Aging-related changes in gut bacteria weaken the intestinal barrier and increase risk of sepsis

To determine whether aging increases susceptibility to sepsis, the mortality rates within 36 h following cecal ligation and puncture (CLP) between aged (22–24 months) and young (8–10 weeks) mice were compared. Results indicated a significant increase in mortality in the aged group compared to the young group (Figure 1A). Specifically, about 50% of the young mice died from sepsis within 15 h after CLP, whereas all aged mice succumbed to the condition. Furthermore, to determine whether aging further exacerbates sepsis-induced intestinal barrier dysfunction, we performed histopathological assays, immunohistochemistry, Western blot analysis, and enzyme-linked immunosorbent assay (ELISA). The results showed that that aged septic mice exhibited more severe intestinal injuries (Figure 1B−C), reduced numbers of goblet cells (Figure 1D−E), and decreased Muc2 expression (Figure 1F−G) compared to young mice. Post-CLP analysis showed reduced expression levels of Nlrp6, Occludin, Claudin3, and autophagy-related proteins (Atg7 and LC3II/LC3I) in the aged group (Figure 1H−1I). Additionally, the serum concentrations of TNF-α and IL-1β were also higher in the aged group (Figure 1J). Similarly, intestinal permeability assays indicated higher serum levels of fluorescein isothiocyanate (FITC) and lipopolysaccharide (LPS) in the aged group than in the young group (Figure 1K). Quantitative reverse transcriptase polymerase chain reaction (qRT‒PCR) analyzes of colon tissues showed increased mRNA expression of Tnf-α, Cxcl1, and Ccl2 in aged septic mice (Figure 1L). These findings collectively suggest that aging enhances susceptibility to sepsis by exacerbating intestinal injury and barrier dysfunction, accompanied by an intensified inflammatory response.

Subsequently, we compared inflammatory markers and intestinal injury parameters between aged and young sham-operated mice to determine whether the worse outcomes of sepsis in aged individuals result from intrinsic aging effects or age-specific responses during sepsis. While some inflammatory markers were elevated in aged mice, indicators of intestinal damage and barrier function showed no differences between the groups (Figure S1), which suggests that the exacerbation of sepsis outcomes in aging may primarily stem from age-specific pathological mechanisms during sepsis progression. We hypothesize that this phenomenon could be linked to gut microbiota dysbiosis, though further validation is needed.

Intestinal microbiota have been implicated in the pathogenesis of sepsis, and their composition is altered by aging.21 To investigate whether changes in the gut microbiota could directly contribute to increased sepsis susceptibility in aged individuals, cecal contents (CCs) from both aged and young mice were collected for analysis. Intraperitoneal injections of these CCs into young mice demonstrated similar sepsis susceptibility across both groups (Figure S2A‒S2D). Subsequently, we eradicated gut microbiota by administering broad-spectrum antibiotics (ABX) for 3 d prior to treatment with CLP based on the previous studies.22^,23 The survival rates of aged and young septic mice with ABX pretreatment were comparable (Figure S2E). Additionally, the results revealed no significant differences in histopathological scores, numbers of goblet cells and Muc2 expression (Figure S2F–S2K) compared to young septic mice. Similarly, serum levels of TNF-α and the mRNA expression of Tnf-α and Ccl2 in the colon tissues showed no differences between the young and aged septic mice with ABX pretreatment (Figure S2L−2M). Consistent with the septic mice subjected to ABX pretreatment, both aged and young mice of sham groups exhibited a similarity on intestinal damage and inflammatory cytokines after pretreatment with ABX (Figure S2N−2U).

