The probiotic strains, L. plantarum CCFM1189 and L. reuteri CCFM1190, were obtained from the Culture Collection of Food Microorganisms at Jiangnan University (Wuxi, China). CCFM1189 was incubated under anaerobic conditions in de Man-Rogosa-Sharpe (MRS) liquid medium at 37°C for a duration of 16–24 h. CCFM1190 was cultivated in MRS liquid medium supplemented with 0.5 g·L⁻¹ L-cysteine hydrochloride within an anaerobic chamber at 37°C for a period of 24–48 h. After strain amplification, centrifuge at 4°C and 8,000 × g for 10 min, discard the supernatant, wash twice with 0.9% sterile saline, resuspend the bacterial slurry in 5 mL of sucrose solution, count, and store in a − 80°C freezer. Before gavage, centrifuge at 4°C and 8,000 × g for 10 min, dilute the bacterial suspension to 5 × 10⁹ CFU·mL⁻¹ with 0.9% sterile saline based on the counting results, and set aside.

The dosage regimen for the ABX solution in this study was adapted from Scott et al. (17): Ampicillin (9 mg·kg⁻¹), metronidazole (9 mg·kg⁻¹), neomycin sulfate (9 mg·kg⁻¹), and vancomycin (4.5 mg·kg⁻¹) were administered via intragastric route at a volume of 0.2 mL per day, commencing from the adaptation phase and continuing until the 30th day, which precedes the initiation of the stimulation protocol.

The IA was solubilized using a Dimethyl sulfoxide (DMSO) solution (18). The composition of the DMSO solution included DMSO, polyethylene glycol 300 (PEG300), tween-80, and saline solution in the ratio of 2:8:1:9, respectively. The IA was dissolved in this DMSO solution. The low-dose IA group (IA1) received 0.2 mL of the low-dose IA solution (4.8 mg·mL⁻¹), whereas the high-dose IA group (IA2) was administered 0.2 mL of the high-dose IA solution (48 mg·mL⁻¹).

The mice used in this experiment were specific pathogen free (SPF) 4-week-old female BALB/c mice, which were purchased from Charles River Laboratory Animal Technology Co., Ltd. (Zhejiang, China), and raised in the barrier facilities of the Experimental Animal Center at Jiangnan University. The initial body weights of the mice ranged from 13 to 16 g. A one-week acclimatization period was observed prior to the commencement of the experiment. The environmental conditions for feeding, the dietary regimen, the establishment of the food allergy model, and the protocols for execution and sampling were all in accordance with the methodologies detailed in our previous study (16), the schematic diagram is shown in Figure 1.

The mice were randomly allocated into 10 distinct groups: control, model, ABX, ABX + CCFM1189, ABX + CCFM1190, CCFM1189, CCFM1190, DMSO, IA1, and IA2. Each group comprised 6–7 mice. Mice in the ABX + CCFM1189, ABX + CCFM1190, CCFM1189, and CCFM1190 groups received a daily gavage of 0.2 mL (5 × 10⁹ CFU·mL⁻¹) of probiotic suspension from day 0 to day 43. Conversely, mice in the control and model groups were administered an equivalent volume of sterile saline solution. For the ABX-treated groups (ABX, ABX + CCFM1189, ABX + CCFM1190), each mouse received a daily gavage of 0.2 mL of ABX solution from the adaptation period until the 30th day (the day preceding the initiation of stimulation), with the ABX being dissolved in saline solution. The DMSO group received intragastric administration of 0.2 mL of DMSO solution from day 0 to day 43. For the IA-treated groups, the IA1 group received 0.2 mL of low-dose IA solution (4.8 mg·mL⁻¹) per mouse from day 0 to day 43, while the IA2 group received 0.2 mL of high-dose IA solution (48 mg·mL⁻¹) per mouse over the same period. The mice were weighed at the same time each week from the adaptive period, with their weights expressed as percentages (weight% = current weight / initial weight (day 0) × 100%).

This investigation was conducted in full compliance with the NIH Guidelines for the Care and Use of Laboratory Animals (NIH Publication No. 85-23 Rev. 985) and received formal approval from the Laboratory Animal Ethics Committee of Jiangnan University (Wuxi, Jiangsu, China; approval numbers JN. No. 20210915b1121201[300] and JN. No. 20211115b0581228[434]).

The calculation method of food allergy symptom score in mice detailed in our previous study (16). Feces of mice were collected for a period of time, and the diarrhea index was calculated (diarrhea index = the number of soft stools / the total amount of feces).

