Skip to main content
NCBI home page
Particle and Fibre Toxicology logoLink to Particle and Fibre Toxicology
. 2026 Mar 31;23:20. doi: 10.1186/s12989-026-00676-7

Maternal PM2.5 exposure induces insulin resistance in adult male mouse offspring via gut–liver axis

Renjie Hu 1,#, Wenbin Zhao 1,#, Wenjun Fan 1,2, Ran Li 1,2, Lu Zhang 1,2, Li Qin 1,2, Huanhuan Zhu 1,2, Xuan Zhu 3, Qinghua Sun 1,2, Guifen Qiang 4, Cuiqing Liu 1,2,✉
PMCID: PMC13085421  PMID: 41917993

Abstract

Background

The rising prevalence of metabolic diseases represents a global health challenge, with metabolically unhealthy normal-weight (MUHNW) individuals remaining largely overlooked. In addition to direct fine particulate matter (PM2.5) inhalation, there is growing recognition that maternal PM2.5 exposure may be a contributing environmental factor for metabolic disorders. However, the mechanisms by which maternal PM2.5 exposure induced metabolic disorders in the offspring remain unknown.

Methods

Eight-week-old pregnant C57BL/6N mice were exposed to either filtered air (FA) or ambient PM2.5 throughout gestation, from gestational day 0 to 18, using a whole-body inhalation exposure system. Eight-week-old male C57BL/6N mice were treated once daily for three consecutive days with an antibiotic cocktail containing 1 g/L ampicillin, 0.5 g/L neomycin, 0.5 g/L vancomycin, and 1 g/L metronidazole to generate pseudo-germ-free mice. Subsequently, fecal microbiota from maternal PM2.5-exposed three-week-old male mouse offspring (donor) were transplanted to pseudo-germ-free mice (recipient) via oral gavage twice weekly for five weeks. After fecal microbiota transplantation (FMT), fecal samples from donor and recipient mice were collected for full-length 16S rRNA sequencing. Liver tissue from donor mice was analyzed by 5R 16S rRNA sequencing.

Results

Maternal PM2.5 exposure induced non-obese insulin resistance in adult male mouse offspring, with the liver identified as a susceptible organ characterized by suppressed AKT phosphorylation. Subsequently, systemic and hepatic insulin resistance were recapitulated in pseudo-germ-free mice, which received gut microbiota from maternal PM2.5-exposed mouse offspring via FMT. Mechanistically, the increased abundance of Helicobacter hepaticus contributed to DNA damage-mediated colonic barrier injury. This impaired colonic barrier facilitated gut-to-liver translocation of bacteria and lipopolysaccharide (LPS), which triggered hepatic inflammation via activation of TLR4 signaling pathway, ultimately leading to insulin resistance.

Conclusions

These findings indicated a causal role for gut microbiota dysbiosis in maternal PM2.5 exposure-induced non-obese insulin resistance in the offspring, providing potential insights into the developmental origins of MUHNW from the perspective of maternal exposure to air pollution.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12989-026-00676-7.

Keywords: Fine particulate matter, Maternal exposure, Insulin resistance, Gut microbiota, Gut−liver axis

Background

In the past two decades, the global prevalence of metabolic diseases has increased significantly, becoming a major public health concern [1]. Previous studies have mainly focused on obesity-related metabolic disorders, largely neglecting the metabolically unhealthy normal weight (MUHNW) phenotype, which affected approximately 20% of individuals with normal weight [2]. Recent studies have demonstrated that early-life environmental exposures increased the risk of metabolic disease in adulthood [3–5], a concept known as the Developmental Origins of Health and Disease (DOHaD). Fine particulate matter (PM2.5), a key constituent of air pollution, has been identified in our previous studies as an important contributor to metabolic disease in both obese and non-obese mice [6–8]. Importantly, exposure to PM2.5 during gestation has been demonstrated to program adverse metabolic outcomes in mouse offspring [9, 10]. Elucidating how maternal PM2.5 exposure causes metabolic diseases in offspring is crucial for understanding the pathophysiology of the MUNW phenotype.

The gut microbiota was recognized as a crucial regulator in the pathogenesis of metabolic diseases, particularly by its role in regulating hepatic metabolism [11, 12]. An epidemiological study demonstrated a correlation between PM2.5 exposure and disrupted gut bacterial colonization in neonates [13]. Animal studies have consistently demonstrated that maternal exposure to PM2.5 led to a significantly altered gut microbiota composition in offspring [14, 15]. However, it remains unknown how gut microbiota dysbiosis contributes to maternal PM2.5 exposure-induced metabolic disturbances in offspring, as well as the underlying mechanisms involved.

The intestinal barrier serves as the primary defense against harmful gut pathogens and their associated molecular patterns (PAMPs), including lipopolysaccharide (LPS). Pathogenic bacteria were found to disrupt intestinal barrier integrity and facilitate the gut-to-liver translocation of bacteria, which promoted hepatocellular carcinoma in mice [16]. Toll-like receptor 4 (TLR4) functions as a key pattern recognition receptor in the liver. It interacted with bacteria and PAMPs to trigger an inflammatory response and contributed to the pathogenesis of liver disease [17]. Interestingly, germ-free mice treated with diethylnitrosamine (DEN) and the hepatotoxin carbon tetrachloride (CCl4) exhibited attenuated hepatic injury due to reduced TLR4 activation [18]. This evidence highlighted the potential contribution of TLR4 activation, induced by gut microbiota dysbiosis, to the development of liver disease. However, whether maternal PM2.5 exposure facilitates the gut-to-liver translocation of bacteria and PAMPs, and the mechanisms by which this process induces disturbances in offspring hepatic metabolism, remain unknown.

Thus, we established the causality of maternal PM2.5 exposure-induced gut microbiota dysbiosis in promoting non-obese insulin resistance in the offspring. Furthermore, we explored the mechanisms of gut–liver axis by which maternal PM2.5 exposure induced metabolic disorder. This study provided both theoretical and experimental foundations for understanding the pathogenesis of the fetal-origin MUHNW phenotype induced by air pollution.

Methods

Animals

C57BL/6N mice (7-week-old) were obtained from Charles River Laboratories and maintained under a 12-h light-dark cycle (06:00–18:00), with unrestricted access to food (Jiangsu Synergy Pharmaceutical Bioengineering Co., Ltd., China) and water. All experimental procedures received approval from the Institutional Animal Care and Use Committee of Zhejiang Chinese Medical University (ZCMU) and were performed in accordance with its guidelines.

Maternal PM2.5 exposure

Following a one-week acclimation, virgin female mice were paired with male mice at 20:00. Vaginal plugs were examined at 08:00 on the subsequent morning. Female mice with a detected vaginal plug, marking gestational day 0 (GD 0), were randomly allocated to either a filtered air (FA) group (n = 13) or a concentrated PM2.5 exposure group (PM2.5; n = 15). The female mice underwent inhalation exposure for 12 h daily (from 08:00 to 20:00) over 19 successive days (GD 0 to 18) using a whole-body inhalation exposure system on the ZCMU campus, as previously described [19]. After GD 18 exposure, pregnant mice were transferred from both FA (n = 5) and PM2.5 (n = 5) chambers to individual cages in a clean air environment until delivery (~ GD 19) to prevent offspring from exposure to PM2.5. Notably, only female mice received to PM2.5 exposure, while male mice used for mating and all offspring were maintained in a clean air environment throughout the study.

