Fluorescent polystyrene (PS) particles were purchased from the Baseline Chromatography Technology Development Center (Tianjin, China). Polyethylene (PE), polycarbonate (PC), polypropylene (PP) and polyethylene terephthalate (PET) particles were obtained from BASF (China) Company Limited. Particle size was characterized by Topsizer Plus Laser Particle Size Analyzer with SCF-126B micro-circulating sampler (Omec, China).

Fermented foods were diluted with sterile PBS and then plated on GAM (Gifu Anaerobic Medium), BBL™ and MRS (de Man, Rogosa, and Sharpe) agar plates. The plates were incubated under anaerobic condition at 37°C for 48 h. Single colonies were streaked into corresponding fresh medium and incubated under anaerobic condition for another 24 h. After incubation, the bacteria were harvested by centrifugation, washed twice with sterile PBS solution, and resuspended to 1 × 10⁹ CFU/mL in PBS solution.

20 μL of 1 × 10⁹ CFU/mL bacterial suspension was added into a 96-well plate in which each well contained 180 μL of 0.16 mg/mL 0.1 μm PS fluorescent particle working solution. The 96-well plate was placed on shaker and incubated for 4 h in the dark at 37°C, 800 rpm. The solution was then centrifuged at 2000 rpm for 10 min, and the fluorescence intensity of the supernatant was measured using a Tecan Microplate Reader (excitation 488 nm, emission 518 nm). The adsorption ratio (AR) was calculated as follows: AR (%) = (A1−A2)/A1 × 100%, where A1 and A2 represent the fluorescence intensity of the supernatant incubated with the PBS control, and candidate bacteria, respectively. Samples were prepared in triplicates. The taxonomy of selected strains was determined by 16S rRNA Sanger sequencing.

Selected Lactobacillus strains were inoculated into MRS medium and anaerobically cultured at 37°C for 24 h. After culturing, the bacterial solution was centrifuged at 4000 rpm for 10 min. The supernatant was discarded, and the pellet was washed twice by adding 450 μL of sterile PBS buffer. Next, PBS was added to resuspend the bacterial solution and adjust its concentration to 1 × 10⁹ CFU/mL. Subsequently, 100 μL of the bacterial suspension was transferred into a 1.5-mL EP tube, and 900 μL of sterile PBS or PS fluorescent particle working solution (0.16 mg/mL, particle size 0.1 μm) was mixed. The mixture was then shaken in the dark for 4 h at 37°C and 800 rpm. At the end of the incubation period, the solution was examined for bacteria-PS particle aggregates. Finally, the incubation solution was centrifuged at 2,000 rpm for 10 min to observe the color of the bacterial precipitate.

The overnight culture of isolated strains was collected by centrifugation at 4,000 rpm for 10 min. The precipitate was washed with PBS solution and resuspended with MRS medium at pH 2.5 or MRS medium containing 0.1% bile salt. Bacterial suspension was then incubated under anaerobic condition at 37°C for 3 h. Lactobacillus rhamnosus GG was used as a control. The survival rate was calculated as follows: Survival rate (%) = C1/C0 × 100%, where C0 and C1 represent the number of viable cells in the culture before and after the incubation, respectively. Samples were prepared in triplicates.

The antioxidant activity was determined by 2,2-Diphenyl-1-picrylhydrazyl (DPPH) radical-scavenging capacity according to the previous method (Li et al., 2012). Briefly, Lactobacillus was suspended in PBS solution at a concentration of 1 × 10⁹ CFU/mL. A 3 μg/mL vitamin C (Vc) solution was used as positive control and PBS as negative control. 500 μL of the bacterial suspension was added to 500 μL of 0.2 mmol/L DPPH ethanol solution. The reaction of bacterial suspension and ethanol was used as blank. Following thorough mixing, the reaction was placed on a shaker at 30°C, 800 rpm for 30 min. The absorbance of 100 μL of supernatant was measured at 517 nm after centrifugation at 4,000 rpm for 10 min. The DPPH radical scavenging activity was calculated as follows: Scavenging activity (%) = (1 – (A1 – A0)/Ac) × 100%, where A1 represents the absorbance value in the experiment group, A0 represents the absorbance value in the no-DPPH blank, and Ac represents the absorbance value in the PBS-treated negative control. Samples were prepared in triplicates.

