Incorporation of polylactic acid microplastics into the carbon cycle as a carbon source to remodel the endogenous metabolism of the gut

Significance

The growing presence of polylactic acid microplastics (PLA-MPs) in aqueous and alimentary contexts has led to increased oral exposure. Understanding primary biological processes of PLA-MPs in the gut is crucial for assessing their potential impact on the host health. We clarified the degradation source of PLA-MPs in the gut and revealed the incorporation of PLA-MPs into the metabolic flux of gut microbiota and epithelial cells as a carbon source in vivo using the ¹³C isotope tracing technique, thereby entailing a distinct biological fate of biodegradable plastics in mammals.

Abstract

Biodegradable polylactic acid (PLA) plastics have been praised as an effective solution to the global pollution caused by petroleum-based plastics, and their widespread use in food packaging and disposable tableware has resulted in increased oral exposure to PLA microplastics (PLA-MPs). Despite their eco-friendly and biodegradable reputation, the in vivo behaviors of PLA-MPs concerning fermentation, carbon cycle, and adverse effects remain unknown. Here, we showed that gut microbiota from the colon can effectively degrade the PLA-MPs by secreting esterase FrsA, whereas esterase FrsA-producing bacteria were identified to dominate this behavior in male C57BL/6 mice. Using isotope tracing and multiomics techniques, we uncovered that ¹³C-labeled PLA-MPs were incorporated into the carbon cycle of gut microbiota as a carbon source. Meanwhile, these degraded PLA-MPs fragments entered the succinate pathway of the tricarboxylic acid cycle within gut epithelial cells. These processes altered the metabolic phenotype of the gut, resulting in the decreased linear short-chain fatty acids that are primary energy sources of the gut epithelium. Furthermore, we found that exposure of PLA-MPs significantly reduced the appetite and body weight of mice. Our findings present an overall process of biodegradable plastics within hosts, with the focus on the entire double carbon cycle of PLA-MPs in the gut, which offers indispensable insights into the potential impact of exposure to PLA-MPs.

The escalating global contamination of microplastics (MPs) presents serious ecological and economic problems (1, 2), which poses a top risk to the human health (3, 4). Despite efforts like the “plastic limit order” (referring to the regulations aimed at reducing plastic pollution) (5, 6), conventional petroleum-based plastics such as polyethylene (PE), polypropylene, polystyrene, polyvinyl chloride (PVC), and PE terephthalate (PET) continue to be widely used, and their sustainability has always been a challenge (7, 8). Estimates suggest that adults consume up to approximately 5 g of MPs per week from tap water, bottled drinks, food, and containers (9, 10). In contrast, infants ingest even higher amounts, with up to 11.2 million particles of MPs per week from feeding bottles (11). These plastics have been found in the lung, spleen, kidney, and placenta, as well as in breast milk and blood of humans (4, 12, 13). Epidemiological studies have demonstrated that patients with inflammatory bowel disease have a higher level of MPs in their feces compared to healthy individuals (14, 15). Given that oral ingestion is the predominant route of MP exposure (16), the gut serves as a primary site for their interaction, accumulation, and potential biological effects. A better understanding of the biological processes of MPs in living organisms, especially in the gut, is thus essential for evaluating their long-term health implications and guiding the development of safer plastic alternatives.

Currently, conventional petroleum-based plastics are the most important contributor to global plastic pollution due to their poor degradability and sustainability (17, 18). Given that, biodegradable plastics like polylactic acid (PLA), which holds the largest market share at around 50%, have been put forward as a major solution to the enduring plastic problem, particularly in the field of food packaging and disposable tableware (19–21). Compared to petroleum-based plastics, PLA plastics are more brittle and easier to form MPs (22). Biodegradable plastics degrade optimally in composting facilities with temperatures around 60 °C and 50 to 70% humidity, achieving over 90% decomposition in 180 d. However, in natural environments, complete degradation may take several years due to low temperatures, dryness, or microbial inactivity (23). Specifically, PLA degrades at a slower rate compared to other biodegradable plastics in aquatic environments (24), which means that if not properly managed, MPs produced by the slow and incomplete degradation of PLA will persist in the aquatic environments. Although biodegradable plastics are often assumed to be safe and sustainable (25, 26), multiple studies have shown that the decomposition of biodegradable MPs can be detrimental (14, 15, 27). For instance, enteritis was caused by the breakdown of PLA plastics into smaller oligomer nanoparticles due to lipase (15). In particular, a recent study highlighted the role of the gut microbiota in controlling the catabolic actions of exogenous carbon nanoparticles, which influence the function of intestinal stem cells through their transformed products (28). Consequently, the dynamic processes taking place in the gut, particularly those governed by the gut microbiota, which are responsible for digesting exogenous dietary fibers (28–30), determine the ultimate fate and safety of invading plastics, yet the detailed in vivo degradation process and possible transformation of PLA-MPs in the gut remain unknown (31).

