the relative content of Helicobacter, Mucispirillum, and Coprococcus were reduced in comparison to the fg-TiO₂F group (Fig. 4G). Furthermore, the Lactobacillus_reuteri proportion was significantly increased, while the Mucispirillum schaedleri proportion was markedly reduced in the fg-TiO₂QF group compared to the fg-TiO₂F group (Fig. S12). Notably, the fecal ILA concentration in the fg-TiO₂F group was only 48.2% of that observed in the ControlF group (Fig. 4H). In contrast, the fecal ILA level in the fg-TiO₂QF group recovered to 223.4% of that in the fg-TiO₂F group (Fig. 4H). The alterations in AHR expression levels were aligned with the observed changes in fecal ILA concentrations in FMT-recipient mice (Fig. 4I and 4J). These findings suggested that FMT effectively mimicked the impact of fg-TiO₂ and quercetin on intestinal microbial composition and fecal ILA levels.

Fig. 4.

FMT effectively duplicated the effects of fg-TiO₂ and quercetin on gut microbiota structure and ILA levels. (A) Chao index. (B) PCoA based on Weighted Unifrac distances. (C) Relative abundances of Bacteroidetes and Firmicutes after FMT. (D) Ratio of Firmicutes to Bacteroidetes after FMT. (E) LEfSe of gut microbiota between ControlF and fg-TiO₂F groups. (F) LEfSe of gut microbiota between fg-TiO₂F and fg-TiO₂QF groups. (G) Relative abundances of gut microbes at genus level after FMT. (H) Fecal concentration of ILA after FMT. (I) Ahr mRNA expression levels in colon tissues after FMT. (J) Colonic AHR content after FMT. Data are presented as mean ± s.e.m., n = 6 in per group, *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001, n.s., non-significant

FMT mimicked the effects of fg-TiO₂ and quercetin on gut mucosal barrier function

The intestinal barrier function and mucin sulfation were assessed in FMT-recipient mice to elucidate the importance of gut microorganisms. Immunofluorescence staining of colonic MUC2 revealed that the MUC2 content in the fg-TiO₂F group was reduced relative to the ControlF group, whereas the fg-TiO₂QF group exhibited an elevation in the MUC2 content compared to the fg-TiO₂F group (Fig. 5A). A notable reduction in crypt depth in the fg-TiO₂F group, relative to both the ControlF and fg-TiO₂QF groups, suggested disruption of crypt architecture (Fig. 5B). Consistent with the immunofluorescence observations, MUC2 mRNA and protein expression levels in the ControlF and fg-TiO₂QF groups were markedly higher than those in the fg-TiO₂F group (Fig. 5C and 5D). Furthermore, the intestinal permeability test showed that the serum FITC-Dextran concentration in the fg-TiO₂F group was 58.9% greater than that in the ControlF group, whereas the fg-TiO₂QF group exhibited a 31.2% reduction in the serum FITC-Dextran concentration compared to the fg-TiO₂F group (Fig. 5E). These findings indicated that the transplanted gut microbiota could mimic the deleterious impact of fg-TiO₂ particles on the intestinal mucosal barrier, as well as the protective effect of quercetin against fg-TiO₂-induced damage to the intestinal mucosal barrier.

Fig. 5.

Effects of FMT on intestinal barrier function and mucin sulfation. (A) Representative MUC2 immunofluorescence staining images of distal colon tissues after FMT. Scale bar: 100 μm. (B) Crypt depth statistics. (C, D) Muc2 mRNA expression levels and MUC2 content in colon tissues after FMT. (E) FITC-dextran concentration in serum. (F) Representative images of distal colonic sections stained with HID-AB after FMT. Scale bar: 200 μm (up) and 100 μm (down). (G) Sulfomucin positive cell number per crypt. (H-J) Papss2, Slc35b3, and GlcNAc6st2 mRNA expression levels of in colon tissues after FMT. (K, L) Representative PAPSS2 immunohistochemical staining images of distal colon tissues and statistics after FMT. Data are presented as mean ± s.e.m., n = 6 in each group in B-E and G-J, n = 5 in each group in L, *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001