At postnatal day 0 (PND 0), pups were randomly culled to a litter size of 6–8 and housed with their dams under clean air conditions (5 litters per group). At weaning (PND 21), 1–2 male and 1–2 female offspring from each litter were randomly selected and group-housed by sex, divided into 4 groups: Male/FA (n = 8), Female/FA (n = 8), Male/PM2.5 (n = 8), and Female/PM2.5 (n = 8). At PND 56, offspring were subjected to glucose tolerance test (GTT), followed by insulin tolerance test (ITT) at PND58. Body composition analysis was performed at PND 59. At PND 60, offspring were anesthetized with isoflurane. Blood was collected via the retro-orbital plexus, after which the mice were euthanized by cervical dislocation for necropsy. Liver, colon, epididymal white adipose tissue (eWAT), inguinal white adipose tissue (iWAT), brown adipose tissue (BAT), and skeletal muscle were collected. Portions of each tissue were fixed in 4% paraformaldehyde (PFA), while the remaining tissues were immediately snap-frozen in liquid nitrogen and stored at −80℃ for subsequent analysis.

PM2.5 concentration

As previously described [19], PM2.5 in the chambers was sampled using Teflon membranes (Pall Corporation, USA) during the exposure period. Membrane weights were measured before and after sampling utilizing a microbalance (Mettler-Toledo, Switzerland) under constant temperature and humidity conditions. The PM2.5 concentration was then calculated according to the following formula: PM2.5 concentration (µg/m3) = weight gain of the membrane (µg)/[sampling time (min) × sampling flow rate (L/min) × 10− 3]. The sampling flow rate was determined using an air flow calibrator (Gilibrator 2, Sensidyne, USA). The average daily PM2.5 concentration inhaled by the mice over the 24-hour cycle was normalized according to the 12-hour exposure protocol as follows: average daily PM2.5 concentration (µg/m3) = [PM2.5 concentration in the PM2.5 chamber (µg/m3) × 12 (hour) + PM2.5 concentration in the FA chamber (µg/m3) × 12 (hour)]/24 (hour).

Fecal microbiota transplantation (FMT)

To generate pseudo-germ-free mice, C57BL/6N male mice (8-week-old) were orally gavaged with a 200 µL antibiotic cocktail (ABX) once daily for three consecutive days as previously described [20]. The ABX solution contained ampicillin (1 g/L; HY-B0522A, MCE, USA), neomycin (0.5 g/L; HY-B0470, MCE, USA), vancomycin (0.5 g/L; HY-17362, MCE, USA), and metronidazole (1 g/L; HY-B0318, MCE, USA). Subsequently, beginning on the first day after the antibiotic regimen, recipient pseudo-germ-free mice were randomly allocated to either a FA-FMT group (n = 8) or a PM2.5-FMT (n = 8), and administered a 200 µL microbial suspension from male donor offspring via oral gavage twice weekly for five weeks. This period corresponded to PND 21 to 56 in the donors.

After gently lifting the mouse, 100 mg of fresh fecal pellets from male donor offspring were collected using autoclaved forceps and immediately placed in to a sterile centrifuge tube containing 1 mL of sterile PBS, followed by homogenization to prepare the microbial suspension. After being centrifuged at 2000g for 2 min, the supernatant was harvested for gavage.

Body composition analysis

Measurements of lean and fat mass in mice were performed with a benchtop time-domain NMR instrument (minispec LF50, Bruker, Germany).

GTT and ITT

Mice were fasted for 12 h before the GTT and 4.5 h before the ITT. Following fasting, the mice received an i.p. injection of glucose (2 mg/g body weight) for GTT or insulin (0.5 U/kg body weight) for ITT. Blood glucose levels were determined pre-injection and at 15, 30, 60, 90, and 120 min after injection using a Precision Neo Blood Glucose Meter (Abbott Diabetes Care Inc., USA).

Homeostasis model assessment of insulin resistance (HOMA-IR)

An ultrasensitive mouse insulin ELISA kit (Crystal Chem, USA) was used to detect fasting serum insulin levels. The HOMA-IR index was calculated as previously described [6, 21]: HOMA-IR = [fasting insulin (ng/mL) × 24 × fasting glucose (mg/dL)]/405. Of this, 1 mg of insulin is equal to 24 IU.

Assessment of serum intestinal permeability parameters

Serum levels of D-Lactate were quantified with a colorimetric assay kit (Elabscience, China), and LPS and diamine oxidase (DAO) levels were detected with mouse ELISA kits (ELK Biotechnology, China).

Assessment of serum testosterone levels

Serum levels of testosterone were quantified with a testosterone assay kit (R&D Systems, USA).

Histology and immunohistochemical analysis

Hematoxylin and eosin (H&E) staining: Liver and colon samples were fixed in 4% PFA, embedded in paraffin, and sectioned at 5 μm for liver and 3 μm for colon. Sections were stained following a standard protocol as previously described [21]. Images were captured with a microscope (Leica Microsystems, Germany). Morphometric analysis of crypt depth, goblet cells count per crypt depth, and inflammatory cell infiltration was performed on six fields per section using Image J. Mucosal structure damage was specifically scored according to an established histopathological scoring system [22, 23].

For immunohistochemistry (IHC), colon sections were stained with a rabbit anti-γH2AX antibody (1:100; 9718, CST, USA), as previously described [19]. Images were captured using the same Leica DM4B microscope. The γH2AX-positive cells were quantified by analyzing five fields per section with Image J.

Analysis of fecal bacterial DNA

After ABX treatment, fecal samples were collected for bacterial DNA extraction with the SteadyPure Stool DNA Kit (Accurate Biology, China). Extracted DNA was quantified on a NanoDrop 2000 spectrophotometer (Thermo Scientific, USA). Quantitative real-time polymerase chain reaction (qRT-PCR) was conducted on a QuantStudio 7 Flex system (Applied Biosystems, USA) with universal primers specific to the bacterial 16S ribosomal RNA (rRNA) gene. The relative abundance of bacterial DNA was determined by normalization to the host 18S rRNA gene in the mouse samples. The sequences for all primers are listed in Supplemental Table 1.

16S rRNA gene V3V4 region sequencing of fecal samples

16S rRNA gene V3V4 region sequencing was conducted by LC-Bio Technology Co., Ltd. (Hangzhou, China). Briefly, the Fecal Genome DNA Extraction Kit (BioTeke, China) was utilized to isolate microbial DNA from fecal samples of ABX-treated mice. DNA quantification was performed using Qubit (Invitrogen, USA). The V3-V4 regions were amplified by PCR with the universal primer 341 F/805R; sequences are shown in Supplemental Table 2. After primer removal (cutadapt v1.9) and paired-end read merging (FLASH v1.2.8), low-quality reads were filtered out using fqtrim (v0.94). Then, Vsearch (v2.3.4) was employed to filter out chimeric sequences. Denoising and inference of amplicon sequence variants (ASVs) were conducted with DADA2, followed by taxonomic classification against the SILVA and NT-16 S databases via the QIIME2 feature-classifier. α and β diversity were subsequently analyzed within the QIIME2 framework.

Full-length 16S rRNA gene sequencing of fecal samples

Full-length 16S rRNA gene sequencing was conducted by LC-Bio Technology Co., Ltd. (Hangzhou, China). Briefly, the CTAB method was applied to extract microbial DNA from the fecal samples of offspring and recipient mice. The full-length 16S rRNA gene was amplified with specific barcode primers 27 F/1492R; sequences are shown in Supplemental Table 2. SMRTbell libraries were constructed and sequenced on the PacBio RS II platform (LC-Bio Technology Co., Ltd., Hangzhou, China). SMRT Link (v6.0) was used to generate circular consensus sequencing (CCS) reads from raw subreads (minPasses = 5, minPredictedAccuracy = 0.9), followed by demultiplexing, primer removal, and quality filtering. Following dereplication and chimera removal using DADA2, ASVs were obtained and taxonomically classified against the SILVA database. Alpha and beta diversity metrics were calculated using QIIME2. The OmicStudio platform (https://www.omicstudio.cn/tool) was used to conduct bioinformatic analyses. For this analysis, n = 6 per group, with at least 1 sample selected to ensure representation from each litter within each group.