To investigate the impact of intestinal dysbiosis on sepsis progression in aging, we first performed fecal microbiota transplantation (FMT) using donor specimens from aged mice and human patients into the young mice of pseudo-germ-free (PGF) treatment by ABX. To further confirm the reproducibility of our FMT methodology, we subsequently recollected fecal specimens from both young and aged septic mice, as well as from patients, and carried out an additional round of FMT experiments, which further substantiate the efficacy of the approach. The results demonstrated that microbial communities from both donor septic mice (Figure S3A−3B) and patients (Figure S3C−3D) were capable of robust engraftment in the recipient PGF mice. The survival analysis showed that young mice receiving feces from young donors exhibited greater resistance to sepsis compared to those receiving feces from older donors, regardless of whether the fecal source was mice or patients (Figure 2A−2B). Histopathological examinations revealed that young mice and patients gavaged with feces from aged mice developed more severe intestinal injuries (Figure 2C−2F), reduced numbers of goblet cells (Figure 2G−2J), and decreased Muc2 expression (Figure 2K−2N) in response to sepsis than those gavaged with feces from young mice. Additionally, the increased serum cytokine levels and the elevated expressions of cytokines and chemokines in colon tissues from mice that received feces from aged septic mice compared to those that received feces from young septic mice following CLP treatment, and aged feces recipients exhibited higher serum FITC and LPS compared to young feces recipients (Figure 2O−2P). Consistently, recipients of aged septic patients feces exhibited higher LPS and cytokine levels in response to sepsis (Figure 2Q−2R). Collectively, these findings suggest that microbiome alterations associated with aging significantly increase susceptibility to sepsis.

We investigated the potential mechanisms by which changes in microbiota during aging increase susceptibility to sepsis. 16S rDNA sequencing revealed that the aged group exhibited distinct alpha diversity and marked intergroup separation in gut microbiota composition (Figure 3A). Alpha and beta diversities were similar between aged and young patients (Figure S4A - S4B). Additionally, aging led to alterations in the genus-level composition of intestinal microbiota in patients (Figure S4C). Linear discriminant analysis (LDA) indicated a higher abundance of Klebsiella in aged septic patients and mice (Figure 3B−3C). To explore the impact of aging on the metabolic profile of intestinal microbiota, metabonomic analyzes were conducted, revealing significant shifts in fecal microbiota metabolic formations in aged septic mice, as demonstrated by principal coordinate analysis (PCoA) (Figure 3D). The enrichment analysis revealed that differential metabolites were predominantly involved in histidine metabolism (Figure 3E). Given that histamine is the principal metabolic product of histidine and its established role in immune regulation, we selected HA as our primary target. This choice was further supported by volcano plot (Figure 3F) and heatmap analyzes (Figure S4D) showing increased HA levels in aged septic mice. Subsequent targeted LC‒MS/MS quantification confirmed significantly higher HA concentrations in aged versus young hosts (Figure 3G), suggesting aging potentiates HA accumulation during sepsis.

Upon confirming the concurrent increase in Klebsiella and HA in aged septic hosts, we observed a significant rise in K. aero abundance in these patients (Figure S4E). We measured the relative abundance of K. aero in fecal samples from aged and young septic patients and found that the relative levels of K. aero were significantly higher in aged patients compared to the young counterparts (Figure S4E). Subsequently, in vitro experiments revealed that HA concentrations in the culture supernatant of K. aero were higher compared to those in the control culture medium (Figure S4F). Given the high variability of the histidine decarboxylase (HDC) gene in Gram-negative bacteria,24 we aimed to further clarify the mechanism of HA production. We assessed HA levels in Klebsiella oxytoca (K. oxy), K. aero, a K. aero isolate engineered to lack the HDC gene (K. aero^-HDC), and control group mice following CLP treatment. The results indicated that only K. aero could produce significant levels of HA, whereas K. oxy, K. aero^-HDC, and control groups exhibited comparable HA levels (Figure 3H). Additionally, in vitro deletion of the HDC gene in K. aero strains effectively inhibited HA production in the supernatants (Figure S4G). Corresponding with the increased presence of K. aero, aging also led to elevated serum HA concentrations in septic mice (Figure S4H), suggesting a critical role for K. aero in the progression of sepsis through enhanced HA production during aging.

To assess the influence of K. aero on sepsis, cytokine levels were measured using ELISA. The results showed that young septic mice pretreated with K. aero exhibited higher serum levels of TNF-α and IL-1β compared to those not receiving K. aero (Figure 4A). Similarly, K. aero pretreatment was associated with decreased survival rates in aged septic mice (Figure 4B). Consequently, the proinflammatory role of K. aero during sepsis was further evaluated to determine its association with HA production. As anticipated, HA elevated the serum concentrations of TNF-α and IL-1β, as well as decreased survival rates in septic mice (Figure 4C−4D). Histological assays indicated that HA treatment exacerbated pathological changes in the colon of septic mice (Figure 4E−4F). Immunohistochemical analyzes showed that HA treatment reduced Muc2 expression and goblet cell numbers in the colon (Figure S5A–S5D). Furthermore, deletion of the HDC gene in K. aero strains alleviated its detrimental effects on pathological changes in the colon and inflammatory response in septic mice (Figure 4G−4I). Collectively, these findings suggest that both K. aero and HA exacerbate the effects of sepsis, influencing inflammatory responses and survival outcomes.