Further HID-AB staining and statistical analyses disclosed a noteworthy 60.0% decrease in the sulfomucin positive cell number per crypt in the fg-TiO₂F group compared to the ControlF group (Fig. 5F and 5G). Conversely, FMT from mice administered fg-TiO₂ plus quercetin resulted in a 111.1% increase in the sulfomucin positive cell number per crypt within colon tissues (Fig. 5F and 5G). Relative to the ControlF group, the colonic mRNA expression levels of Papss2, Slc35b3, and GlcNAc6st in the fg-TiO₂F group were markedly decreased by 61.3%, 66.7%, and 83.3%, respectively (Fig. 5H-5J). In contrast, the fg-TiO₂QF group showed notable increases in the mRNA expression levels of Papss2, Slc35b3, and GlcNAc6st by 2.4 folds, 1.1 folds, and 3.7 folds, respectively, compared to the fg-TiO₂F group (Fig. 5H-5J). In addition, immunohistochemical evaluation and statistical analyses revealed that the PAPSS2 positive staining area in the fg-TiO₂F group was less than 70% of the positive area observed in the ControlF and fg-TiO₂QF groups (Fig. 5K and 5L), aligning with the findings from qRT-PCR analyses (Fig. 5H). These results suggested that the intestinal microbiome was essential for the impact of fg-TiO₂ and quercetin on mucin sulfation.

FMT replicated the impact of fg-TiO₂ and quercetin on intestinal inflammation

Subsequently, the impact of FMT on intestinal inflammation was assessed. A significant shortening of the colon was noted in the fg-TiO₂F group, in comparison to both the ControlF and fg-TiO₂QF groups (Fig. S13), indicated the presence of colonic inflammation in mice receiving FMT from the fg-TiO₂ group. H&E staining showed significant disruption of the colon structure in mice that underwent FMT from the fg-TiO₂ group, whereas the colon structure in the ControlF and fg-TiO₂QF groups remained relatively intact (Fig. 6A). qRT-PCR analysis of inflammatory cytokines demonstrated that the Il-1β, Il-6, and Tnf-α mRNA expression levels in the colon of the fg-TiO₂F group were significantly elevated as opposed to the ControlF group (Fig. 6B-6D). In contrast, the fg-TiO₂QF group had notably lower gene expression levels of these inflammatory cytokines than the fg-TiO₂F group (Fig. 6B-6D). Furthermore, the fg-TiO₂F group exhibited a notable increase in the fecal LPS concentration compared to both the ControlF and fg-TiO₂QF groups (Fig. 6E), which might be associated with FMT-induced alterations in gut microbiota composition. The serum LPS concentration in fg-TiO₂F mice was observed to be 22.9% higher than that in the ControlF mice (Fig. 6F). However, mice in the fg-TiO₂QF group exhibited a significant reduction in serum LPS concentration by 18.1% compared to the fg-TiO₂F group (Fig. 6F). Moreover, IL-1β, IL-6, and TNF-α levels in sera were markedly elevated in the fg-TiO₂F group relative to the other two FMT-recipient groups (Fig. 6G-6I), corroborating the findings from the LPS concentration measurements (Fig. 6F). Together, these outcomes demonstrated that gut microbes mediated the effects of fg-TiO₂ and quercetin on colonic inflammation.

Fig. 6.

Effects of FMT on intestinal inflammation. (A) Representative H&E staining images of distal colon tissues after FMT. Scale bar: 100 μm. (B-D) Gene expression levels of Il-1β, Il-6, and Tnf-α in colon tissues after FMT. (E) Fecal LPS concentration after FMT. (F) Serum LPS concentration after FMT. (G-I) Serum concentration of IL-1β, IL-6, and TNF-α after FMT. Data are shown as mean ± s.e.m., n = 6 in each group, *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001, n.s., non-significant