5R 16S rRNA gene sequencing of liver samples

5R 16S rRNA sequencing was conducted by LC-Bio Technology Co., Ltd. (Hangzhou, China) as previously described [24]. Briefly, the CTAB method was used to extract DNA from liver tissues of offspring. Negative controls, including those for sampling, DNA extraction, and no-template PCR amplification, were processed in parallel to account for contamination at all stages of handling and from the laboratory environment. The 16S rRNA gene was amplified across five regions; sequences are shown in Supplemental Table 2. Sequencing of the libraries was carried out on an Illumina NovaSeq 6000 platform. Following sequencing, reads were demultiplexed, filtered, and aligned to their respective target regions. Next, sequencing reads from the five regions were integrated using the Short Multiple Regions Framework (SMURF) approach [25]. Taxonomic profiling was then performed using the GreenGenes database, followed by contamination filtering as previously described [24]. Alpha and beta diversity metrics were calculated using QIIME2. For this analysis, n = 6 per group, with at least 1 sample selected to ensure representation from each litter within each group.

qRT-PCR

RNAiso Plus (TaKaRa, Japan) was employed to isolate total RNA from liver and colon samples. A NanoDrop 2000 spectrophotometer (Thermo Scientific, USA) was applied for quantification. PrimeScript RT Master Mix (TaKaRa, Japan) was used to reverse transcribe RNA to cDNA. qRT-PCR was performed on a QuantStudio 7 Flex system (Applied Biosystems, USA). Relative gene expression levels were determined using the 2−ΔΔCt method after normalization to β-actin. All primer sequences used are provided in Supplemental Table 1.

Western blotting

RIPA lysis buffer (Boster, China) was employed to isolate total protein. BCA assay kit (Beyotime, China) was employed to quantified the protein concentration. Equal amounts of protein were loaded onto SDS-PAGE for separation and then transferred to PVDF membranes. The membranes were blocked with 5% non-fat milk and incubated overnight at 4°C with specific primary antibodies (for a complete list of antibodies, catalog numbers, and dilutions, see Supplemental Table 3). After washing, the membranes were incubated with HRP-conjugated secondary antibodies at room temperature. Protein bands were detected with a ChemiDoc imaging system (12003153, Bio-Rad, USA) and quantified using Image J. For this analysis, n = 4 per group, and all samples within each group were obtained from different litters.

Statistical analysis

All data are presented as the mean ± standard error of the mean (SEM). Statistical analyses were carried out in GraphPad Prism software (version 8.0). For comparisons between two groups, an unpaired, two-tailed Student’s t-test was used. For comparisons within the same subjects across multiple time points, a repeated-measures analysis of variance (ANOVA) followed by Bonferroni’s post hoc test was applied. For the nonparametric tests, a two-tailed Mann-Whitney U test was employed for two-group comparisons. A P-value of less than 0.05 was considered statistically significant. The specific sample sizes (n values) for each experiment are listed in the figure legends.

Results

Maternal PM2.5 exposure induced insulin resistance in adult male mouse offspring

To investigate the impacts of maternal PM2.5 exposure on developmental and metabolic outcomes in offspring, pregnant mice were exposed to PM2.5 throughout gestation (GD 0 to 18). The animal experimental design is schematically presented in Fig. 1A. During the exposure period, the average PM2.5 concentration in FA and PM2.5 chambers were monitored daily and compared with concurrent outdoor ambient PM2.5 levels at the exposure site (Fig. 1B). The average daily inhaled PM2.5 concentration for pregnant mice in the PM2.5 group was 62.18 ± 11.80 µg/m3, which was remarkably close to the actual human inhalation dose (72.33 ± 8.43 µg/m3), whereas that in the FA group was 3.87 ± 1.50 µg/m3 (Fig. 1C).

Fig. 1.

Fig. 1

Open in a new tab

Effects of maternal PM2.5 exposure on metabolic health in adult male mouse offspring. A Schematic experimental design. B The average daily PM2.5 concentration curves in the FA chamber (white solid dot), PM2.5 chamber (red solid dot), and monitoring station (gray solid dot) in the area where the exposure system was located. C The average daily PM2.5 concentration in the FA chamber, PM2.5 chamber, and monitoring station during the exposure period. D Body weight of male offspring. n = 8 per group. E Fat mass of adult male offspring. n = 8 per group. F, G ITT and the area under curve of adult male offspring. n = 8 per group. H, I GTT and the area under curve of adult male offspring. n = 8 per group. J Liver mass of adult male offspring. n = 8 per group. (K) Western blotting and quantification analysis for AKT phosphorylation levels in the liver of adult male offspring. n = 4 per group. All data were expressed as means ± SEM. Data were compared using Student’s t-test (E, G, and I–K) and two-way repeated-measures ANOVA followed by Bonferroni’s post hoc test (D, F, H). *P<0.05, **P<0.01, and ***P<0.001

The growth trajectory exhibited that maternal exposure to PM2.5 resulted in a trend toward reduced body weight in male offspring throughout the entire observation period (Fig. 1D), whereas no significant effect was observed in female offspring (Supplemental Fig. 1A). Furthermore, maternal exposure to PM2.5 did not result in significant changes in body composition (e.g., lean and fat mass) in either male or female offspring at PND56 (Fig. 1E and Supplemental Fig. 1B–D). Notably, adult male offspring from PM2.5-exposed dams exhibited significant impairment in insulin sensitivity (Fig. 1F, G), despite unaltered glucose tolerance (Fig. 1H, I), fasting insulin levels, or HOMA-IR (data not shown). However, no significant differences in glucose tolerance or insulin sensitivity were detected in adult female offspring from PM2.5-exposed dams (Supplemental Fig. 1E–H). Therefore, we primarily focused on the effects of maternal exposure to PM2.5 on adult male offspring in the subsequent investigations.

To identify the primary metabolic organ affected by maternal exposure to PM2.5 in adult male offspring, primary insulin-response tissues were collected, including the liver, skeletal muscle, and various adipose tissue such as eWAT, iWAT, and BAT. Maternal PM2.5 exposure significantly reduced liver mass and hepatic AKT phosphorylation (Fig. 1J, K), indicating impaired hepatic insulin signaling. However, no significant structural abnormalities or lipid accumulation were observed in the liver of maternal PM2.5-exposed adult male offspring (data not shown). Moreover, no significant changes in organ mass or AKT phosphorylation were detected in either adipose tissues or skeletal muscle (Supplemental Fig. 2). These results suggested that maternal exposure to PM2.5 led to non-obese insulin resistance in adult male offspring, with the liver being a primary susceptible organ.

Maternal PM2.5 exposure induced gut microbiota dysbiosis and colonic barrier injury in adult male mouse offspring

The gut microbiota exerts a profound influence on the pathogenesis of insulin resistance, particularly in hepatic metabolism [26–28]. To examine the impacts of maternal exposure to PM2.5 on the gut microbiota of adult male offspring, we collected fecal samples and performed full-length 16S rRNA sequencing. Although maternal exposure to PM2.5 did not markedly change the α diversity of the gut microbiota in adult male offspring (Fig. 2A), principal coordinate analysis (PCoA) based on Bray-Curtis distances exhibited a distinct separation in microbial community structure between the two groups (Anosim; R = 0.6352, P = 0.002) (Fig. 2B). At the phylum level, adult male offspring from PM2.5-exposed dams showed significantly increased relative abundances of Firmicutes, Desulfobacterota, Campilobacterota, and Deferribacterota, as well as a significant decrease in Bacteroidota (Fig. 2C). At the genus level, these offspring exhibited higher abundances of Lachnospiraceae NK4A136 group, Desulfovibrio, Lachnoclostridium, and Helicobacter, and lower abundances of Muribaculaceae and Prevotellaceae UCG-001 (Fig. 2D). Subsequent linear discriminant analysis effect size (LEfSe) analysis showed that Clostridium sp., Dorea sp., Clostridium fusiformis, Mucispirillum scheadleri, and Helicobacter hepaticus were significantly enriched in maternal PM2.5-exposed adult male offspring (Fig. 2E). Notably, the abundance of H. hepaticus exhibited a notable positive correlation with the area under curve of ITT (Fig. 2F), whereas no significant correlations were observed for the other four bacterial species (Supplemental Fig. 3). Next, PICRUSt2 prediction analysis exhibited a considerable upregulation in the predicted abundance of genes encoding the cytolethal distending toxin (CDT) subunit (Fig. 2G), which is a critical genotoxin produced by H. hepaticus [29]. Spearman correlation analysis consistently demonstrated a positive correlation between the predicted abundance of the CDT subunit and the relative abundance of H. hepaticus (Fig. 2H). These results indicated that maternal exposure to PM2.5 let to gut microbiota dysbiosis in adult male offspring, accompanied by a marked increase in the abundance of H. hepaticus.