Serum and jejunal tissue samples were procured and subjected to analysis using mouse-specific OVA-sIgE, IL-4, IL-5, IL-13, and HIS enzyme-linked immunosorbent assay (ELISA) kits (Nanjing Senbeijia Biological Technology Co., Ltd., Nanjing, China), in accordance with the manufacturer’s guidelines. HIS, a biogenic amine present in the human body, is primarily recognized for its function as a mediator in allergic responses; however, it also serves as a crucial signaling molecule within the nervous system, gastrointestinal tract, integumentary system, and immune system. The absorbance was quantified at a wavelength of 450 nm utilizing a microplate reader (Thermo Fisher Scientific, Waltham, MA, United States).

The jejunal tissues were preserved in a 4% paraformaldehyde solution. Subsequently, the tissues were processed for immunohistochemical sectioning, encompassing steps such as repair, dehydration, transparency, waxing, embedding, sectioning, dewaxing, antigen retrieval, sealing, incubation with primary and secondary antibodies, Diaminobenzidine coloration, termination of coloration, and redyeing and dewatering. The sections were digitized and examined at 400× magnification using a digital slide scanner (3DHISTECH Ltd., Hungary). The positive staining area and total area were quantified using Image J 1.8.0 (National Institutes of Health, United States).

The CCFM1189 and CCFM1190 strains were activated using MRS liquid medium over two generations. The second generation was harvested at the logarithmic phase, centrifuged at 4000 × g for 10 min, and the supernatant was subsequently discarded. An equivalent volume of the pre-reduced isotonic potassium phosphate buffer (PPS; 10 mM MgSO₄, 21.46 mM KH₂PO₄, 18.54 mM K₂HPO_4, pH 7.0) was prepared for the purpose of cleansing bacterial cells. The cells were thoroughly resuspended in the PPS, which was subsequently removed under the same centrifugation conditions as previously described. Following two cycles, an equivalent volume of PPS supplemented with 1 mM tryptophan was added and incubated in an anaerobic workstation for 24 h. Subsequently, the mixture was centrifuged at 8000 × g for 10 min, and the supernatant was filtered through a 0.22 μm filter membrane for subsequent detection.

The collected fecal samples were subjected to lyophilization under conditions of −107°C and 0.2 mbar. The freeze-dried fecal samples were precisely weighed to 50 mg, and subsequently augmented with 800 μL of a methanol aqueous solution (methanol: water = 1: 1). The samples underwent vortex mixing for 15 s, followed by tissue crushing at a frequency of 65 Hz, each cycle lasting 30 s (repeated 5 times). The homogenized samples were incubated at 4°C overnight. Following centrifugation at 16000 × g for 10 min, 500 μL of the supernatant was aspirated and concentrated under vacuum at 45°C for 2–4 h. The supernatant was resuspended in 200 μL of a methanol solution (methanol: water = 1: 9), centrifuged at 14000 × g for 10 min, and subsequently filtered through a 0.22 μm filter membrane.

A stock solution of IA with a concentration of 1,000 ppm was prepared by diluting a methanol-aqueous solution (methanol: water = 1: 9) to yield concentrations of 50 ppm, 25 ppm, 10 ppm, 5 ppm, 1 ppm, 500 ppb, 250 ppb, 125 ppb, 100 ppb, 50 ppb, 25 ppb, 12.5 ppb, 6.25 ppb, 3.125 ppb, and 1.5625 ppb. These varying concentrations of IA were employed to construct a standard curve correlating peak area with concentration. IA was quantified using ultra-performance liquid chromatography coupled with high-resolution mass spectrometry (UPLC-HRMS; Thermo Fisher Scientific, Waltham, MA, USA; C18 UPLC Columns). The column temperature was maintained at 36.5°C, while the automatic sampler temperature was set at 4°C. Each sample was analyzed in positive ion mode with mobile phases comprising (A) acetonitrile and (B) 0.1% formic acid in aqueous solution. The injection volume was 2 μL, and the flow rate was 300 μL/min. The gradient elution program settings are detailed in Table 1.

The FastDNA Spin Kit for Feces (MP Biomedicals, CAT.NO.6570200) was employed to extract DNA from the fecal samples of mice, adhering strictly to the manufacturer’s protocol. The polymerase chain reaction (PCR) was utilized to amplify the DNA within the V3–V4 region. Post-PCR, the agarose gel was excised for purification purposes, utilizing the DNA Gel/PCR Purification Miniprep Kit (BW-DC3511, Beiwo Meditech Co., Ltd., Hangzhou, China), in strict accordance with the manufacturer’s guidelines. The PCR primers (341F and 806R), sequencing platform (Illumina NovaSeq 6,000), quality control method of the raw data, and analysis platform [QIIME 2 (19)] were referred to Lu et al. (20). Conduct a differential genus analysis utilizing the integrated LEfSe (Linear Discriminant Analysis Effect Size) utility within OmicStudio (21), accompanied by visualization, and employing the subsequent filters: Kruskal-Wallis test threshold set at 0.05; Wilcoxon test threshold set at 0.05; LDA score threshold set at 3.