Increased levels of K. aero and its metabolite HA during aging are associated with impaired intestinal autophagy, compromised barrier function, elevated inflammatory cytokine levels, increased gut microbiota translocation, and heightened susceptibility to sepsis. To elucidate how HA contributes to increased susceptibility to sepsis in aged mice, we collected colonic tissues from both aged and young septic mice for transcriptomic analysis. Differentially expressed genes were mainly enriched in pathways related to bacterial gene ontology (GO) functions, immune responses, response to bacterium and innate immune processes (Figure S5E). Nlrp6 is a key immune sensor in the gut that serves as a critical receptor for intestinal bacteria and their metabolites.25^,26 Previous studies have suggested that Nlrp6 may also be a target of histamine produced by gut microbiota.26 Consistently, our results showed reduced Nlrp6 expression in aged septic mice compared to the young group (Figure S5F). Therefore, we hypothesized that HA contributes to sepsis, intestinal injury, and barrier dysfunction in aged mice by downregulating Nlrp6 expression and impairing intestinal autophagy.

To assess whether HA exacerbates intestinal permeability and systemic inflammation by inhibiting Nlrp6 expression in septic mice, we used ELISA to quantify serum concentrations of FITC, LPS, and cytokines. Systemic inflammatory responses were compared across different groups. Serum levels of FITC, LPS, TNF-α, and IL-1β were significantly higher in the aged group than in the young group. Notably, Nlrp6 overexpression reduced TNF-α, IL-1β, FITC, and LPS levels compared with the CLP group. However, HA intervention led to increased inflammatory response and greater intestinal barrier permeability compared to the adeno-associated virus-mediated Nlrp6 overexpression combined with CLP (AAV_Nlrp6+CLP) group (Figures 5A and S6A). These findings suggest that HA may promote inflammation and intestinal permeability in septic mice by downregulating Nlrp6 expression. These effects were especially pronounced in aged septic mice.

To further investigate the role of HA in mediating intestinal injury and barrier dysfunction through inhibition of Nlrp6 expression, we performed H&E and AB–PAS staining, immunohistochemical analysis, and WB assays. The findings indicated that aged septic mice exhibited more severe epithelial cell shedding in the upper and middle colon than young septic mice, along with pronounced villus shortening and widening and severe inflammatory infiltration, resulting in higher Chiu's scores. Nlrp6 overexpression markedly improved inflammatory infiltration, villus shortening, and epithelial shedding in septic mice, resulting in lower Chiu's scores compared with the CLP group. However, subsequent HA intervention significantly worsened these pathological changes, as indicated by shorter villi, increased ulcerative and inflammatory infiltration, and higher Chiu's scores compared with the AAV_Nlrp6+CLP group (Figures 5B−5C). These results indicate that HA may worsen intestinal injury by suppressing Nlrp6 expression in septic mice, with this effect being more pronounced in aged mice.

Additionally, AB-PAS staining and immunohistochemical analysis showed a decrease in goblet cell numbers and Muc2 expression in the colonic tissues of aged mice compared with the young group. Nlrp6 overexpression in septic mice further reduced goblet cell numbers and Muc2 expression compared with the CLP group. Notably, HA administration also reduced goblet cell numbers and Muc2 expression relative to the AAV_Nlrp6+CLP group (Figure 5D–G). Furthermore, the expression of the tight junction proteins Claudin1 and Claudin3 was higher in the AAV_Nlrp6+CLP group than in the CLP group. HA treatment reduced the expression of these proteins compared with the AAV_Nlrp6+CLP group (Figures 6A−6B and S6B−S6C). These findings demonstrate that HA impairs the intestinal mucus barrier by inhibiting Nlrp6 expression in septic mice, an effect that is more pronounced in aged mice. In summary, HA inhibits Nlrp6 expression, thereby promoting intestinal injury, barrier dysfunction, and inflammatory responses in septic mice, with these effects being amplified by aging.