Fig. 2.

Fig. 2

Open in a new tab

Effects of maternal PM2.5 exposure on gut microbiota and colonic barrier in adult male offspring. A α-diversity of gut microbiota. n = 6 per group. B PCoA plot based on Bray-Curtis distances of gut microbiota. n = 6 per group. C Relative abundance of gut bacteria at the phylum level. D Relative abundance of gut bacteria at the genus level. E LEfSe analysis (LDA score > 3) of gut microbiota at the species level. F Spearman correlation analysis between the abundance of H. hepaticus and the area under curve of ITT. n = 6 per group. G PICRUSt2-predicted abundance of genes encoding CDT subunits. n = 6 per group. H Spearman correlation analysis between the relative abundance of H. hepaticus and the predicted expression of the CDT subunit. n = 6 per group. I Serum levels of D-Lactate, LPS, and DAO in adult male offspring. n = 8 per group. J–N H&E staining and quantitative analysis of colonic pathological morphology. Scale bars, 100 μm. n = 8 per group. O qRT-PCR analysis of intestinal barrier-related gene expression in the colon of adult male offspring. n = 7 per group. P, Q Western blotting and quantification analysis for intestinal barrier-related protein expression in the colon of adult male offspring. n = 4 per group. R, S IHC staining and quantification of γH2AX in the colon of adult male offspring. Scale bars, 50 μm. n = 6 per group. T qRT-PCR analysis of DNA damage and repair-related gene expression in the colon of adult male offspring. n = 7 per group. All data were expressed as means ± SEM. Data were compared using Student’s t-test (I, K–O, Q, S, and T) and Mann-Whitney U test (A and G). *P<0.05, **P<0.01, and ***P<0.001

The intestinal barrier serves as the primary defense mechanism against pathogenic bacteria. Adult male offspring from PM2.5-exposed dams exhibited significantly increased serum levels of D-lactate, LPS, and DAO (Fig. 2I). Since H. hepaticus primarily colonizes the colon [30], colonic barrier integrity in maternal PM2.5-exposed adult male offspring was subsequently assessed. H&E staining revealed that maternal PM2.5 exposure caused colonic barrier injury in adult male offspring, as evidenced by reduced crypt depth, decreased goblet cell numbers, increased inflammatory cell infiltration, and disruption of mucosal architecture (Fig. 2J–N). qRT-PCR analysis revealed that maternal PM2.5 exposure markedly downregulated mRNA expression levels of Cldn4, Cldn5, and Tjp1, but upregulated that of Cldn2 in the colon of adult male offspring (Fig. 2O). Western blot analysis confirmed that maternal PM2.5 exposure significantly reduced the protein expression of Claudin-4 and ZO1 (encoded by Tjp1) and increased Claudin-2 in the colon of adult male offspring (Fig. 2P, Q). These results indicated that maternal exposure to PM2.5 caused colonic barrier injury in adult male offspring.

CDT acts as a genotoxin by inducing DNA damage [31]. To elucidate the mechanism by which maternal PM2.5 exposure-induced gut microbiota dysbiosis led to colonic barrier injury in adult male offspring, we examined the markers related to colonic DNA damage. IHC staining showed that maternal PM2.5 exposure markedly upregulated expression of γH2AX, a marker of DNA damage, in the colon of adult male offspring (Fig. 2R, S). The mRNA levels of DNA damage and repair-related genes (e.g., Trp53, Apex1, Ogg1, and Xrcc1) were increased in the colon of adult male offspring from PM-exposed dams (Fig. 2T). These results indicated that maternal exposure to PM2.5 resulted in colonic barrier injury in adult male offspring by inducing DNA damage.

Fecal microbiota from maternal PM2.5-exposed adult male mouse offspring induced insulin resistance in recipient mice

To delineate the role of gut microbiota dysbiosis in maternal PM2.5 exposure-induced insulin resistance in adult male offspring, we performed FMT of microbiota from both maternal PM2.5-exposed (PM2.5-FMT) and control (FA-FMT) adult male offspring into pseudo-germ-free mice (Fig. 3A). Prior to FMT, recipient mice received ABX treatment to deplete the endogenous gut microbiota. Following the three-day regimen, the ABX treatment significantly altered the microbial community structure (Anosim; R = 0.9640, P = 0.002) (Fig. 3B) and markedly reduced both fecal bacterial load and gut microbiota alpha diversity (Fig. 3C and Supplemental Fig. 4), collectively confirming the successful establishment of the pseudo-germ-free mouse model.

Fig. 3.

Fig. 3

Open in a new tab

FMT induced insulin resistance in recipient mice. A FMT experimental design. B PCoA plot based on Bray-Curtis distances of gut microbiota. n = 7 per group. C α-diversity of gut microbiota (based on Chao1). n = 7 per group. D Body weight of recipient mice. n = 8 per group. E Fat mass of recipient mice. n = 8 per group. F, G ITT and the area under curve of recipient mice. n = 8 per group. H, I GTT and the area under curve of recipient mice. n = 8 per group. J Liver mass of recipient mice. n = 8 per group. K Western blotting and quantification analysis for AKT phosphorylation levels in the liver of recipient mice. n = 4 per group. All data were expressed as means ± SEM. Data were compared using Student’s t-test (E, G, and I–K), Mann-Whitney U test (C), and two-way repeated-measures ANOVA followed by Bonferroni’s post hoc test (D, F, H). *P<0.05 and ***P<0.001

Next, we found that PM2.5-FMT did not change body weight, fat mass, or lean mass in recipient mice (Fig. 3D and E and Supplemental Fig. 5A). Notably, PM2.5-FMT significantly impaired insulin sensitivity and glucose tolerance in recipient mice (Fig. 3F–I), despite unaltered fasting insulin levels and HOMA-IR (Supplemental Fig. 5B and C). Furthermore, PM2.5-FMT significantly reduced liver mass and suppressed hepatic AKT phosphorylation in recipient mice (Fig. 3J, K). These results demonstrated that fecal microbiota from maternal PM2.5-exposed adult male offspring induced insulin resistance in recipient mice.

Fecal microbiota from maternal PM2.5-exposed adult male mouse offspring led to colonic barrier injury in recipient mice

To detect the specific gut microbiota responsible for PM2.5-FMT-induced insulin resistance, full-length 16 S rRNA sequencing was performed on fecal samples from recipient mice. Our results showed that PM2.5-FMT did not markedly change the α and β diversity of gut microbiota in recipient mice (Fig. 4A, B). At the phylum level, PM2.5-FMT markedly increased the relative abundances of Campilobacterota in recipient mice (Fig. 4C). At the genus level, the relative abundance of Helicobacter was markedly elevated in PM2.5-FMT recipient mice (Fig. 4D). Furthermore, LEfSe analysis revealed that H. hepaticus was significantly enriched in PM2.5-FMT recipient mice, whereas Lachnospiraceae and Desulfovibrio fairfieldensis were enriched in FA-FMT recipient mice (Fig. 4E). Notably, the abundance of H. hepaticus exhibited a significant positive correlation with the area under curve of ITT (Fig. 4F). PICRUSt2 analysis revealed that PM2.5-FMT considerably increased the predicted abundance of genes encoding the CDT subunits in recipient mice (Fig. 4G). These results suggested that fecal microbiota from maternal PM2.5-exposed adult male offspring recapitulated gut microbiota dysbiosis in recipient mice, accompanied by a marked increase in the abundance of H. hepaticus.