ILA ameliorated fg-TiO₂-induced gut barrier damage via promoting mucin sulfation

Having identified ILA as a key metabolite significantly altered by both fg-TiO₂ and quercetin, we investigated its potential role in mucin sulfation and the preservation of the intestinal mucosal barrier. To this end, we conducted ILA supplementation experiments. The results indicated that ILA supplementation did not significantly impact the body weight, food intake, and water intake of mice exposed to fg-TiO₂ (Fig. S14). Importantly, ELISA analyses revealed that fecal ILA levels increased by 1.83-fold in the ILA-supplemented group relative to the fg-TiO₂-only group (Fig. 7A). Remarkably, the marked elevation in colonic AHR expression levels of observed in the fg-TiO₂+ILA group, as opposed to the fg-TiO₂ group alone, suggested that ILA supplementation activated the AHR signaling pathway (Fig. 7B, S15). Furthermore, the reduction in MUC2 expression levels induced by fg-TiO₂ was significantly ameliorated following ILA treatment (Fig. 7C and 7D, and S16), corroborated by crypt depth measurements and intestinal permeability tests (Fig. S17, 7E). More importantly, following ILA supplementation, there was a significant increase of more than 2.3-fold in the colonic sulfur-positive cell number per crypt in mice exposed to fg-TiO₂ (Fig. 7F and 7G). qRT-PCR analyses revealed that the colonic Papss2, Slc35b3, and GlcNAc6st mRNA expression levels in the fg-TiO₂+ILA group were elevated by 1.7 folds, 2.0 folds, and 4.8 folds, respectively, in comparison to the fg-TiO₂ group (Fig. 7H-7J). Immunohistochemical staining and subsequent statistical analysis demonstrated that the positive area for PAPSS2 in the colon of mice in the fg-TiO₂+ILA group was 1.5 times larger than that in the fg-TiO₂ group (Fig. 7K and 7L). These data indicated that ILA supplementation mitigated gut mucosal barrier damage induced by fg-TiO₂ through the enhancement of mucin sulfation.

Fig. 7.

ILA reversed intestinal mucosal barrier damage and weakened mucin sulfation induced by fg-TiO₂. (A) Fecal ILA levels. (B) Colonic AHR content. (C) Representative MUC2 immunofluorescence staining images of distal colon tissues. Scale bar: 100 μm. (D) Colonic MUC2 content. (E) FITC-dextran concentration in serum. (F) Representative images of distal colonic sections stained with HID-AB. Scale bar: 100 μm. (G) Sulfomucin positive cell number per crypt. (H-J) Papss2, Slc35b3, and GlcNAc6st2 mRNA expression levels in colon tissues. (K, L) Representative PAPSS2 immunohistochemical staining images of distal colon tissues and statistics. Data are presented as mean ± s.e.m., n = 6 in each group in A, B, D, E, and G-J, n = 5 in each group in L, *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.

ILA markedly mitigated intestinal inflammation induced by fg-TiO₂

Finally, the effects of ILA supplementation on fg-TiO₂-induced intestinal inflammation was evaluated. Statistical analysis of colon length indicated that ILA treatment effectively mitigated the colon shortening induced by fg-TiO₂ (Fig. S18). Histological examination using H&E staining revealed that the colonic architecture in the fg-TiO₂+ILA group was more intact compared to the only fg-TiO₂ group (Fig. 8A). The colonic Il-1β, Il-6, and Tnf-α mRNA expression levels were found to be significantly lower in the fg-TiO₂+ILA group than in the fg-TiO₂ group, approaching levels observed in the control group, as demonstrated by qRT-PCR analysis (Fig. 8B-8D). Although ILA treatment lowered fecal LPS concentration in mice exposed to fg-TiO₂, this decrease was not statistically significant (Fig. 8E). Interestingly, the serum LPS concentration in the fg-TiO₂+ILA group was obviously decreased by 30.0% relative to the fg-TiO₂ group (Fig. 8F), suggesting that ILA conferred protective effects on the gut mucosal barrier, thus inhibiting the translocation of LPS from the intestinal lumen into the bloodstream. LPS from translocating from the intestinal lumen to the bloodstream. The serum IL-1β, IL-6, and TNF-α concentrations in mice exposed to fg-TiO₂ were significantly decreased by 16.6%, 23.5%, and 22.1%, respectively, following ILA supplementation (Fig. 8G-8I). These findings suggested that ILA effectively mitigated the inflammation induced by fg-TiO₂.

Fig. 8.