In this study, we generated PLA-MPs by grinding commercial PLA plastic particles with an electrical mill to simulate natural plastic crushing processes, and used the multiomics approach (proteomics, metagenomics, and metabolomics) and stable isotope ¹³C tracing to reveal the in vivo behaviors of PLA-MPs in male C57BL/6 mice. We demonstrated that gut microbiota effectively degraded PLA-MPs through secreting esterase FrsA, with Helicobacter muridarum and Barnesiella viscericola dominating this process. Importantly, we uncovered the entire double carbon cycle of ¹³C-PLA-MPs as a carbon source in gut microbiota and epithelial cells. The succinate pathway of the tricarboxylic acid (TCA) cycle played an important role in the transformation of PLA-MPs in epithelial cells. Notably, gut microbiota converted L-lactate (L-LA) derived from PLA-MPs into D-lactate (D-LA) due to their unique chiral fermentation, thus resulting in elevated serum urate levels. The metabolic phenotype of gut microbiota and epithelial cells was altered by these interconnected actions. Moreover, exposure of PLA-MPs induced the damage of the gut barrier and reduced food intake of mice. These findings made a significant advancement by introducing the complete life cycle of exogenous PLA-MPs in hosts, encompassing degradation, fermentation, transformation, and adverse outcomes. As biodegradable plastics continue to gain traction, such knowledge becomes imperative for assessing and addressing the biosafety concerns posed by the growing presence of biodegradable MPs, represented by PLA-MPs, in aqueous and alimentary contexts.

Results

Gut Microbiota Degrade PLA-MPs In Vivo and In Vitro.

We prepared PLA-MPs by pulverizing commercial PLA plastic particles using an electrical mill to mimic natural plastic crushing processes. The size (~50 μm), surface charge (−21.7 mV), particle number (41,600/mg), endotoxin content (<0.002 EU/mg), chirality composition (97.93% L-LA and 2.07% D-LA), element composition (oxygen/carbon: 72.23%/27.77%), infrared spectrum, molecular weight (Mw, 31,400), and crystallinity (54.01%) of these PLA-MPs are presented in SI Appendix, Fig. S1, demonstrating no apparent impurities in the prepared samples.

We first investigated the transit of Cy5-labeled PLA-MPs through the mouse gastrointestinal tract by oral gavage at a single dose of 200 mg/kg (Fig. 1 A, Upper). Individuals may consume up to 5 g of MPs per week (10), equivalent to an oral dose of ~147 mg/kg/d in mice (based on body surface area conversion; 20 g mouse, 60 kg human) (32), and even more for infants (11). Therefore, this single dose likely underestimates human MP exposure (33). Over a 48-h period, PLA-MPs translocated swiftly from the stomach to the small intestine within 0.5 h and then resided in the colon for up to 36 h (Fig. 1B), indicating they resided in the colon for over 2/3 of the time. The PLA-MPs’ size significantly decreased from an initial 50 ± 24 μm to 17 ± 9 μm in fecal samples at 48 h (Fig. 1 C–E), suggesting effective degradation of PLA-MPs by intestinal components.

Fig. 1.