Fig. 4.

Fig. 4

Open in a new tab

FMT recapitulated gut microbiota dysbiosis and led to colonic barrier injury in recipient mice. A α-diversity of gut microbiota. n = 6 per group. B PCoA plot based on Bray-Curtis distances of gut microbiota. n = 6 per group. C Relative abundance of gut bacteria at the phylum level. D Relative abundance of gut bacteria at the genus level. E LEfSe analysis (LDA score > 3) of gut microbiota at the species level. F Spearman correlation analysis between the abundance of H. hepaticus and the area under curve of ITT. n = 6 per group. G PICRUSt2-predicted abundance of genes encoding CDT subunits. n = 6 per group. H Serum levels of D-Lactate, LPS, and DAO. n = 8 per group. I–M H&E staining and quantitative analysis of colonic pathological morphology. Scale bars, 100 μm. n = 8 per group. N qRT-PCR analysis of intestinal barrier-related gene expression in the colon of recipient mice. n = 7 per group. O, P Western blotting and quantification analysis for intestinal barrier-related protein expression in the colon of recipient mice. n = 4 per group. Q, R IHC staining and quantification of γH2AX in the colon of recipient mice. Scale bars, 50 μm. n = 6 per group. S qRT-PCR analysis of DNA damage and repair-related gene expression in the colon of recipient mice. n = 7 per group. All data were expressed as means ± SEM. Data were compared using Student’s t-test (H, J–N, P, R, S) and Mann-Whitney U test (A, G). *P<0.05, **P<0.01, and ***P<0.001

Next, colonic barrier integrity was compared between PM2.5-FMT and FA-FMT recipient mice. PM2.5-FMT recipient mice exhibited significantly elevated serum levels of D-lactate, LPS, and DAO (Fig. 4H). H&E staining revealed that PM2.5-FMT caused colonic barrier injury in recipient mice, as evidenced by reduced crypt depth, decreased goblet cell numbers, disrupted mucosal architecture, and a modest increase in inflammatory cell infiltration (Fig. 4I–M). qRT-PCR analysis showed that PM2.5-FMT markedly downregulated the mRNA expression levels of Cldn1, Cldn4, Cldn5, and Tjp1, but slightly increased that of Cldn2 in the colon of recipient mice (Fig. 4N). Consistently, western blot analysis showed that PM2.5-FMT recipient mice exhibited significantly decreased protein expression levels of Claudin-1, Claudin-4, and ZO1, and increased Claudin-2 in the colon of recipient mice (Fig. 4O, P). These results demonstrated that fecal microbiota from maternal PM2.5-exposed adult male offspring led to colonic barrier injury in recipient mice.

To determine whether gut microbiota dysbiosis induced DNA damage, we evaluated relevant biomarkers in the colon of TMF recipient mice. PM2.5-FMT significantly upregulated the expression of γH2AX in the colon of recipient mice (Fig. 4Q, R), accompanied by increased mRNA expression of genes associated with DNA damage and repair (Fig. 4S). These results suggested that fecal microbiota from maternal PM2.5-exposed adult male offspring induced DNA damage in recipient mice, which in turn led to colonic barrier injury.

Maternal PM2.5 exposure promoted gut-to-liver translocation of bacteria and TLR4 activation, resulting in inflammation in adult male mouse offspring

To determine whether bacteria translocated from the gut to the liver in adult male offspring from PM2.5-exposed dams, we analyzed the hepatic microbiome in adult male offspring (donor mice) using 5R 16 S rRNA sequencing. Maternal PM2.5 exposure markedly increased the α diversity of the liver bacteria in donor mice (Fig. 5A). PCoA based on Bray-Curtis distances exhibited a significantly alteration in liver bacteria community structure in donor mice from PM2.5-exposed dams (Anosim; R = 0.2458, P = 0.005) (Fig. 5B). At the phylum level, maternal PM2.5 exposure significantly elevated the relative abundance of Firmicutes and decreased that of Proteobacteria in the liver of donor mice (Fig. 5C), consistent with the alterations observed in fecal samples. At the genus level, maternal PM2.5 exposure significantly increased the relative abundance of Lactobacillus, Dorea, and Blautia in the liver of donor mice (Fig. 5D). LEfSe analysis showed that maternal PM2.5 exposure resulted in a significant enrichment of Lactobacillus reuteri, Lactobacillus agilis, and Blautia producta in the liver of donor mice (Fig. 5E). These results indicated that maternal exposure to PM2.5 caused a considerable increase in bacterial load and altered the bacterial composition in the liver of donor mice.

Fig. 5.

Fig. 5

Open in a new tab

Effects of maternal PM2.5 exposure on bacterial translocation and hepatic inflammation in adult male offspring. A α-diversity of bacteria in the liver of donor mice. n = 8 per group. B PCoA plot based on Bray-Curtis distances of bacteria in the liver of donor mice. n = 8 per group. C Relative abundance of liver bacteria at the phylum level in donor mice. D Relative abundance of liver bacteria at the genus level in donor mice. E LEfSe analysis (LDA score > 3) of liver bacteria at the species level in donor mice. F, G Western blotting and quantification analysis for phosphorylation levels of P65, ERK, P38, and JNK in the liver of donor mice. n = 4 per group. H, I Western blotting and quantification analysis for protein expression of IL-1β, TNF-α, and IL-6 in the liver of donor mice. n = 4 per group. J, K Western blotting and quantification analysis for phosphorylation levels of P65, ERK, P38, and JNK in the liver of recipient mice. n = 4 per group. L, M Western blotting and quantification analysis for protein expression of IL-1β, TNF-α, and IL-6 in the liver of recipient mice. n = 4 per group. All data were expressed as means ± SEM. Data were compared using Student’s t-test (G, K, I, M) and Mann-Whitney U test (A). *P<0.05, **P<0.01, and ***P<0.001

TLR4 is a key pattern recognition receptor that detects bacterial pathogens and PAMPs, particularly LPS. To investigate the mechanism by which gut microbiota dysbiosis contributed to hepatic insulin resistance induced by maternal exposure to PM2.5 in donor mice, we examined the impacts of maternal exposure to PM2.5 on hepatic TLR4 signaling pathway in donor mice. Although maternal PM2.5 exposure did not affect the Tlr4 mRNA expression in the liver of donor mice (Supplemental Fig. 6 A), it markedly activated MAPK and NF-κB signaling pathways downstream of TLR4, as evidenced by enhanced phosphorylation of P38, P65, and ERK (Fig. 5F, G). Consistently, protein expression levels of the inflammatory cytokines, including IL-1β, TNF-α, and IL-6, were significantly increased in the liver of donor mice from PM2.5-exposed dams (Fig. 5H, I). These results indicated that maternal exposure to PM2.5 triggered hepatic inflammation by activating the TLR4 signaling pathway in donor mice.

To elucidate the role of gut microbiota dysbiosis in maternal PM2.5 exposure-induced hepatic inflammation in donor mice, we detected the aforementioned TLR4 signaling pathway and inflammatory cytokines in the FMT recipient mice. PM2.5-FMT significantly upregulated the mRNA expression of Tlr4 in the liver of recipient mice (Supplemental Fig. 6B). Consistent with the results in maternal PM2.5-exposed donor mice, we also detected significant activation of both MAPK and NF-κB signaling pathways in PM2.5-FMT recipient mice, accompanied by markedly increased protein expression of downstream inflammatory cytokines (Fig. 5J–M). These results demonstrated that fecal microbiota from maternal PM2.5-exposed donor mice induced hepatic inflammation by activating the TLR4 signaling pathway in recipient mice.