To investigate whether HA influences autophagy in intestinal epithelial cells of septic mice by modulating Nlrp6 expression, we utilized WB assays, TEM, and confocal microscopy to assess autophagy-related proteins and autophagic flux. Additionally, we employed Caco2 cells to examine the induction of cellular senescence by H₂O₂. We confirmed successful senescence induction using β-galactosidase staining and flow cytometry to analyze the cell cycle (Figure S7A−S7B).

WB analysis of colon tissues revealed that Nlrp6, LC3, and Atg7 levels were reduced in the aged group. In the AAV_Nlrp6+CLP group, Nlrp6, LC3, and Atg7 expression was increased compared with the CLP group. However, HA intervention reduced the expression of Nlrp6, LC3, and Atg7 compared with the AAV_Nlrp6+CLP group (Figure 6A−6B). Protein expression in Caco2 cells mirrored that in the intestinal tissues of septic mice, confirming the reliability of both cellular and tissue models (Figure S7C−S7D). These findings indicate that HA exacerbates autophagy inhibition and increases intestinal barrier permeability by downregulating Nlrp6 expression.

TEM showed fewer autophagosomes in the intestinal tissues of the aged group compared with the young group. The number of autophagosomes increased in the AAV_Nlrp6+CLP group compared with the CLP group, while HA treatment reduced autophagosome formation compared with the AAV_Nlrp6+CLP group, indicating that HA impedes autophagosome formation by inhibiting Nlrp6 expression (Figure 6C). Similarly, confocal microscopy showed that autophagic flux in Caco2 cells from the aged group was lower than in those from young septic mice. Nlrp6 overexpression significantly enhanced autophagic flux compared with the LPS group, while HA treatment suppressed autophagic flux compared with the AAV_Nlrp6+CLP group (Figure S7). These results indicate that HA inhibits autophagic flux by suppressing Nlrp6 expression.

To elucidate the mechanism by which HA or aging suppresses autophagy via inhibition of Nlrp6 expression and its interaction with LC3, we performed colocalization fluorescence and coimmunoprecipitation (Co-IP) assays. Immunofluorescence analysis showed reduced colocalization of Nlrp6 and LC3 in colonic tissue from the aged group compared with the young group. In contrast, colocalization increased in the AAV_Nlrp6+CLP group compared with the CLP group, whereas HA treatment reduced co-localization compared with the AAV_Nlrp6+CLP group (Figure 6D). Similar results were observed in Caco2 cells, confirming consistency across experimental models (Figure S7F). Co-IP assays further confirmed that Aged may be associated with a reduction of the interacI on between Nlrp6 and LC3 compared with the young group (Figure S8). Collectively, these findings indicate that HA inhibits autophagy by suppressing Nlrp6 expression, thereby reducing the interaction between Nlrp6 and LC3 and contributing to autophagy deficiency in intestinal cells of septic mice. This effect is more pronounced in aged mice, highlighting increased vulnerability to HA-mediated autophagy suppression with aging (Graphical Abstract).

Despite our findings that clearly demonstrate the mediating role of Nlrp6 in linking intestinal flora imbalance to barrier dysfunction in aged sepsis hosts, the mechanisms by which HA inhibits Nlrp6 expression in this context remain incompletely clarified. Additionally, while this study focused on the colon due to its high Muc2 and Nlrp6 expression and central role in mucus barrier maintenance, sepsis-induced damage in the jejunum and ileum warrants further investigation. Future studies will systematically compare regional responses to delineate segment-specific mechanisms. Moreover, our study was specifically focused on intestinal barrier dysfunction in aging sepsis hosts, without a stratified age analysis, which limits the generalizability of our results across different age groups. The relationship between intestinal flora imbalance and infectious diseases across these groups has not been clearly defined. Finally, the underlying reasons for the observed increase in K. aero prevalence in aging also warrant further investigation to fully understand their implications and mechanisms.

The CLP-induced sepsis mouse model was generated following previously described protocols.59 The mice were deeply anesthetized, and a 1–2 cm midline laparotomy was performed under sterile conditions. Half of the distal cecum was ligated and punctured once with a 21 G needle, allowing a small amount of fecal content to be extruded from the perforation site. The incision was then sutured layer by layer, and 1 ml of sterile normal saline was administered subcutaneously for resuscitation.