Gut microbiota degraded the PLA-MPs. (A) Experimental design for PLA-MPs degradation by colon microbiota in vivo (Upper) and in vitro (Lower). In vivo: Mice received a single oral dose of 200 mg/kg Cy5-labeled PLA-MPs. PLA-MPs-Cy5 localization in the gastrointestinal tract was visualized at 0 to 48 h, and particle morphology and size in fecal samples were analyzed at 48 h. In vitro: Colon microbiota were incubated with PLA-MPs (1 g) for 21 d (OD₆₀₀ = 1; 1:30 w/v PLA-MPs/medium), with degradation measured every 3 d. (B) The transport process of Cy5 labeled PLA-MPs in mouse gastrointestinal tract after oral gavage. (C) The in vivo degradation of PLA-MPs labeled with Cy5. (Scale bar: 50 μm.) (D and E) Size distribution of PLA-MPs before and after oral gavage in mice (n = 200). Particle sizes in feces were measured by counting 200 particles. (F) Mass loss of PLA-MPs in vitro due to degradation by gut microbiota (n = 3, three replicates). (G–J) The changes of Mw (G), crystallinity (H), particle morphology (I, Scale bar: 50 μm), and surface erosions (J, Scale bar: 1 μm) of PLA-MPs during the degradation process mediated by gut microbiota in vitro (n = 3, three replicates). The culture medium served as the control. Data are expressed as mean ± SD. Two-way ANOVA was used for statistical analysis in Fig. 1 F–H, with comparisons to the control group. ns (not significant), *P < 0.05, **P < 0.01, and ***P < 0.001.

We next determined the specific components responsible for the degradation of PLA-MPs by incubating them separately in simulated gastric fluid, simulated small intestinal fluid, or simulated colonic fluid (SCF) for 3 d (Particles/simulated fluid was 1:30 w/v). Meanwhile, we supplemented each simulated fluid with corresponding mouse gastrointestinal contents (extracted from the stomach, small intestine, and colon) to ensure the presence of gut microbiota and various digestive enzymes (SI Appendix, Table S1) (34). The results revealed that under anaerobic conditions, the degradation of PLA-MPs in the SCF containing colon contents was significantly higher than that in the other groups (SI Appendix, Fig. S2 A and B). Given that colonic microbiota exclusively thrives under anaerobic conditions (28, 29), we hypothesized that microbiota from the colon are the primary factor responsible for PLA-MPs degradation. This hypothesis was confirmed when gut microbiota extracted from the colon prominently broke down the PLA-MPs after 3 d of incubation under anaerobic conditions (SI Appendix, Fig. S2 C and D). Thus, these results clearly demonstrate the dominant capability of colonic microbiota to degrade PLA-MPs, ruling out the significant contribution of other factors, such as gastrointestinal digestive enzymes and bile salts.

We further studied the degradation process of PLA-MPs by gut microbiota in vitro (Fig. 1 A, Lower). After 21-d incubation (OD₆₀₀ = 1; 1:30 w/v PLA-MPs/medium, refreshed medium every 3 d), approximately 15% of the PLA-MPs underwent degradation by gut microbiota, whereas under the natural hydrolysis condition (control group), only about 3% degradation of PLA-MPs occurred (Fig. 1F). PLA-MPs degradation generated fragmented molecules (SI Appendix, Fig. S3A), reduced molecular weight from 31,400 to 13,000 (Fig. 1G; the culture medium served as the control), and increased polydispersity index from 1.87 to 28.84 (SI Appendix, Fig. S3B). Notably, the crystallinity of PLA-MPs increased from 54.01 to 64.20% (Fig. 1H), suggesting that preferential degradation of amorphous regions with defects. Especially, the degradation rate of PLA-MPs was highest (up to 5%) in the initial stage (0 to 3 d), slowed during the mid-term stage (3 to 15 d), and increased again in the later stage (15 to 21 d). Moreover, Microscopy showed that gut microbiota caused surface erosion on the particles, making them fragile and leading to recurrent breakdown (Fig. 1 I and J). Overall, these findings demonstrate that gut microbiota efficiently degrade PLA-MPs both in vivo and in vitro.

Key Microbial Enzymes and Gut Microbes Involved in the Degradation of PLA-MPs.

We next investigated how gut microbes degrade PLA-MPs through a 21-d cocultivation experiment of PLA-MPs. We collected proteins absorbed on the surface of PLA-MPs (Fig. 2 A, Upper) on day 3 (initial stage), day 12 (intermediate stage), and day 21 (later stage). Proteomic analysis of the collected samples revealed significant changes in protein expression profiles over time (SI Appendix, Fig. S4A). On day 12, compared to day 3, there were 84 upregulated and 94 downregulated proteins. Interestingly, between day 12 and day 21, there were only three differentially expressed proteins (SI Appendix, Fig. S4B), indicating the stabilization of the dominant proteins responsible for degrading PLA-MPs on their surface by day 12 (SI Appendix, Fig. S4C). After screening (SI Appendix, Fig. S5), we identified 19 significantly enriched functional proteins on PLA-MPs (Fig. 2B and SI Appendix, Table S2). Among them, esterase FrsA (Q0TL77) and L-LA dehydrogenase (B8DSV5) levels were 7.4- and 7.3-fold higher, respectively, on day 12 compared to day 3 (Fig. 2 C and D). This suggests that gut microbiota may degrade PLA-MPs by secreting esterase FrsA and then converting them using L-LA dehydrogenase.