Effects of ILA on fg-TiO₂-induced intestinal inflammation. (A) Representative H&E staining images of distal colon tissues. Scale bar: 100 μm. (B-D) Gene expression levels of Il-1β, Il-6, and Tnf-α in colon tissues. (E) Fecal LPS concentration. (F) Serum LPS concentration. (G-I) Serum concentration of IL-1β, IL-6, and TNF-α. Data are shown as mean ± s.e.m., n = 6 in each group, *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001, n.s., non-significant

Discussion

Over recent decades, the safety of TiO₂ has been a subject of considerable debate. Increasing evidence indicates that the oral consumption of TiO₂ may pose health risks [1]. In particular, recent advancements in detection technologies and the development of standardized detection methods have revealed that fg-TiO₂ contains a higher percentage of nanoparticles than the previously noted 36%, ranging from 40% to 80% [4, 40]. In 2021, the European Food Safety Authority (EFSA) re-evaluated the safety of the fg-TiO₂ using a nano-specific approach and determined it to be unsafe, prompting the European Union to prohibit its use as a food additive [41]. Conversely, regulatory bodies in other regions, such as the China Food and Drug Administration (CFDA), the United States Food and Drug Administration (FDA), the British Food Standards Agency (FSA), Health Canada, and Food Standards Australia New Zealand, have stated that fg-TiO₂ does not pose safety concerns [1, 36]. This divergence in policy can be attributed to several factors, including advancements in testing methodologies, regional variations in exposure levels, and the impact of dietary culture [1]. Consequently, there is a pressing need for more comprehensive and systematic research into the safety of TiO₂ as an additive in food. This study looked into the adverse effects of prolonged consumption of fg-TiO₂ on the intestinal mucus barrier in mice, focusing on its impact on the interaction between intestinal flora and mucin sulfation. To the best of our knowledge, our study appears to provide the first evidence to elucidate that fg-TiO₂ induced gut microbiota dysbiosis, particularly the decline of Lactobacillus, leading to a reduction in intestinal ILA levels, which in turn impaired mucin sulfation and ultimately compromised the intestinal mucus barrier.

In this study, mice were exposed to fg-TiO₂ at 40 mg/kg body weight, a level consistent with with previous studies [8, 9]. This dosage corresponds to a human equivalent dose of 3.24 mg/kg body weight after body surface area normalization [42]. Recent statistical analyses indicate that the average oral consumption of TiO₂ in the population ranges from 0.045 to 10.5 mg/kg body weight per day [1]. Notably, the exposure levels in children are significantly higher than those in adults, attributable to children’s preference for foods containing TiO₂ (candies, cakes, and other sweets) [36]. The dosage of fg-TiO₂ employed in this study may more closely approximate the exposure levels observed in children, whose gut microbiota and intestinal barriers are more susceptible to environmental and dietary factors [4, 43]. Early life exposure to TiO₂ has been shown to alter intestinal microbial composition, increase intestinal barrier permeability, and heighten the risk of developing colitis [44]. It is important to note that adults with medium-to-high dietary exposure can also reach the daily exposure level of 3.24 mg/kg TiO₂ body weight, potentially due to their dietary habits [4]. Overall, the fg-TiO₂ dosage used in this study falls within the limits of human exposure, applicable to both adults and children.

Previous researches on mammals have similarly indicated a strong association between oral ingestion of fg-TiO₂ and the onset of intestinal inflammation [8, 45]. Certain studies have identified that fg-TiO₂ particles can stimulate intestinal cells to produce reactive oxygen species (ROS), leading to oxidative stress. This subsequently activates inflammasomes such as NOD-like receptor family pyrin-containing 3 (NLRP3), enhance the secretion of inflammatory cytokines like IL-1β, and exacerbate or initiate localized intestinal inflammation [46, 47]. The current study investigated the impact of oral fg-TiO₂ exposure on the intestinal microbiota and found that fg-TiO₂ significantly enriched Firmicutes while diminished Bacteroides, aligning with both our previous research and the findings of other study [8, 14]. An elevated Firmicutes-to-Bacteroides ratio is a characteristic feature of the intestinal microbes in obese individuals [48], and this research also noted a minor rise in body weight among mice treated with fg-TiO₂ (Fig. S2). Furthermore, fg-TiO₂ led to significantly reduced levels of Lactobacillus, which have been shown to mitigate colonic inflammation and preserve the gut barrier integrity [49, 50]. Subsequent FMT experiments successfully replicated the changes in the abundance of Lactobacillus (Fig. 4G), suggesting that modifications in gut microbiota are pivotal in the intestinal inflammation induced by fg-TiO₂.