To explore the effects of HA on septic mice, 20 mg/Kg of HA (MedChemExpress, Monmouth Junction, NJ, USA) dissolved in sterile water was administered intraperitoneally before CLP. Additionally, the mice received intravenous treatment with adeno-associated-virus-9 (AAV9) targeting Nlrp6 (Hanbio Tech, Shanghai, China) at a dose of 1.4 × 10¹¹ Vg per mouse 3‒4 weeks prior to CLP to induce overexpression of the Nlrp6 gene. Furthermore, HA was administered intraperitoneally to AAV9-Nlrp6 mice 2 h before CLP. All the samples were collected from all mice 8 h after either sham operation or cecal ligation and puncture. The mice were euthanized according to the specific requirements of the study.

CCs were harvested from both young and aged mice and then suspended at a 1:10 dilution with phosphate-buffered saline (PBS) based on weight/volume. Subsequently, 500 µl of the CC suspension was injected intraperitoneally into young mice to induce septic peritonitis.

Fresh fecal samples were collected from both young and aged septic patients and were immediately frozen at −80 °C for preservation. The baseline characteristics of the septic patients are presented in Table S1. These fecal samples were obtained in accordance with the Sepsis 3.0 criteria from the First Affiliated Hospital of Zhengzhou University. The study protocols were approved by the Institutional Ethical Committee, with approval numbers 2021-KY-0467-003 and 2022-KY-0215.

Caco2 cells (h032, iCell) were cultured following the manufacturer's instructions. To induce senescence, the cells were treated with 200 μmol/L of 30% hydrogen peroxide (H₂O₂, Sigma-Aldrich (Shanghai) Trading Co., Ltd.) for 2 h, followed by a recovery period of 72 h (termed the aged group). For the overexpression of the Nlrp6 gene, Caco2 cells were seeded at a density of 4 × 10⁵ cells per ml using the pcDNA3.1 Nlrp6 vector (Genechem, Shanghai, China). The vector was constructed with the forward primer 5′-CACACTGGACTAGTGGATCCCGCCACCATGGACCAGCCAGAGGCCCCCTGCTCC-3′ and the reverse primer 5′-AGTCACTTAAGCTTGGTACGAAGGTCGAGATGAGTTCCTTGGGAGG-3′. Transfection was carried out using Lipofectamine™ 3000 (ThermoFisher, MA, USA) according to the manufacturer's guidelines, with protein expression analyzed 36 h post-transfection.

Additionally, the mRFP-GFP-LC3 vector (Hanbio Tech, Shanghai, China) was transfected into Caco2 cells as per the manufacturer's instructions. Autophagic flux was assessed using confocal microscopy 36 h after transfection.

K. aero was cultured in Luria-Bertani (LB) medium at 37 °C with shaking at 180 rpm for 10 h under aerobic conditions. For the animal experiments, mice received intragastric administration of K. aero at a dose of 2 × 10⁸ CFU per mouse once daily for 3 d. Subsequently, the mice were subjected to CLP treatment.

To deplete the intestinal microbiota, mice were treated orally with a broad-spectrum antibiotic cocktail consisting of 100 mg/Kg vancomycin, 200 mg/Kg neomycin sulfate, 200 mg/Kg metronidazole, and 200 mg/Kg ampicillin for 3 d.16 Following this treatment, approximately 3 g of fecal material from either aged or young septic mice was resuspended in PBS to achieve a fecal concentration of 0.125 g/ml. This suspension was then administered intragastrically to young recipient mice. These recipient mice were subsequently utilized for further experimental assessments.

Total genomic DNA was extracted from cecal feces stored at −80 °C using the cetyltrimethylammonium bromide/sodium dodecyl sulfate (CTAB/SDS) method. The extracted DNA was diluted to a concentration of 1 ng/μl with sterile water for subsequent analyzes. Amplification of the 16S rDNA V3-V4 regions was performed using specific primers (341F: CCTAYGGGRBGCASCAG; 806 R: GGACTACNNGGGTATCTAAT) that included a barcode for identification. The PCR products were sequenced on the Illumina NovaSeq 6000 platform.