Fig. 2.

Key microbial enzymes and gut microbes involved in the degradation of PLA-MPs. (A) Experimental design for screening the target proteins (Upper) and gut microbiota (Lower) involved in PLA-MP degradation. Colon microbiota were incubated with PLA-MPs (1 g) for 21 d (OD₆₀₀ = 1; 1:30 w/v PLA-MPs/medium). Surface proteins on fragments were identified on days 3, 12, and 21 via proteomics. Microbiota were analyzed for specific degrading microbes by metagenomics. (B) Z-score heat map of 19 target functional proteins, presented as UniProt accession (n = 3, three replicates). (C and D) Relative abundance of esterase FrsA (C) and L-LA dehydrogenase (D) capable of degrading ester bonds and metabolizing L-LA (n = 3, three replicates). (E) The predicted structure of esterase FrsA (Q0TL77) and docked L-LA dimer. (F and G) The recombinant esterase FrsA (Q0TL77) expressed from Escherichia coli degraded PLA-MPs to small fragments (F) and L-LA monomers (G) (n = 3, three replicates). (H) Predominant gut bacteria involved in the PLA-MPs degradation and metabolic transformation (n = 6, biologically independent mice; P < 0.05). Values are standardized Z-scores based on relative abundances. Only bacteria with an occurrence frequency > 0.5 (present in at least three samples in each group) are displayed. (I) Network diagram of interactions between 17 target gut bacteria and proteins (n = 3, three replicates; P < 0.05, Spearman correlation coefficient; bacteria in lowercase as in Fig. 2H; line thickness indicates the degree of significant difference; shape size indicates the number of connected nodes). Data are expressed as mean ± SD. Statistical analysis was performed using the two-tailed t test for Fig. 2 C and D, and two-way ANOVA for Fig. 2G, with comparisons to the control group. ns (not significant), *P < 0.05, **P < 0.01, and ***P < 0.001.

The existence of gut microbiota–secreted esterase capable of hydrolyzing PLA-MPs is a remarkable finding, as previous studies have mainly focused on bacterial enzymes degrading conventional petroleum-based plastics (35–37). To further characterize esterase FrsA, we predicted its structure using AlphaFold2 and confirmed its identity via the UniProt database. Homology modeling with Vibrio vulnificus FrsA (PDB: 4I4C; 43% sequence similarity) as a template revealed a structure with parallel β-sheets and α-helices, forming a typical α/β hydrolase active center (38) (SI Appendix, Fig. S6 A and B). Molecular docking of the L-LA dimer with this structure using Glide showed a stable ligand conformation matching the original position in V. vulnificus FrsA (Fig. 2E). Molecular dynamics simulations demonstrated the stability of the enzyme–substrate complex over 500 ns (SI Appendix, Fig. S6C), with most residues stable except the protein’s C/N termini (SI Appendix, Fig. S6D). These results suggest that the L-LA dimer can stably bind to and be degraded by the esterase FrsA’s active center. Experimentally, coincubation of esterase FrsA (Q0TL77) from the recombinant Escherichia coli O6:K15:H31 with PLA-MPs resulted in small fragments of PLA-MPs (Fig. 2F) and an increased level of L-LA monomers in the medium (Fig. 2G), further confirming the enzymatic ability of esterase FrsA to degrade PLA-MPs. Typically, the prevalent methods for managing plastic waste, such as thermochemical depolymerization, require high temperatures (200 °C to 500 °C) and substantial energy inputs (hundreds of kW·h per ton of plastic) (23). Furthermore, many biodegradable plastics require specific composting conditions (precise temperature, humidity, and high concentration of microbes) to undergo full degradation, and the current infrastructure is insufficient to handle the vast volume of plastic waste generated. Our finding of an effective enzymatic decomposition system involving esterase FrsA for PLA-MPs provides a promising alternative approach to managing MPs. This enzymatic degradation presents a viable and environmentally friendly option that is distinct from traditional thermochemical depolymerization methods (7).