The allergy symptom scores, diarrhea index, ELISA results, and IA content were subjected to statistical analysis using IBM SPSS Statistics 26 (SPSS Inc., Chicago, IL, United States). The assessment of significant differences between the groups was conducted through one-way analysis of variance (ANOVA) followed by Duncan’s multiple range test. All data are presented as mean values ± the standard error of the mean (SEM). A p-value of <0.05 was deemed indicative of a statistically significant difference between the groups, corresponding to a confidence interval exceeding 95%.

Initially, we performed exhaustive evaluations and statistical analyses on the food allergy-associated phenotypic indices across all experimental mice groups. In alignment with our prior research, CCFM1189 and CCFM1190 exhibited the outstanding capacity to alleviate weight loss, allergy scores, and diarrhea symptoms in normal food allergic mice (Figures 2A–C). Nevertheless, in contrast to these two groups, ABX treatment intensified the disease phenotypes in food-allergic mice, and the oral administration of CCFM1189 and CCFM1190 failed to significantly alleviate the symptoms in ABX-treated mice.

Subsequently, we determined the levels of serum OVA-sIgE across all mice groups. As shown in Figure 2D, administration of ABX resulted in a reduction of serum OVA-sIgE levels in mice when compared to the model group. In the case of CCFM1189 intervention, no significant difference was observed in the response to OVA-sIgE whether or not ABX was treated. Regarding the CCFM1190 group, it is exhibited a significant reduction in serum OVA-sIgE levels in mice (p < 0.01), whereas the ABX-treated CCFM1190 group did not demonstrate such a reduction. The results of jejunum HIS levels (a critical mediator in allergic responses; measured in supernatant) showed that effective bacterial strains have the potential to substantially mitigate the heightened HIS levels induced by food allergies; however, probiotic interventions in mice subjected to ABX treatment failed to diminish the HIS content within the jejunum (Figure 2E).

The levels of Th2 cytokines (IL-4, IL-5, and IL-13) in the jejunum supernatant of mice indicated that ABX treatment mitigated the inhibitory impact of the effective strains CCFM1189 and CCFM1190 on the Th2 immune response. In comparison to the model group, CCFM1189 and CCFM1190 notably diminished the IL-4 levels in the jejunum of mice, with the alleviating effect of the ABX-treated strain being less pronounced (Figure 2F). Regarding IL-5 and IL-13, CCFM1189 decreased the levels of these cytokines; however, this reduction was not statistically significant when compared to the ABX + CCFM1189 group; CCFM1190, in the absence of ABX treatment, significantly lowered the levels of IL-5 (Figure 2G) and IL-13 (Figure 2H; p < 0.01). Collectively, the ABX + CCFM1189 and ABX + CCFM1190 groups exhibited a lesser inhibitory effect on the Th2-type immune response in food allergic mice compared to the CCFM1189 and CCFM1190 groups.

Moreover, the immunohistochemical analysis of tight junction proteins ZO-1 and Occludin in the jejunum of mice was conducted, with the proportion of positively stained regions being quantified (Figures 3A,B). The results indicated that ABX treatment led to a reduction in the expression levels of ZO-1 and Occludin in the jejunum of mice. The expression levels of ZO-1 and Occludin in the ABX + CCFM1189 and ABX + CCFM1190 groups were markedly lower compared to those in the CCFM1189 and CCFM1190 groups without ABX treatment (p < 0.05). After histopathological examination of the jejunum sections in mice, it is found that the small intestinal villi of the model group mice exhibited disorganization and abscission, while the lamina propria appeared disordered and lax. The intestinal villi of mice in the CCFM1189 and CCFM1190 groups were orderly arranged, with the lamina propria being intact and closely aligned, resembling the phenotype of the control group and thereby restoring the integrity of the intestinal barrier to a certain extent. In contrast, the ABX treatment groups (ABX, ABX + CCFM1189, and ABX + CCFM1190) displayed comparable phenotypes and elevated levels of inflammatory cell infiltration, akin to the model group.

In our prior research endeavors, we have ascertained that the probiotic strains CCFM1189 and CCFM1090 significantly mitigate food allergy in ovalbumin-induced mice, with IA being identified as a crucial microbial-derived metabolite. To elucidate the mechanisms responsible for the variations in IA levels within the gastrointestinal tract under probiotic intervention, we initially utilized an in vitro fermentation methodology. By incorporating tryptophan into the culture medium, we assessed the capacity of CCFM1189 and CCFM1190 to metabolize tryptophan and yield IA (Figure 4A). The results indicated that none of these strains were capable of producing IA through tryptophan metabolism during in vitro fermentation. Subsequently, to verify whether CCFM1189 and CCFM1190 augment the IA content in mouse feces by influencing other symbiotic bacteria within the intestinal tract, the IA content of effective strain groups treated with ABX (ABX + CCFM1189, ABX + CCFM1190) was compared with those not treated with ABX (CCFM1189, CCFM1190). The results demonstrated that the IA content in the feces of ABX-treated mice was lower than that of both the control and model groups. The fecal IA content of the CCFM1189 and CCFM1190 groups without ABX treatment exhibited an increase compared to the ABX-treated groups (Figure 4B). The aforementioned findings imply that CCFM1189 and CCFM1190 could potentially mitigate food allergy symptoms by augmenting the IA concentration in the fecal matter of mice, with the elevated IA levels possibly attributable to the probiotic strains’ influence on the intestinal symbiotic microbiota of the mice.