Discussion

To our knowledge, this is the first study to establish the fetal origins of gut microbiota dysbiosis in PM2.5 exposure-mediated insulin resistance within the largely ignored context of MUHNW. The main findings of this study are as follows: (i) maternal PM2.5 exposure induced non-obese insulin resistance in the adult male offspring, identifying the liver as a primary susceptible organ; (ii) maternal PM2.5 exposure led to gut microbiota dysbiosis and DNA damage-mediated colonic barrier injury in the offspring; (iii) fecal microbiota from maternal PM2.5-exposed offspring recapitulated insulin resistance, gut microbiota dysbiosis, and DNA damage-mediated colonic barrier injury in recipient mice; (iv) maternal exposure to PM2.5 promoted bacterial translocation from the gut to the liver, activating TLR4-mediated hepatic inflammation in the offspring.

The MUHNW phenotype was observed in approximately 20% of individuals with normal body weight [2]. Previous studies have demonstrated that early-life environmental exposures increased the risk of metabolic disease in adulthood [3–5]. Given the persistence of air pollution, this study explored the fetal origins of the MUHNW phenotype from the perspective of maternal PM2.5 exposure. In this study, maternal PM2.5 exposure induced insulin resistance in adult male offspring, but not in adult female offspring. Importantly, this fetal-origin insulin resistance was not accompanied by significant alterations in body weight, indicating an MUHNW-like phenotype in adult male offspring from PM2.5-exposed dams. Moreover, our further investigation into various insulin-targeting organs in the offspring identified the liver as the primary organ affected by maternal PM2.5 exposure, characterized by significantly decreased hepatic AKT phosphorylation. Therefore, this non-obese, fetal-origin insulin resistance resulting from maternal exposure to PM2.5 established a critical etiological link between maternal environmental insult and the MUHNW phenotype, highlighting maternal exposure to PM2.5 as a risk factor for this clinically significant yet underrecognized patient population. In addition, this MUHNW-like phenotype in adult male offspring from PM2.5-exposed dams extended into middle age and progressively affected extrahepatic organs such as iWAT [32], suggesting the long-term and dynamic metabolic effects.

This pronounced sexual dimorphism in non-obese insulin resistance induced by maternal PM2.5 exposure is consistent with the higher prevalence of diabetes currently observed in males compared to females [33]. Sex hormones play a pivotal role in mediating sex differences observed in health and disease [34–36]. Our results showed that maternal PM2.5 exposure significantly elevated serum androgen levels in adult male offspring while concurrently reduction in females (Supplemental Fig. 7). Elevated androgen levels are implicated in the pathogenesis of metabolic disturbances like insulin resistance in both sexes [37, 38], whereas estrogen confers protection against metabolic dysfunction in premenopausal women [39]. This sex-specific disparity in hormone levels, observed under both physiological and environmental challenges, may underlie the differential sensitivity of offspring to maternal PM2.5 exposure-induced non-obese insulin resistance.

It is noteworthy that we identified gut microbiota dysbiosis as a causal driver of maternal PM2.5 exposure-induced insulin resistance in offspring by application of FMT. Firstly, we confirmed that maternal exposure PM2.5 caused gut microbiota dysbiosis in offspring. Previous epidemiological and animal studies consistently demonstrated that maternal exposure to PM2.5 disrupted gut microbiota colonization in neonates and led to dysbiosis in adulthood [13, 15, 40]. Next, FMT of microbiota from maternal PM2.5-exposed offspring into pseudo-germ-free mice duplicated insulin resistance, impaired glucose tolerance and hepatic insulin signals without altering body weight in recipient mice, recapitulating the MUHNW-like phenotype.

Another novelty was that we identified H. hepaticus as the key pathogenic bacterium in maternal PM2.5 exposure-induced insulin resistance in offspring. H. hepaticus, which was positively correlated with insulin resistance, was the only bacterium species whose relative abundance was significantly increased in both the donor (maternal PM2.5-exposed offspring) and the FMT recipient mice. H. hepaticus mainly colonized the colon and was a genotoxin-producing pathogenic bacterium implicated in a spectrum of intestinal and hepatic diseases, including inflammatory bowel disease, colorectal carcinoma, chronic hepatitis, and hepatocellular carcinoma [29, 41, 42]. Subsequently, PICRUSt2 analysis indicated that the predicted abundance of the CDT subunits was remarkably upregulated in both the donor and the FMT recipient mice, showing a strong positive correlation with the abundance of H. hepaticus. CDT is a known genotoxin produced by various bacterial species, including H. hepaticus [29], and functions to induce DNA damage [31]. Consistently, we detected significantly increased colonic DNA damage and impaired colonic barrier in both the donor and the recipient mice, providing a mechanistic basis for gut microbiota dysbiosis-mediated insulin resistance in maternal PM2.5-exposed offspring.

Ultimately, our research revealed gut-to-liver translocation of microbiota through the damaged colonic barrier in maternal PM2.5-exposed offspring, triggering a pro-inflammatory cascade involving the activation of TLR4 signaling (e.g., MAPK and NF-κB signaling pathways) and hepatic inflammation. Notably, maternal PM2.5 exposure considerably elevated the abundances of L. reuteri, L. agilis, and B. producta in the liver of offspring. Although we failed to observe a significant increase in these bacteria within the gut of offspring from maternal PM2.5-exposed dams, Yang et al. did demonstrate that intestinal B. producta exacerbated non-alcoholic steatohepatitis by activating hepatic NF-κB signaling pathway through the production of 2-oleoylglycerol [43]. Furthermore, FMT of microbiota from maternal PM2.5-exposed offspring into recipient mice recapitulated the enhanced hepatic inflammatory response, confirming that maternal PM2.5 exposure-induced gut microbiota dysbiosis triggered liver inflammation in the offspring by activating the TLR4 signaling pathway. Mechanistically supporting our results, Dapito et al. demonstrated that gut microbiota promoted chronic liver injury, inflammation, fibrosis, and hepatocarcinogenesis through TLR4 activation, using germ-free and TLR4-inactivated mutant mouse models [18]. These results highlighted the critical role of gut microbiota-mediated TLR4 activation in the progression of liver injury in response to maternal PM2.5 exposure.

However, our study has some limitations: (i) while this study established a causal role of gut microbiota dysbiosis in maternal PM2.5-induced insulin resistance in the offspring, the mechanism by which maternal PM2.5 reprogrammed the offspring’s gut microbiota remains elusive; (ii) although pseudo-germ-free mice were used as FMT recipients in this study, it is important to note that ABX treatment cannot completely eliminate all endogenous microbes, potentially affecting the transplantation efficiency. Therefore, future investigations using germ-free mice are required to validate these findings; (iii) this study provided preliminary evidence for a potential role of H. hepaticus in maternal PM2.5 exposure-induced colonic barrier injury, hepatic inflammation, and insulin resistance in the offspring. However, future studies should colonize germ-free mice with either H. hepaticus or a cdtB (bioactive subunit of CDT)-deficient H. hepaticus strain to elucidate the underlying mechanisms; (iv) although this study indicated that maternal PM2.5 exposure facilitated gut-to-liver translocation of bacteria in offspring, the precise impacts of these translocated bacteria on hepatic metabolism and immune responses remain to be elucidated.

Conclusions

In summary, this study established the critical role of maternal PM2.5 exposure-induced gut microbiota dysbiosis in the development of insulin resistance in offspring. H. hepaticus was identified as a key pathogenic bacterium that contributed to disrupting the colonic barrier by promoting DNA damage. This colonic barrier injury, in turn, facilitated the translocation of gut bacteria and LPS to the liver, which activated the TLR4 signaling pathway, triggered hepatic inflammation, and ultimately led to insulin resistance. This study not only offered crucial etiological clues of maternal PM2.5 exposure for the MUHNW phenotype, which could pose long-term metabolic and cardiovascular risks to offspring, but also provided the potential intervention target of gut microbiota dysbiosis.