Furthermore, we also conducted metagenomic sequencing, with the specific steps as follows: the total DNA was extracted from samples using the Fecal Genome DNA Extraction Kit (AU46111-96, BioTeke, China). DNA libraries were constructed using the TruSeq Nano DNA Library Preparation Kit-Set (#FC-121-4001, Illumina, USA) following the manufacturer's instructions. Metagenome libraries were then sequenced on an Illumina NovaSeq 6000 platform with PE150. Fastp software (v0.23.4) were used to remove the reads that contained adapter contamination, low-quality bases and undetermined bases. The sequence quality was subsequently verified using fastp. The quality-filtered reads were first aligned to the genome by using bowtie (v2.2) to filter out host contaminations. Then, the remaining reads were subjected to de novo assembly for each sample using MEGAHIT (v1.2.9) and used to assign microbial functions and taxonomy. Metaprodiga(v2.6.3) was used to predict the coding regions (CDS) of the assembled contigs, and the CDS sequences of all the samples were clustered using MMseq2 (v15-6f452) to obtain unigenes. DIAMOND (v 0.9.14) was used to perform a taxonomic assessment of the microbiota based on the NR database. The Wilcoxon test was used to identify the differentially abundant species, and significances were declared at p < 0.05 and |log2-fold change|>1.

We constructed the HDC-deficient mutant of K. aero using double homologous recombination. Initially, single clones were inoculated in liquid medium with corresponding antibiotic resistance and cultured overnight at 37 °C to confirm antibiotic resistance. Genomic DNA was extracted from these cultures using the PureLink™ Microbiome DNA Purification Kit (A29790↗, Thermo). This DNA served as a template to amplify the HDC knockout regions using the upper and lower homology arms. The primer sequences used were: HDC-19-F/HDC-up-R: TGTAAAACGACGGCCAGTTATTTTCCAGTACCGCACGCATA/ACATAAATATGGAGATTCATGAGTGAGCAGCGGCAAACTG;HDC-down-F/HDC-19-R:GTTTGCCGCTGCTCACTCATGAATCTCCATATTTATGTTG/CTATGACCATGATTACGCCATGTCCCGATGAAAGGTCTGATT.

The gel-purified PCR fragments of the HDC gene were used for subsequent ligation. The purified HDC UP and HDC DOWN fragments were ligated into the pUC19 vector backbone. This construct was transformed into E. coli DH5α competent cells, and colonies were selected on Amp-IPTG-X-Gal plates using the blue‒white screening method. Positive clones were confirmed by Sanger sequencing to determine whether the fragment had the intended mutations.

The extracted plasmids from the verified positive clones served as templates for amplifying the HDC repair homology arm using primers HDC-up-F and HDC-down-R. The PCR products were then digested and purified, ready for CRISPR/Cas9 gene editing.

For transformation, 10 µl of plasmid containing Cas9 was added to the prepared competent cells and kept on ice for 5 min. The mixture was then electrotransformed at 2500 V, followed by the addition of 1 ml of LB medium and incubation at 37 °C for 1 h. Transformed cells were plated on kanamycin-resistant plates to select positive clones.

Cas9-expressing competent cells were prepared similarly and used for a subsequent transformation with the hdc repair homology arm and the prepared hdc sgRNA plasmid. After transformation, cells were plated on kanamycin/tetracycline plates to select for HDC deletion strains. Positive HDC-deficient clones were identified using primers HDC-up-F (TATTTTCCAGTACCGCACGCATA) and HDC-down-R (ATGTCCCGATGAAAGGTCTGATT) and further verified by Sanger sequencing.

The concentrations of TNF-α, IL-1β, LPS, and HA were quantified using ELISA kits provided by Cusabio and Elabscience, both located in Wuhan, China. All assays were performed strictly according to the manufacturer's instructions.

Total RNA was extracted using TRIzol reagent (Solarbio, Beijing, China). Reverse transcription was performed using a TaqMan Reverse Transcription Kit (UE, Suzhou, China). Quantitative PCR was conducted using qPCR SuperMix (Yeasen, Shanghai, China), and the results were analyzed using the 2^−ΔΔCT method to quantify gene expression levels. Details of all the qRT‒PCR primers used are provided in Table S2.