Then, we conducted an analysis of the diversity and composition of the gut microbiota across various groups of mice. As depicted in Figure 4C, CCFM1189 and CCFM1190 exhibit the capability to reverse the reduction in microbial species diversity induced by food allergy models. Following treatment with ABX, a marked decrease in the diversity of the gut microbiota in mice was observed, and the intervention of probiotic strains failed to mitigate the dysbiosis induced by antibiotics (Figure 5A). The results pertaining to β diversity indicated that the distributions of ABX treatment groups (ABX, ABX + CCFM1189, and ABX + CCFM1190) was more concentrated, while were dispersed in the control, model, CCFM1189, and CCFM1190 groups (Figure 4D).

Moreover, an analysis was conducted to identify the differential bacterial genera within the gut microbiota of mice between the probiotic intervention group and the group receiving both ABX treatment and probiotic intervention. As shown in Figure 4E, it was observed that within the CCFM1189 group, the dominant bacterial species included Lactobacillus, Alistipes, Ruminococcaceae UCG-014, Alloprevotella, Rikenellaceae RC9 gut group, Ruminococcaceae UCG-010, Mucispirillum, Prevotellaceae UCG-001, Bifidobacterium, Butyricimonas, Turicibacter, Clostridium sensu stricto 1, and Candidaatus saccharimonas. In the CCFM1190 group and the ABX + CCFM1190 group, Lactobacillus, Alistipes, Alloprevotella, Rikenellaceae RC9 gut group, Ruminococcaceae UCG-014, Lachnospiraceae NK4A136 group, Clostridium sensu stricto 1, Turicibacter, Muribaculum, Romboutsia, Prevotellaceae UCG-001, Ruminococcaceae UCG-010, Enterorhabdus, Oscillibacter, Mucispirillum, Rumiclostridium, and Butyricimonas were found to be enriched within the CCFM1190 group. Among the bacterial genera enriched under the intervention of effective probiotic strains, Clostridium, Prevotella, Bifidobacterium, and Lachnospira have been identified as bacterial genera capable of producing tryptophan-indole derivatives (22). Consequently, we assert that the ingestion of the effective strains CCFM1189 and CCFM1190 can modulate the gut microbiota, promoting the proliferation of bacteria that produce elevated levels of IA and resulting in an increased concentration of IA within the gut.

In order to further elucidate the mechanisms and underlying material basis through which probiotics effectively mitigate food allergies, we conducted direct interventions using low-dose and high-dose IA in murine models of food allergy. The results indicated that neither dosage of IA significantly influenced the weight loss observed in the mice (Figure 5A). However, in comparison to the DMSO group, the IA intervention exhibited remarkable efficacy in ameliorating the phenotypic manifestations (Figure 5B) and the diarrhea index (Figure 5C) in food allergy mice, with the high-dose IA demonstrating a more pronounced effect. Furthermore, neither dosage of IA produced a significant reduction in serum OVA-sIgE levels in food allergy mice (Figure 5D). Regarding the HIS levels in the jejunal tissue supernatant of the mice, both low-dose and high-dose IA were effective in decreasing HIS concentrations, with the high-dose IA showing a notably greater impact (Figure 5E).

IA has the potential to inhibit the Th2-type immune response to a certain degree. In comparison to the DMSO group, low-dose IA was found to significantly reduce the levels of IL-4 in the jejunal supernatant of mice (Figure 6A). Regarding IL-5 (Figure 6B) and IL-13 (Figure 6C), high-dose IA exhibited a decreasing effect. Furthermore, IA was observed to enhance the positive expression of ZO-1 and Occludin immunohistochemical staining. In the DMSO group, the expression levels of ZO-1 (Figure 6D) and Occludin (Figure 6E) in the jejunum of mice were diminished; however, IA administration resulted in an elevation of these expression levels. Notably, there was no significant difference in the regulatory effects of IA on tight junction proteins between the two dosage levels. In conclusion, we have confirmed that IA serves as a crucial metabolite in the mitigation of food allergies and have elucidated the mechanisms of action of the probiotic strains CCFM1189 and CCFM1190.