Supplementary Information

Supplementary Material 1 (22.9MB, docx)

Abbreviations

Apex1

Apurinicapyrimidinic endonuclease 1

ABX

Antibiotic cocktail

BAT

Brown adipose tissue

CCl4

Carbon tetrachloride

CDT

Cytolethal distending toxin

Cldn1/2/4/5

Claudin 1/2/4/5

DAO

Diamine oxidase

DEN

Diethylnitrosamine

DOHaD

Developmental Origins of Health and Disease

eWAT

Epididymal white adipose tissue

FA

Filtered air

FMT

Fecal microbiota transplantation

Gadd45a

Growth arrest and DNAdamageinducible 45 alpha

GD

Gestational day

GTT

Glucose tolerance test

H&E

Hematoxylin and eosin

HOMA-IR

Homeostasis model assessment of insulin resistance

IHC

Immunohistochemistry

ITT

Inulin tolerance test

iWAT

Inguinal white adipose tissue

LEfSe

Linear discriminant analysis effect size

LPS

Lipopolysaccharide

MUHNW

Metabolically unhealthy normal weight

Ocln

Occludin

Ogg1

8-oxoguanine DNA-glycosylase 1

PAMPs

Pathogen-associated molecular patterns

PCoA

Principal coordinate analysis

PM2.5

Fine particulate matter

PND

Postnatal day

qRT-PCR

Quantitative real-time polymerase chain reaction

SEM

Standard error of the mean

TLR4

Toll-like receptor 4

Trp53

Transformation related protein 53

Tjp1

Tight junction protein 1

Xrcc1

X-ray repair complementing defective repair in Chinese hamster cells 1

ZCMU

Zhejiang Chinese Medical University

Author contributions

RJH performed the experiments, generated and analyzed data, and drafted the manuscript. WBZ, WJF, RL, LZ, LQ, and HHZ performed the experiments and generated data. RJH and WBZ helped with data collection and analysis. WBZ, QHS, and CQL helped with manuscript edition. XZ, GFQ, and CQL designed the experiments. CQL is the guarantor of this work, had full access to all the data, and takes full responsibility for the integrity of data and the accuracy of data analysis. All authors read and approved the final manuscript.

Funding

This work was supported by the National Natural Science Foundation of China (Grant numbers 82273590 to C.L and 82173480 to R.L) and the Key Foundation of Zhejiang Chinese Medical University (Grant number 2025JKZDZC03 to C.L).

Data availability

The datasets generated during the current study are available from the corresponding author upon reasonable request.

Declarations

Ethics approval and consent to participate

All animal experimental procedures were approved by the Institutional Animal Care and Use Committee of Zhejiang Chinese Medical University.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.

Footnotes

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Renjie Hu and Wenbin Zhao have contributed equally to this work.