Caco2 cells were cultured at a density of 4 × 10⁵ cells per ml. The pcDNA3.1 Nlrp6 plasmid with a Flag tag was used to achieve targeted overexpression of Nlrp6 in Caco2 cells. Transfection was carried out using Lipofectamine™ 3000 (ThermoFisher, MA, USA) following the manufacturer's instructions, with the transfection process continuing for 12 h. Subsequently, transfection with mRFP-GFP-LC3 (Hanbio Tech, Shanghai, China) was performed according to the manufacturer's guidelines, and 36 h post-transfection, the cells were incubated overnight at 4 °C with an anti-Nlrp6 antibody (Bioss, Beijing, China). The cells were then incubated with Dynabeads™ Protein G magnetic beads (Invitrogen, CA, USA) at 4 °C to facilitate the immunoprecipitation. WB analysis was subsequently performed using Nlrp6 antibodies (Bioss, Beijing, China) and LC3 antibodies (CST, MA, USA) to detect the proteins of interest.

Caco2 cells and colon tissues were lysed using RIPA buffer (Solarbio, Beijing, China) on ice, followed by denaturation in SDS loading buffer by heating for 10 min. WB was conducted according to the manufacturer's guidelines. The proteins were probed using the following primary antibodies: β-actin (Abcam, MA, USA), LC3 (CST, MA, USA), Nlrp6 (Bioss, Beijing, China), Atg7 (CST, MA, USA), p62 (CST, MA, USA), Occludin (Proteintech, Wuhan, China), Claudin1 (Abcam, MA, USA), and Claudin3 (Abcam, MA, USA). Secondary antibodies were sourced from Proteintech (Wuhan, China). Protein expression was detected using an enhanced chemiluminescence (ECL) assay (UE, Suzhou, China).

Colon tissues from septic mice were collected and fixed in 4% paraformaldehyde for 24 h. The tissues were then embedded in paraffin, and histological sections were prepared for H&E staining. The severity of intestinal injury was evaluated based on parameters such as congestion, intestinal wall thickening, ulceration, and adhesion. These features were scored using a standardized scale ranging from 0 (absent) to 4 (severe), as previously described.60 The stained sections were examined under a microscope (NanoZoomer S60, Hamamatsu, Japan).

To assess apoptosis and evaluate cellular senescence, we utilized the FITC Annexin V Apoptosis Detection Kit I (BioLegend, CA, USA). Caco2 cells were first washed twice with cold PBS and then resuspended in 1× binding buffer to a concentration of 1 × 10⁶ cells/ml. Next, 100 µl of this cell suspension (1 × 10⁵ cells) was transferred to a 5 ml culture tube. To each tube, 5 µl of FITC Annexin V and 5 µl of propidium iodide (PI) were added. After gently vortexing, the cells were incubated for 15 min at room temperature in the dark. Following incubation, 400 µl of 1× binding buffer was added to each tube. Flow cytometric analysis was conducted within 1 h to ensure optimal assessment of the cells.

Paraffin-embedded tissue sections were dewaxed and rehydrated before being incubated with 3% hydrogen peroxide for 20 min to block endogenous peroxidase activity. The sections were then blocked with goat serum for 30 min to prevent nonspecific binding. Subsequently, the sections were incubated overnight at 4 °C with an anti-Muc2 antibody (Servibio, Wuhan, China). Following this, sections were incubated with a biotinylated secondary antibody and streptavidin-coupled horseradish peroxidase (HRP). After three washes, the sections were stained with diaminobenzidine (DAB) and counterstained with hematoxylin. For each sample, at least three fields were randomly selected for observation. Images were captured and analyzed using ImageJ software.

Immunofluorescence colocalization of Nlrp6 and LC3 was performed in both Caco2 cells and colon paraffin-embedded tissue sections following the manufacturer's instructions. Initially, both cells and tissue sections were incubated overnight at 4 °C with a primary antibody against Nlrp6 (Bioss, Beijing, China). For detection, tissue sections were incubated with Cy3-conjugated donkey anti-rabbit IgG (Servicebio, Wuhan, China), and cells with iF488-Tyramide (Servicebio, Wuhan, China) for 1 h at room temperature. Afterward, a 10min incubation with Tyramide signal amplification (TSA) was conducted in the dark, with the tissue sections also undergoing antigen retrieval using an antigen repair buffer. Subsequently, cells and tissue sections were incubated overnight at 4 °C with an antibody against LC3 (CST, MA, USA). For visualization, cells were treated with Cy3-conjugated donkey anti-rabbit IgG (Servicebio, Wuhan, China), and tissue sections with iF488-Tyramide (Servicebio, Wuhan, China), each for 1 h at room temperature. Finally, DAPI (Servicebio, Wuhan, China) was applied to stain the nuclei.