Notably, a significant outcome of our study was that the intestinal inflammation induced by fg-TiO₂ was partially attributable to its detrimental effects on the intestinal mucin sulfation. Sulfated mucin, a prevalent form of O-linked glycan modification, is plentiful in the small intestine and colon, with particularly elevated levels observed in the distal colon. The increased concentration of sulfated mucin in the distal colon serves to protect against the degradation of the mucosal layer by bacterial enzymes, thereby effectively preventing close contact between bacteria, harmful agents, and the intestinal epithelium [21]. Numerous studies have demonstrated the role of intestinal flora in regulating mucin sulfation [51]. Notably, tryptophan metabolites produced by the intestinal microorganisms have been reported to upregulate the expression of essential enzymes responsible for mucin sulfation (Papss and Slc35b3) through the activation of AHR signaling pathway [20]. This study revealed that fg-TiO₂ reduced the content of a microbial tryptophan metabolite ILA, leading to the inhibition of the AHR pathway and a significant decrease in mucin sulfation levels. This finding offered a novel explanation for the disruption of the intestinal mucus barrier caused by fg-TiO₂.

Our findings indicated that quercetin significantly ameliorated fg-TiO₂-induced colonic inflammation by modulating the gut microbial-ILA-mucin sulfation axis. Quercetin, a widely consumed dietary flavonoid, is noted for its multiple biological activities and favorable safety profile [27]. Previous research has demonstrated that quercetin at a dose of 2 g/kg body weight does not elicit significant adverse effects in humans [52]. In this study, the employed quercetin dosage of 50 mg/kg is consistent with those commonly utilized in studies investigating quercetin’s effects on intestinal disorders [31, 32, 53]. Translated to human equivalent dose, 50 mg/kg in mice corresponds to approximately 4.1 mg/kg in humans, equivalent to a daily intake of 287 mg for a 70‑kg adult. This intake falls within the typical dietary range of 10–500 mg of quercetin in humans, as reported by statistical analyses [54]. Quercetin has shown anti-inflammatory properties via various mechanisms; however, its efficacy is constrained by its low bioavailability [55, 56]. Recent researches indicated that quercetin could significantly ameliorate inflammation-associated diseases by altering the gut microbial composition and metabolites [30]. In obese mammalian models, quercetin supplementation has been observed to lower the presence of Firmicutes and raise the presence of Bacteroides, thereby enhancing gut microbial metabolites, including SCFAs and bile acids, which ultimately reduced metabolic inflammation and improved lipid metabolism [53, 57, 58]. At the genus level, quercetin has been documented to promote the growth of specific beneficial bacteria, including Lactobacillus, Akkermansia, and Bifidobacterium [32, 57, 59]. These findings aligned with this study’s observations on quercetin’s role in the enrichment of Lactobacillus. Furthermore, several studies have reported that quercetin can ameliorate intestinal barrier dysfunction in various mouse models [60]. Quercetin supplementation has been reported to increase goblet cell numbers and MUC2 levels in colitis mouse models, thereby maintaining the integrity of the colonic mucus layer [35, 61]. Another study demonstrated that quercetin notably raised the levels of indole-3-propionic acid (IPA), a tryptophan metabolite, in the feces of obese mice and provided protective effects on the intestinal barrier against damage [31]. Notably, quercetin supplementation was proved to enhance microbial ILA production to promote mucin sulfation via activating the AHR signaling pathway, which strengthened the colonic mucosal barrier in this study. Many other flavonoids (e.g., luteolin, kaempferol, apigenin) also exert comparable anti-inflammatory and gut microbiota-modulating activities [62]. However, our study focused on the gut microbiota-ILA-mucin sulfation axis and specifically identified quercetin as a key regulator of this axis. Therefore, while anti-inflammatory and microbial-regulating properties are common among flavonoids, quercetin may act specifically in regulating ILA metabolism and the mucin sulfation pathway, warranting further comparative validation with other flavonoids. Our findings offered novel insights into dietary strategies aimed at ameliorating intestinal barrier dysfunction induced by fg-TiO₂.