References

  • 1.Chew NWS, Ng CH, Tan DJH, Kong G, Lin C, Chin YH, et al. The global burden of metabolic disease: data from 2000 to 2019. Cell Metab. 2023;35(3):414-28.e3. 10.1016/j.cmet.2023.02.003. [DOI] [PubMed] [Google Scholar]
  • 2.Stefan N, Schick F, Häring HU. Causes, characteristics, and consequences of metabolically unhealthy normal weight in humans. Cell Metab. 2017;26(2):292–300. 10.1016/j.cmet.2017.07.008. [DOI] [PubMed] [Google Scholar]
  • 3.Huang H, Balzer NR, Seep L, Splichalova I, Blank-Stein N, Viola MF, et al. Kupffer cell programming by maternal obesity triggers fatty liver disease. Nature. 2025;644(8077):790–8. 10.1038/s41586-025-09190-w. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Huang Z, Niu R, Xu Q, Zhang R, Hu W, Qin Y, et al. Impact of maternal BPA exposure during pregnancy on obesity in male offspring: a mechanistic mouse study of adipose-derived exosomal miRNA. Environ Health Perspect. 2025;133(1): 17011. 10.1289/ehp14888. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Yao N, Kinouchi K, Katoh M, Ashtiani KC, Abdelkarim S, Morimoto H, et al. Maternal circadian rhythms during pregnancy dictate metabolic plasticity in offspring. Cell Metab. 2025;37(2):395-412.e6. 10.1016/j.cmet.2024.12.002. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Li R, Sun Q, Lam SM, Chen R, Zhu J, Gu W, et al. Sex-dependent effects of ambient PM(2.5) pollution on insulin sensitivity and hepatic lipid metabolism in mice. Part Fibre Toxicol. 2020;17(1): 14. 10.1186/s12989-020-00343-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Liu C, Bai Y, Xu X, Sun L, Wang A, Wang TY, et al. Exaggerated effects of particulate matter air pollution in genetic type II diabetes mellitus. Part Fibre Toxicol. 2014;11:27. 10.1186/1743-8977-11-27. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Liu C, Xu X, Bai Y, Wang TY, Rao X, Wang A, et al. Air pollution-mediated susceptibility to inflammation and insulin resistance: influence of CCR2 pathways in mice. Environ Health Perspect. 2014;122(1):17–26. 10.1289/ehp.1306841. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Chen M, Liang S, Qin X, Zhang L, Qiu L, Chen S, et al. Prenatal exposure to diesel exhaust PM(2.5) causes offspring β cell dysfunction in adulthood. Am J Physiol Endocrinol Metab. 2018;315(1):E72-e80. 10.1152/ajpendo.00336.2017. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Chen M, Liang S, Zhou H, Xu Y, Qin X, Hu Z, et al. Prenatal and postnatal mothering by diesel exhaust PM(2.5)-exposed dams differentially program mouse energy metabolism. Part Fibre Toxicol. 2017;14(1): 3. 10.1186/s12989-017-0183-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Hsu CL, Schnabl B. The gut-liver axis and gut microbiota in health and liver disease. Nat Rev Microbiol. 2023;21(11):719–33. 10.1038/s41579-023-00904-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Tilg H, Adolph TE, Trauner M. Gut-liver axis: pathophysiological concepts and clinical implications. Cell Metab. 2022;34(11):1700–18. 10.1016/j.cmet.2022.09.017. [DOI] [PubMed] [Google Scholar]
  • 13.Liu Y, Zhang L, Wang J, Sui X, Li J, Gui Y, et al. Prenatal PM(2.5) exposure associated with neonatal gut bacterial colonization and early children’s cognitive development. Environ Health. 2024;2(11):802–15. 10.1021/envhealth.4c00050. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Liu W, Zhou Y, Qin Y, Li Y, Yu L, Li R, et al. Sex-specific effects of PM(2.5) maternal exposure on offspring’s serum lipoproteins and gut microbiota. Sci Total Environ. 2020;739:139982. 10.1016/j.scitotenv.2020.139982. [DOI] [PubMed] [Google Scholar]
  • 15.Liu W, Zhou Y, Qin Y, Yu L, Li R, Chen Y, et al. Quercetin intervention alleviates offspring’s oxidative stress, inflammation, and tight junction damage in the colon induced by maternal fine particulate matter (PM(2.5)) exposure through the reduction of bacteroides. Nutrients. 2020. 10.3390/nu12103095. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Wang X, Fang Y, Liang W, Cai Y, Wong CC, Wang J, et al. Gut-liver translocation of pathogen Klebsiella pneumoniae promotes hepatocellular carcinoma in mice. Nat Microbiol. 2025;10(1):169–84. 10.1038/s41564-024-01890-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Valenti L, Fracanzani AL, Fargion S. The immunopathogenesis of alcoholic and nonalcoholic steatohepatitis: two triggers for one disease? Semin Immunopathol. 2009;31(3):359–69. 10.1007/s00281-009-0152-9. [DOI] [PubMed] [Google Scholar]
  • 18.Dapito DH, Mencin A, Gwak GY, Pradere JP, Jang MK, Mederacke I, et al. Promotion of hepatocellular carcinoma by the intestinal microbiota and TLR4. Cancer Cell. 2012;21(4):504–16. 10.1016/j.ccr.2012.02.007. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Hu RJ, Pan J, Zhu JY, Li SD, Zhang GQ, Wang PP, et al. β3-adrenergic receptor methylation mediates fine particulate matter inhalation-impaired white adipose tissue browning. Nano Today. 2024;55: 13. 10.1016/j.nantod.2024.102167. [Google Scholar]
  • 20.Bárcena C, Valdés-Mas R, Mayoral P, Garabaya C, Durand S, Rodríguez F, et al. Healthspan and lifespan extension by fecal microbiota transplantation into progeroid mice. Nat Med. 2019;25(8):1234–42. 10.1038/s41591-019-0504-5. [DOI] [PubMed] [Google Scholar]
  • 21.Hu R, Zhang L, Qin L, Ding H, Li R, Gu W, et al. Airborne PM(2.5) pollution: a double-edged sword modulating hepatic lipid metabolism in middle-aged male mice. Environ Pollut. 2023;324:121347. 10.1016/j.envpol.2023.121347. [DOI] [PubMed] [Google Scholar]
  • 22.Erben U, Loddenkemper C, Doerfel K, Spieckermann S, Haller D, Heimesaat MM, et al. A guide to histomorphological evaluation of intestinal inflammation in mouse models. Int J Clin Exp Pathol. 2014;7 8:4557–76. [PMC free article] [PubMed] [Google Scholar]
  • 23.Liu X, Liu C, Qian X, Zhang S, Yao Z, Chai Y, et al. Fecal microbiota transplantation alleviated heat-induced colonic tissue damage, epithelial apoptosis, and oxidative stress. Appl Environ Microbiol. 2025;91(10):e0097625. 10.1128/aem.00976-25. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Nejman D, Livyatan I, Fuks G, Gavert N, Zwang Y, Geller LT, et al. The human tumor microbiome is composed of tumor type-specific intracellular bacteria. Science. 2020;368(6494):973–80. 10.1126/science.aay9189. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Fuks G, Elgart M, Amir A, Zeisel A, Turnbaugh PJ, Soen Y, et al. Combining 16S rRNA gene variable regions enables high-resolution microbial community profiling. Microbiome. 2018;6(1): 17. 10.1186/s40168-017-0396-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Takeuchi T, Kubota T, Nakanishi Y, Tsugawa H, Suda W, Kwon AT, et al. Gut microbial carbohydrate metabolism contributes to insulin resistance. Nature. 2023;621(7978):389–95. 10.1038/s41586-023-06466-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Tilg H, Adolph TE, Dudek M, Knolle P. Non-alcoholic fatty liver disease: the interplay between metabolism, microbes and immunity. Nat Metab. 2021;3(12):1596–607. 10.1038/s42255-021-00501-9. [DOI] [PubMed] [Google Scholar]
  • 28.Zhao L, Zhang F, Ding X, Wu G, Lam YY, Wang X, et al. Gut bacteria selectively promoted by dietary fibers alleviate type 2 diabetes. Science. 2018;359(6380):1151–6. 10.1126/science.aao5774. [DOI] [PubMed] [Google Scholar]
  • 29.Fox JG, Ge Z, Whary MT, Erdman SE, Horwitz BH. Helicobacter hepaticus infection in mice: models for understanding lower bowel inflammation and cancer. Mucosal Immunol. 2011;4(1):22–30. 10.1038/mi.2010.61. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Ahn EH, Liu X, Alam AM, Kang SS, Ye K. Helicobacter hepaticus augmentation triggers Dopaminergic degeneration and motor disorders in mice with Parkinson’s disease. Mol Psychiatry. 2023;28(3):1337–50. 10.1038/s41380-022-01910-2. [DOI] [PubMed] [Google Scholar]
  • 31.Bezine E, Vignard J, Mirey G. The cytolethal distending toxin effects on Mammalian cells: a DNA damage perspective. Cells. 2014;3(2):592–615. 10.3390/cells3020592. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32.Hu R, Li R, Wang L, Li S, Chu H, Zhang L, et al. Maternal fine particulate matter exposure impairs inguinal white adipose tissue plasticity in middle-aged male mouse offspring. ACS Nano. 2026;20(6):5054–71. 10.1021/acsnano.5c18921. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Zhou YC, Liu JM, Zhao ZP, Zhou MG, Ng M. The national and provincial prevalence and non-fatal burdens of diabetes in China from 2005 to 2023 with projections of prevalence to 2050. Military Med Res. 2025;12(1): 28. 10.1186/s40779-025-00615-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34.Kautzky-Willer A, Leutner M, Harreiter J. Sex differences in type 2 diabetes. Diabetologia. 2023;66(6):986–1002. 10.1007/s00125-023-05891-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Li F, Xing X, Jin Q, Wang XM, Dai P, Han M, et al. Sex differences orchestrated by androgens at single-cell resolution. Nature. 2024;629(8010):193–200. 10.1038/s41586-024-07291-6. [DOI] [PubMed] [Google Scholar]
  • 36.Tonnus W, Maremonti F, Gavali S, Schlecht MN, Gembardt F, Belavgeni A, et al. Multiple oestradiol functions inhibit ferroptosis and acute kidney injury. Nature. 2025;645(8082):1011–9. 10.1038/s41586-025-09389-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Azziz R, Carmina E, Chen Z, Dunaif A, Laven JS, Legro RS, et al. Polycystic ovary syndrome. Nat Rev Dis Primers. 2016;2:16057. 10.1038/nrdp.2016.57. [DOI] [PubMed] [Google Scholar]
  • 38.Gao A, Su J, Liu R, Zhao S, Li W, Xu X, et al. Sexual dimorphism in glucose metabolism is shaped by androgen-driven gut microbiome. Nat Commun. 2021;12(1):7080. 10.1038/s41467-021-27187-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39.Vieira-Potter VJ, Mishra G, Townsend KL. Health of adipose tissue: oestrogen matters. Nat Rev Endocrinol. 2026;22(2):76–91. 10.1038/s41574-025-01180-2. [DOI] [PubMed] [Google Scholar]
  • 40.Yang Z, Zhang Y, Ran S, Zhang J, Gao Y, Zhang Y, et al. Exposure to ambient air pollution over developmental stages induced neurodevelopmental impairment in mice offspring via microbiome-gut-brain axis. Part Fibre Toxicol. 2025;22(1):20. 10.1186/s12989-025-00637-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41.Cao S, Zhang Y, Bao R, Wang T, Zhu L, Zhang Q. Helicobacter hepaticus promotes liver fibrosis through oxidative stress induced by hydrogenase in BALB/c mice. Helicobacter. 2023;28(5):e13001. 10.1111/hel.13001. [DOI] [PubMed] [Google Scholar]
  • 42.Su D, Nie Y, Zhu A, Chen Z, Wu P, Zhang L, et al. Vitamin D signaling through induction of Paneth cell defensins maintains gut microbiota and improves metabolic disorders and hepatic steatosis in animal models. Front Physiol. 2016;7:498. 10.3389/fphys.2016.00498. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 43.Yang M, Qi X, Li N, Kaifi JT, Chen S, Wheeler AA, et al. Western diet contributes to the pathogenesis of non-alcoholic steatohepatitis in male mice via remodeling gut microbiota and increasing production of 2-oleoylglycerol. Nat Commun. 2023;14(1):228. 10.1038/s41467-023-35861-1. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary Material 1 (22.9MB, docx)

Data Availability Statement

The datasets generated during the current study are available from the corresponding author upon reasonable request.

Articles from Particle and Fibre Toxicology are provided here courtesy of BMC

RESOURCES