MP particles were resuspended in a PBS solution containing 0.1% Tween-80. 1 × 10⁸ CFU of bacteria and 0.144 mg MP were mixed in a solution, placed on a shaker and incubated for 4 h at 37°C, 800 rpm. After incubation, the precipitate was harvested by centrifugation at 4000 rpm for 10 min and subjected to SEM imaging (TESCAN MIRA LMS, Czechia).

Six-week-old male SPF C57 mice were purchased from Gempharmatech Co., Ltd. (Nanjing, China). Mouse care and all experimental procedures were approved by the Animal Ethics Committee of HuaYuan Shidai Biotech Co., Ltd. (Beijing, China) according to institutional animal ethics guidelines (Permission No. HYSD2022-10). All animals were housed under conditions with a room temperature of 22°C ± 2°C and humidity of 50% ± 10%.

Fourty six-week-old male SPF C57 mice were equally divided into 4 groups and were administered with 1 × 10⁹ CFU of probiotics or equal volume of saline daily by oral gavage for 7 days. Following the final gavage, mice were fasted for 16 h, then given a single oral gavage with 100 μL of 10 mg/mL 5 μm PS solution. After 20 min of PS gavage, mouse intestine was dissected from the cardia to the cecum. The fluorescence position of PS particles was photographed using IVIS® Lumina III In Vivo Imaging System (PerkinElmer, USA). Images were processed by ImageJ software. Excretion rate (%) = L1/L0 was calculated, where L1 represents the movement length of fluorescent PS particles in the intestine, and L0 represents the total length of the intestine from pylorus to cecum.

Six-week-old male SPF C57 mice were purchased from BesTest BioTech Co., Ltd. (Zhuhai, China) and acclimatized for 1 week. Mouse care and all experimental procedures were approved by the Animal Ethics Committee of TOP Biotech Co., Ltd. (Shenzhen, China) according to institutional animal ethics guidelines (Permission No. TOPGM-IACUC-2023-0057). All animals were housed under conditions with a room temperature of 22°C ± 2°C and humidity of 50% ± 10%.

Fifty six-week-old male SPF C57 mice were equally divided into five groups and were treated with 100 μL of 10 mg/mL 5 μm PS solution by intragastric administration for 7 days to establish PS exposure model (Jin et al., 2021; Zolotova et al., 2022; Wang et al., 2023). Three groups were orally gavaged with 1 × 10⁹ CFU probiotics in the morning and exposed to fluorescent PS particles in the afternoon. In the control group, mice were given an equal volume of saline in the morning and exposed to PS particles in the afternoon. In the blank group, mice were given saline twice every day. Following the final PS particles gavage, mice were fasted for 16 h. Mice were anesthetized and sacrificed. Blood, ileum and cecum samples were collected for subsequent analysis.

The amount of fluorescent PS particles in the intestine and feces was determined using flow cytometry referring to previous studies (Nielsen et al., 2014; Jin et al., 2021; Liang et al., 2021). Feces and freshly collected ileum and cecum samples were weighed and homogenized in 400 μL of digestion solution (1 g/L proteinase K, 5 g/L SDS, 23 g/L Na₂HPO₄, and 4.6 g/L NaH₂PO₄). After incubation at 37°C for 12 h, the homogenate was diluted with PBS buffer, and then filtered through a 100 μm cell strainer.

The resulting samples were subjected to flow cytometry analysis using NovoCyte 2100YB Flow Cytometers (Agilent, United States). The minimum threshold of FITC-H was set at 2.0 × 10⁶ and the minimum threshold of FSC-H was set at 60000 to exclude background noise in samples. Collected data were analyzed with FlowJo (v10.8.1) software and the amount of PS particles was determined using a standard curve.