Paraffin-embedded tissue sections were stained using the Alcian Blue PAS staining kit (Solarbio, Beijing, China) to evaluate the number of goblet cells. Stained sections were then examined under a microscope (NanoZoomer S60, Hamamatsu, Japan) for detailed analysis.

Fresh tissue blocks were collected promptly within 1–3 min specifically for TEM analysis. The tissues were fixed overnight at 4 °C using fresh TEM fixative, composed of 2.5% glutaraldehyde in PBS. Following initial fixation, tissues were further fixed with 1% osmium tetroxide in PBS at room temperature for 2 h. The samples were then dehydrated in a graded series of ethanol, infiltrated, and embedded in resin. They were subsequently polymerized in a 65 °C oven. Ultrathin sections (60–80 nm) were cut and placed on copper grids for examination. Micrographs were obtained using a Jeol JEM 1400 Plus electron microscope (Jeol, USA).

Septic mice were fasted for 8 h prior to the assay. FITC-dextran (3–5 kDa, Sigma, MA, USA) was then administered intragastrically at a dose of 60 mg/Kg. Blood samples were collected 3 h postadministration. The fluorescence intensity of the samples was measured using a microplate reader equipped with a multiwavelength measurement system, setting the excitation at 488 nm and emission at 520 nm.

Approximately 25 mg of each sample was weighed and placed into an EP tube, to which 500 μl of extraction solution was added. The extraction solution consisted of methanol, acetonitrile, and water in a 2:2:1 ratio, enriched with an isotopically-labeled internal standard mixture. The mixture was thoroughly vortexed and centrifuged, and the resulting supernatant was transferred to a fresh glass vial for subsequent analysis. LC–MS/MS analyzes were conducted using a Vanquish UHPLC system (Thermo Fisher Scientific) equipped with a UPLC BEH amide column (2.1 × 100 mm, 1.7 μm). The system was coupled to a Q Exactive HFX mass spectrometer (Orbitrap MS, Thermo). Metabolites were annotated using an in-house MS2 database (BiotreeDB) to facilitate accurate metabolite identification.

Sample preparation: Samples were initially centrifuged at 12,000 rpm for 15 min at 4 °C. After centrifugation, 20 µl of H₂O was added to 40 µl of the supernatant. This mixture was homogenized and sonicated in three cycles, then left to settle at −20 °C overnight. Following this, the samples were again centrifuged at 12,000 rpm and 4 °C for 15 min.

Chemical derivatization: 80 µl of the resulting supernatant was transferred into an EP tube. The samples underwent derivatization by adding 40 µl of 100 mmol/L sodium carbonate solution and 40 µl of 2% benzoyl chloride in acetonitrile, followed by incubation for 30 min. After derivatization, 10 µl of an internal standard was added, and the samples were centrifuged at 12,000  rpm for 15 min at 4 °C.

UHPLC-MS/MS analysis: 40 µl of the supernatant was diluted with 20 µl H₂O and transferred to an auto-sampler vial. The UHPLC separation was performed using an ACQUITY Premier (Waters) system equipped with a Waters ACQUITY UPLC HSS T3 column (100 × 2.1 mm, 1.8 µm). The mobile phase consisted of 0.1% formic acid and 1 mM/L ammonium acetate in water (Phase A) and acetonitrile (Phase B). The column temperature was maintained at 40 °C, with the autosampler temperature set at 10 °C. The injection volume was 1 µl. Detection was carried out using a SCIEX Triple Quad™ 6500+ mass spectrometer.

Colonic tissues were collected for RNA extraction using TRIzol reagent. Subsequent library construction and Illumina sequencing were performed to evaluate gene expression. These processes were conducted by Novogene Co., Ltd. (Beijing, China).

All data supporting the findings of this study are available within the article or from the corresponding author upon reasonable request. The 16S rRNA gene sequencing raw data have been deposited in the NCBI database (accession numbers: PRJNA1394304 and PRJNA1394310).