We further investigated the gut microbiota responsible for PLA-MP degradation via metagenomic analysis (Fig. 2 A, Lower). As a carbon source, PLA-MPs significantly altered gut microbiota diversity, leading to distinct microbial clustering (SI Appendix, Fig. S7). At the genus level, five genera (Bacteroides, Clostridium, Duncaniella, Roseburia, and Chitinophaga) increased in abundance, while three (Prevotella, Paramuribaculum, and Dorea) decreased (SI Appendix, Fig. S8A). At the species level, 17 gram-negative bacteria changed significantly, with eight increasing (Fig. 2H and SI Appendix, Fig. S8 B–E). Notably, the most prominent increase was observed with H. muridarum (1,040-fold higher than control), Helicobacter japonicus (1,037-fold), B. viscericola (49-fold), and Muribaculaceae bacterium Isolate-007 (10-fold). These findings indicate that PLA-MPs promote the growth of these bacteria species as a carbon source. In contrast, the top three bacteria with the most significant decrease were Prevotella sp. MGM1 (0.007-fold of control), Bacteroides bouchesdurhonensis (0.02-fold), and Helicobacter sp. MIT 03-1614 (0.07-fold), suggesting a selective inhibition effect of PLA-MPs on these bacteria. This observation suggests that gram-negative bacteria may be more susceptible to PLA-MPs than gram-positive bacteria, likely due to their thinner cell walls.

Due to the consistency between in vitro and in vivo experiments in terms of gut microbiota composition (SI Appendix, Fig. S9), especially on day 12, we subsequently conducted a correlation analysis between the target bacterial species identified in vivo and the two key enzymes to determine which specific gut bacteria were primarily involved in the degradation and metabolic conversion of PLA-MPs. The results show that H. muridarum and B. viscericola were the two gut bacteria most significantly positively correlated with esterase FrsA (Fig. 2I). We then verified the degradation of PLA-MPs by two bacterial strains. Both strains showed strong degradation capabilities, with B. viscericola outperforming H. muridarum (SI Appendix, Fig. S10). Meanwhile, esterase FrsA was significantly enriched on the fragment surfaces. These results validated the dominant roles of H. muridarum and B. viscericola in the degradation of PLA-MPs. Additionally, four Bacteroidetes species (Parabacteroides sp. D26 and Muribaculaceae bacterium Isolate-007/110/037) were identified as having positive interactions with L-LA dehydrogenase, which likely play significant roles in the metabolic transformation of PLA-MPs. These interactions between gut bacteria and key enzymes further highlight the complex process by which gut microbiota contribute to the degradation of PLA-MPs (SI Appendix, Table S3).

Gut Microbiota and Epithelial Cells Incorporate PLA-MPs into the Carbon Cycle.

Given that gut microbiota are the primary exposure targets of orally ingested MPs (39, 40), we delved into the specific metabolic pathways of PLA-MPs within the gut microbiota. After the mice were treated with PLA-MPs (200 mg/kg) by oral gavage for 21 d, we performed whole genome sequencing to analyze colonic microbiota samples (Fig. 3 A, Upper). The functional gene analysis identified 8,631 microbial genes, showing significant alterations in Kyoto Encyclopedia of Genes and Genomes (KEGG) Orthology (KO), with 369 functional genes exhibiting differences (SI Appendix, Fig. S11 A–C and Table S4). These genes were mapped to three levels of KEGG pathways and subjected to Linear Discriminant Analysis (LDA) Effect Size (LEfSe) analysis. As shown in Fig. 3B, the innermost circle (L1) represents broad functional categories (SI Appendix, Fig. S11D), the middle circle (L2) represents specific pathways belonging to L1 categories, and the outermost circle (L3) represents detailed pathways within L2. The results revealed that over half of the differential KEGG pathways (LDA > 2) were related to metabolic processes, especially amino acid (e.g., tyrosine and phenylalanine) and lipid metabolism. This indicates that PLA-MPs, when degraded by gut microbiota, may act as a carbon source to affect microbial metabolism at the genetic level.

Fig. 3.