The intestinal tissues were harvested as described above. Concentrations of IL-10, IL-6, TNF-α and IL-1β in the intestinal tissues were determined by respective ELISA kits (Meimian, China) according to the manufacturer’s protocol.

Fecal samples taken before the final gavage from mice, as described in Materials and Methods 2.7 were weighed and stored in methanol containing 5 mg/kg sodium acetate-D4 (Sigma Aldrich). The samples were homogenized with 20 min ultrasonication. After the esterification process, the SCFA concentration was analyzed with gas chromatography coupled to mass spectrometry (GC–MS).

Statistical differences between two groups were determined using two-tailed Student’s t-test. Differences between more than two groups were tested using one-way analysis of variance (ANOVA). Statistical analysis was performed using GraphPad Prism v8.0. p-value <0.05 was considered statistically significant.

784 probiotic strains from fermented foods were isolated and served as candidate bacteria (see Methods). We implemented a high-throughput screening method with fluorescein isothiocyanate (FITC)-labeled PS particles to screen for bacterial strains with high MP adsorption capacity (Figure 1A). The PS fluorescence intensity in solution decreases upon bacteria adsorbing and precipitating these particles after incubation and centrifugation. Using the relative fluorescence intensity decrease, i.e., adsorption ratio (AR), we found that 87 strains had an AR greater than 60%, with the highest AR being 80.5% (Figure 1B).

Three probiotic strains with high or low AR values were selected for further analysis. Two strains designated as Lacticaseibacillus paracasei DT66 and Lactiplantibacillus plantarum DT88 have AR values of 71.4 and 79.8%, respectively, while the other strain, Lactiplantibacillus plantarum DT22 has an AR value of 6.2% (Figure 1C). Lacticaseibacillus paracasei DT66 and Lactiplantibacillus plantarum DT88 co-aggregated and formed flocculent precipitate with PS particles after incubation, while there was minimal precipitate formed in the bacterial solution without PS particles (Figure 1D). After centrifugation, white precipitates were observed in bacterial solution without co-incubation with PS particles. The precipitate of Lacticaseibacillus paracasei DT66 and Lactiplantibacillus plantarum DT88 after co-incubation with PS particles turned green while Lactiplantibacillus plantarum DT22 remained white, demonstrated the different PS adsorbing capacity of selected Lactobacillus strains (Supplementary Figure S1).

To identify candidate probiotics with high application potential, we further validated for their tolerance to bile salt and gastric acid, as well as their antioxidant potentials. All three Lactobacillus strains demonstrated good tolerance for bile salts and gastric acids (Figure 1E), and showed good antioxidant potentials (Figure 1F). To confirm the screening results for MP adsorption, we subjected the three strains to SEM visualization after co-incubation of bacteria and PS particles (Figure 2). Indeed, Lacticaseibacillus paracasei DT66 and Lactiplantibacillus plantarum DT88 were fully covered by 0.1 μm PS particles, while much fewer PS particles adhered to the surface of Lactiplantibacillus plantarum DT22 (Figure 2A). Similarly, we observed extensive aggregation of Lacticaseibacillus paracasei DT66 and Lactiplantibacillus plantarum DT88 cells, but less Lactiplantibacillus plantarum DT22 cells, aggregated on the surface of 5 μm PS particles (Figure 2B).

In addition to PS particles, we further assessed these bacteria’s adsorption ability to other MP particles commonly found in the environment, including PE, PC, PP, and PET (Supplementary Figure S2). Aggregation of bacterial cells on the surface of different MP particles was observed for both Lacticaseibacillus paracasei DT66 and Lactiplantibacillus plantarum DT88, but much less for Lactiplantibacillus plantarum DT22 (Figures 2C–F).

Taken together, our high-throughput screening and multi-step confirmation resulted in two probiotic strains, Lacticaseibacillus paracasei DT66 and Lactiplantibacillus plantarum DT88, that exhibited high adsorption capacity on a variety of common MPs.