The discoveries in this research indicated that ILA played a significant role in the intestinal mucus barrier in the context of fg-TiO₂ and quercetin exposure. ILA is a gut microbial metabolite of tryptophan, with its biosynthesis primarily facilitated by specific intestinal symbiotic bacteria [63]. Probiotics and symbiotic bacteria, including Lactobacillus and Bifidobacterium, have been identified as major ILA producers [64, 65]. In this study, ILA concentrations varied in parallel with Lactobacillus abundance, further supporting the ability of this genus to produce ILA. Additionally, certain members of the Bacteroidetes have also been recognized for their capability to generate ILA [66]. As a crucial signaling molecule, ILA exerts direct effects on intestinal epithelial cells, particularly goblet cells, ultimately modulating the structure and function of the mucus layer. It was demonstrated that ILA significantly upregulated the gene expression of Muc2 in the human LS174T goblet cell line in a dose-dependent manner in vitro [38]. This finding was corroborated by animal model studies [38, 39]. Mice supplemented with Lactobacillus johnsonii, a bacterium capable of producing ILA, or its culture supernatant exhibited an observed increase in both goblet cell number and colonic MUC2 level [39]. Our study elucidated that ILA supplementation activated the AHR pathway, upregulated mucin sulfation-related enzymes, promotes mucin sulfation, and thereby strengthened the intestinal mucosal barrier to alleviate fg-TiO₂-induced damage. These findings provided a foundational basis for developing therapeutic strategies targeting intestinal barrier injuries through the application of ILA.

Nevertheless, several limitations remain to be addressed in this study. Although the present study employed human-relevant doses of fg-TiO₂ and quercetin to investigate their effects on the gut microbiota-ILA-mucin sulfation axis, the use of a single dose precludes the assessment of dose-response relationships and thereby limits evaluation of the effect magnitude and potential thresholds. In addition, our co-administration design of ILA and fg-TiO₂ only demonstrates the preventive effect of ILA, not its therapeutic efficacy on established intestinal injury, which limits its translational relevance. Future studies with dose gradients and post-injury treatment designs are needed to address these issues.

Conclusion

In conclusion, the present study indicated that prolonged consumption of 40 mg/kg of fg-TiO₂ particles could precipitate notable gut microbiota dysbiosis, especially the decline of Lactobacillus, consequently diminishing intestinal ILA levels. The reduction in ILA caused decreased mucin sulfation due to the inhibition of the AHR signaling pathway, culminating in significant impairment of the intestinal mucosal barrier and subsequent inflammation. Notably, supplementing with quercetin has been shown to effectively ameliorate the fg-TiO₂-induced damage to the intestinal mucosal barrier by increasing Lactobacillus abundances, thereby activating the ILA-AHR-mucin sulfation axis (Fig. 9). This investigation not only elucidated a novel mechanism underlying fg-TiO₂ induced impairment of the intestinal mucosal barrier, but also proposed a novel dietary intervention strategy to mitigate fg-TiO₂-induced intestinal injury.

Fig. 9.

Quercetin ameliorated fg-TiO₂-induced intestinal mucosal barrier damage via modulating gut microbiota-ILA-mucin sulfation interactions. Quercetin administration improved the gut microbiota dysbiosis induced by long-term fg-TiO₂ ingestion, notably by counteracting the depletion of Lactobacillus. The augmentation in Lactobacillus populations subsequently elevated the concentration of its metabolite ILA, which in turn activated the AHR. This activation subsequently upregulated the expression of Papss2 and Slc35b3, thereby enhancing mucin sulfation and safeguarding the intestinal mucosal barrier against damage induced by fg-TiO₂

Author contributions

X. Z., X. G., and Y. C. designed the experiments and co-wrote the manuscript; L. L., A. W., and L. D. performed the experiments; T. W., W. H., Y. Y., J. L., and X. B. assisted the experiments; X. Z., L. L., L. D., and A. W. analyzed the data; X. G., Y. C., and X. Z. revised the manuscript. All authors approved the manuscript’s final version. ^#These authors contributed equally.

Funding

National Natural Science Foundation of China (82304152), Natural Science Foundation (Exploration Project) of Wuhan City (2024040801020373), the Funding for Scientific Research Projects from Wuhan Municipal Health Commission (WX23Q28), Natural Science Foundation of Hubei Province (2026AFC1312, 2023AFB392), Zhongguancun Precision Medicine Foundation, Top Talent Co-Cultivated Project of Wuhan Municipal Health Commission, and Chenxing Project of Wuhan Municipal Health Commission.

Data availability

No datasets were generated or analysed during the current study.

Declarations

Competing interests

The authors declare no competing interests.