Gut microbiota and epithelial cells incorporated PLA-MPs into the carbon cycle as a carbon source. (A) Experimental design for PLA-MPs’ effects on gut microbiota functional genes (Upper) and metabolic flux of ¹³C-PLA-MPs in the gut (Lower). Mice were given PLA-MPs (200 mg/kg/d) orally for 21 d. Colon microbiota were sequenced for functional gene analysis. Another group received ¹³C-PLA-MPs (200 mg/kg/d) orally for 21 d. Colon microbiota, gut epithelial cells, and luminal supernatants were collected to detect ¹³C-labeled metabolites using stable isotope tracing. (B) LEfSe analysis of the KEGG pathway at the L2 and L3 level (LDA > 2) with significantly different functional genes (KO) in the gut microbiota of mice treated with PLA-MPs by oral gavage (200 mg/kg/d) for 21 d (n = 6, biologically independent mice). The innermost circle (L1) represents broad functional categories, the middle circle (L2) represents specific pathways belonging to L1 categories, and the outermost circle (L3) represents detailed pathways within L2. (C) The identified ¹³C labeled metabolites derived from ¹³C-PLA-MPs transformation in the gut microbiota and epithelial cells, respectively (n = 3, biologically independent mice).

We further investigated whether gut microbiota can exploit the PLA-MPs fragments as a carbon source to synthesize endogenous metabolites using the advanced stable isotope ¹³C tracing technique [A powerful technique for tracking metabolic flux of exogenous substances (41, 42)] (Fig. 3 A, Lower). To this end, we synthesized ¹³C-PLA-MPs using ¹³C-labeled L-LA monomer (with 99% ¹³C), and these particles displayed similar properties to pristine PLA-MPs without obvious impurities and residues (SI Appendix, Fig. S12). After oral gavage (200 mg/kg/d) for 21 d, we found that gut microbiota fermented ¹³C-labeled PLA-MPs into multiple endogenous molecules through two major metabolic pathways (Fig. 3 C, Upper). The first pathway involves the fermentation of ¹³C-PLA-MPs into the final product tyrosine via the metabolic chain of ¹³C-PLA-MPs—¹³C-lactate—¹³C-phenylalanine— ¹³C-cinnamate—¹³C-tyrosine. The increased L-LA dehydrogenase (Fig. 2D) was essential for the metabolic transformation of ¹³C-L-LA into downstream products like ¹³C-phenylalanine. These functional molecules, especially the significantly increased tyrosine (SI Appendix, Fig. S13A), can influence the immune system of the gut (43). The second pathway of ¹³C-PLA-MPs involved in the production of central ¹³C-aspartate mediated by L-LA dehydrogenase, which branched into four metabolic chains: 1) ¹³C-aspartate—¹³C-pantothenate; 2) ¹³C-aspartate—¹³C-threonine—¹³C-glycine; 3) ¹³C-aspartate—¹³C-xanthine—¹³C-hypoxanthine or transforming into ¹³C-urate; 4) ¹³C-aspartate—¹³C-lysine—¹³C-biotin/¹³C-isovalerate. Of note, the content of ¹³C-biotin, which can only be synthesized by gut microbiota in mammals (44), in ¹³C-PLA-MPs-treated mice was 28-fold higher than in the control (SI Appendix, Fig. S13B). This suggests that biotin is a major product of microbial anaerobic fermentation of PLA-MPs in the gut. In this process, ¹³C-PLA-MPs served as the preferable carbon source of gut microbiota to produce the biotin, surpassing the role of conventional dietary fiber (45). Notably, the generation of ¹³C-urate raises concerns about the potential adverse effects of ¹³C-PLA-MPs on the kidneys and joints of hosts (46).

Like gut microbiota, gut epithelial cells can also incorporate the PLA-MPs fragments into their carbon cycle, synthesizing multiple amino acids and nucleotide components through two key metabolic pathways (Fig. 3 C, Lower). The first pathway is involved in the synthesis of amino acids and antioxidants in gut epithelial cells. These cells converted ¹³C-PLA-MPs into the metabolic chain of ¹³C-lactate—¹³C-glycine/¹³C-glutamate/¹³C-alanine, which was further converted into final products ¹³C-serine, ¹³C-proline, ¹³C-ornithine, and ¹³C-arginine. Since the production of arginine was especially increased when the epithelial barrier was compromised (47), these amino acids modulate the intestinal mucosal immunity (48). In particular, ¹³C-glycine and ¹³C-glutamate contributed to the synthesis of ¹³C-glutathione, an antioxidant and free radical scavenger that transforms harmful toxins into harmless substances (49). The second pathway involved the transformation of ¹³C-PLA-MPs through the succinate pathway in the TCA cycle, comprising two main branches: 1) ¹³C-succinate— ¹³C-malate—¹³C-oxaloacetate—¹³C-aspartate—¹³C-xanthine— ¹³C-hypoxanthine— ¹³C-adenosine. These molecules contributed to producing the final products of ¹³C-AMP and ¹³C-GMP, which are essential for the nucleotide and energy metabolism of gut epithelial cells (50); 2) The second branch was responsible for the synthesis of nucleic acid precursors including ¹³C-uridine and ¹³C-thymidine through the TCA cycle.