We hypothesize that the larger size of bacteria-MP aggregates may facilitate MP excretion from the gastrointestinal (GI) tract. We next investigated whether the administration of Lacticaseibacillus paracasei DT66 and Lactiplantibacillus plantarum DT88 could promote MP excretion in a mouse model. Monitoring the movement of gavaged material from the pylorus to the cecum can be used to evaluate short-time intestinal excretion rate in vivo (Zhou et al., 2022). Accordingly, we administered mice with each Lactobacillus strain or saline control for seven consecutive days, followed by one shot oral gavage of fluorescent PS particles to examine the immediate MP excretion rate (Figure 3A). Compared to the saline control and Lactiplantibacillus plantarum DT22 groups, mice treated with Lacticaseibacillus paracasei DT66 and Lactiplantibacillus plantarum DT88 showed the fluorescent PS moving further down the intestinal tract (Figure 3B). Quantification results showed that the average excretion rate of PS was 41.0% in the control group, while the number increased to 55.9 and 55.2% in the Lacticaseibacillus paracasei DT66 and Lactiplantibacillus plantarum DT88 groups, respectively (Figure 3C). This result indicated that these probiotics’ high MP adsorption capacity correlates with high MP excretion effect in mice.

Besides MP aggregation, we explored whether GI motility may also play a role in MP removal, as probiotics can produce SCFAs that regulate GI motility (Martin-Gallausiaux et al., 2021). The acetic acid and propionic acid levels in the feces did not show significant changes in all three probiotic treated groups compared to the saline-treated control after 7 days of probiotic administration, while the butyric acid level increased to different extents (Supplementary Figure S3). These results suggested that GI motility may also play a role in the acceleration of PS excretion.

With candidate Lactobacillus strains showing promising MP adsorption and excretion capacities, we next examined if MP accumulation and the subsequent immune response in vivo can be ameliorated by probiotic treatment. To this end, we orally gavaged mice with 1 × 10⁹ CFU bacterial strains or saline in the morning and 1 mg of fluorescent PS in the afternoon for seven consecutive days (Figure 4A).

Pilot time-series excretion experiment indicated that >99% of the PS in the feces were excreted at 16 h post the last PS gavage (Supplementary Figure S4). To determine the residual MP in the intestine, we collected the ileum and cecum samples at 16 h after the last PS gavage. For the ileum tissues, the PS particle amount was found reduced by 61.9 and 66.8% in mice treated with Lacticaseibacillus paracasei DT66 and Lactiplantibacillus plantarum DT88 compared to saline group, respectively (Figure 4B). A similar decrease by these two strains was also observed for the cecum samples (Figure 4B). The low-adsorbing strain Lactiplantibacillus plantarum DT22 did not change the residual PS levels in both ileum and cecum. Taken together, these results showed that Lacticaseibacillus paracasei DT66 and Lactiplantibacillus plantarum DT88 could effectively reduce MP retention in the intestinal tract.

MP exposure could trigger an immune response in the intestine (Liu et al., 2022). To evaluate the impact of PS removal on gut inflammation, we collected intestine tissue samples to examine the levels of different inflammatory biomarkers in mice, including the anti-inflammatory cytokine Interleukin 10 (IL-10) and pro-inflammatory cytokines IL-6, TNF-α and IL-1β. We also tested serum cytokine to assess systemic inflammation, in addition to intestinal cytokine levels which represent localized immune effects. We observed decreased IL-10 levels in the serum and ileum in the PS-treated mice compared to the blank group (Figure 4C, Supplementary Figure S5).

Administration of Lacticaseibacillus paracasei DT66 significantly alleviated the IL-10 reduction only in the serum, and treatment with Lactiplantibacillus plantarum DT88 significantly prevented the IL-10 reduction in both serum and ileum compared to the control group (Figure 4C, Supplementary Figure S5). With Regard to the pro-inflammatory cytokines, we found higher levels of IL-6, TNF-α and IL-1β in the ileum after PS exposure, and treatment with Lactiplantibacillus plantarum DT88 restored these cytokines to the blank group levels (Figure 4C). These results suggested that Lactiplantibacillus plantarum DT88 is an effective probiotic strain to ameliorate MP-induced inflammatory responses.