Notably, we found substantial increase in short-chain fatty acids (SCFAs), ¹³C-isovalerate (SI Appendix, Fig. S13C), and the presence of ¹³C-L-tryptophan, secreted by gut microbiota into the supernatant of lumen and further utilized by gut epithelial cells to generate energy and ¹³C-5-hydroxy-L-tryptophan (a key substrate for serotonin production) (51), respectively. This underscores an interactive network between gut microbiota and epithelial cells mediated by intermediary metabolic products. These combined findings reveal that both gut microbiota and epithelial cells can harness PLA-MPs as an unconventional carbon source for their respective fermentation and metabolic processes.

The Incorporation of PLA-MPs as a Carbon Source Induces the Adaptive Alteration of Gut Microbial Metabolism.

The metabolic flux has shown ¹³C-labeled metabolites from microbial fermentation of ¹³C-PLA-MPs. To assess the ultimate metabolic outcomes, we conducted targeted and nontargeted metabolomics on gut microbiota from mice administered PLA-MPs (oral gavage, 200 mg/kg/d) for 21 d. Targeted analysis of SCFAs revealed a decrease in linear SCFAs compared to the control: acetate (−20%), propionate (−20%), and butyrate (−4%), while branched SCFAs isovalerate and isobutyrate increased by 37% and 7% (Fig. 4 A and B). Normally, acetate typically serves as a precursor for butyrate production via acetyl-CoA (28), and both butyrate and propionate are key energy sources for the gut epithelium, modulating metabolism and reinforcing gut barrier function (30, 52, 53). Reduced levels of these SCFAs may deplete gut epithelial energy, potentially disrupting the gut barrier.

Fig. 4.

PLA-MPs enabled the metabolic alteration of gut microbiota. (A) The changes of SCFAs in gut microbiota of mice treated with PLA-MPs (200 mg/kg/d) by oral gavage for 21 d (n = 6, biologically independent mice), analyzed by gas chromatography-mass spectrometry (GC-MS). (B) The acetate levels of PLA-MPs-treated mice and control (***P < 0.001, two-tailed t test). Data are expressed as mean ± SD. (C) The nontargeted metabolome shows distinct clusters of gut microbial metabolites in PLA-MPs-treated mice and control using the principal coordinates analysis (PCoA) based on Bray–Curtis dissimilarity (n = 8, biologically independent mice; 95% confidence ellipses) by liquid chromatography-mass spectrometry (LC–MS). (D) The 20 most affected KEGG pathways in gut microbiota of PLA-MPs-treated mice. Circle size indicates the number of metabolites in each pathway (n = 8, biologically independent mice). (E) The metabolic profile of top 10 KEGG metabolic pathways in the gut microbiota of PLA-MPs-treated mice. Significant metabolites are highlighted (orange). Cyan nodes (a-j) correspond to the top 10 pathways in Fig. 4D. (F) Changes in the gut microbiota metabolic profile of mice after a 21-d recovery period using the PCoA based on Bray–Curtis dissimilarity (n = 8, biologically independent mice; 95% confidence ellipses). Mice treated with PLA-MPs for 21 d were maintained under the same condition as the control group during the recovery period, without PLA-MPs treatment. (G) Statistics of significantly altered gut microbiota metabolites before and after the recovery experiment.

Nontargeted metabolomics showed a distinct metabolic phenotype in PLA-MPs-treated mice compared to controls (Fig. 4C), with 125 metabolites increased and 50 decreased significantly (SI Appendix, Fig. S14A). Among them, the 20 metabolites with the highest relative abundance (SI Appendix, Fig. S14B) and the 20 metabolites contributing the most to grouping were highlighted (SI Appendix, Fig. S14C). For instance, L-lysine (2.5-fold), L-phenylalanine (1.7-fold), L-threonine (1.8-fold), and hypoxanthine (1.4-fold) were elevated in treated mice. Mapping significantly altered metabolites to KEGG pathways revealed that 20 pathways, especially amino acids, SCFAs, and energy metabolism, were the most affected by PLA-MPs (Fig. 4D), consistent with the ¹³C metabolic flux results in Fig. 3 C, Upper. Furthermore, the network diagram showed clear links between significant metabolites (orange nodes) and the 10 most impacted pathways (cyan nodes) (Fig. 4E). Notably, L-lysine was directly linked to the lysine degradation pathway (Node h), highlighting the profound influence of incorporating PLA-MPs as a carbon source on the metabolic alteration of the gut microbiota.

We then explored which bacteria dominated the fermentation of PLA-MPs as their carbon source into specific metabolites. Correlation analysis between 17 significantly changed gut bacteria (Fig. 2H) and 38 representative differential metabolites revealed that acetate-producing bacteria (e.g., Butyricimonas sp. Marseille P4593, Bacteroides bouchesdurhonensis, Bacteroides plebeius, and Muribaculaceae bacterium Isolate-001/080/039; P < 0.001) were predominant in reducing acetate level (SI Appendix, Fig. S15). Conversely, increased abundances of H. muridarum, H. japonicus, and Muribaculaceae bacterium Isolate-007/037 (P < 0.001) were positively correlated with L-lysine production. This underscores a clear link between specific gut bacteria and PLA-MPs metabolism, revealing a competitive relationship between two gut microbial communities exploiting PLA-MPs to generate metabolites. Taken together, these results suggest that the metabolic phenotype of gut microbiota can be adaptively altered through the microbial fermentation of PLA-MPs.

To assess the persistence of PLA-MP-induced metabolic alterations, mice exposed to PLA-MPs for 21 d underwent an equivalent recovery period under control diet conditions. Despite this recovery phase, gut microbiota metabolism failed to fully revert to the control (Fig. 4F). Although total significantly altered metabolites decreased, 42 metabolites (10 upregulated, 32 downregulated) remained persistently dysregulated (Fig. 4G and SI Appendix, Fig. S16A), predominantly affecting fatty acid, nucleic acid, and amino acid pathways (SI Appendix, Fig. S16B). These findings reveal that PLA-MP-driven metabolic alteration exhibit profound and persistent effects, with core biochemical networks resisting restoration even after exposure cessation.

PLA-MPs Alter the Metabolic Profile of the Gut Epithelium and Induce Potential Health Hazards.

We then investigated whether PLA-MPs incorporation alters gut epithelium metabolism. Nontarget metabolome analysis showed distinct metabolic clustering in gut epithelial cells of PLA-MPs-treated mice compared to the control (Fig. 5A), with 75 downregulated and 73 upregulated metabolites (SI Appendix, Fig. S17A). We identified 20 metabolites with the highest relative abundance (SI Appendix, Fig. S17B) and with the greatest contribution to grouping (SI Appendix, Fig. S17C). The highest abundance was observed in succinate (27-fold), while carbohydrates (e.g., sucrose, glucose) and their downstream products (e.g., 6-phosphogluconate, dAMP, dCMP) contributed most to grouping. The TCA cycle was an essential pathway to the conversion of carbohydrates into these downstream metabolites which were closely related to carbon metabolism, amino acid and nucleic acid synthesis, and energy metabolism in the gut epithelium (Fig. 5B and SI Appendix, Fig. S17D). These results indicate that PLA-MPs can serve as a carbon source, leading to the metabolic alterations in gut epithelial cells.

Fig. 5.

PLA-MPs altered the metabolic profile of gut epithelial cells and induced health hazards. (A) Nontargeted metabolome identified distinct gut epithelial metabolite clusters in PLA-MPs-treated and control mice using PCoA based on Bray–Curtis dissimilarity (n = 8, biologically independent mice; 95% confidence ellipses) by LC–MS. (B) The 20 most affected KEGG metabolic pathways in gut epithelial cells of PLA-MPs-treated mice. Circle size indicates the number of metabolites in each pathway (n = 8, biologically independent mice). (C) Urate levels in gut microbiota, gut epithelial cells, blood, and urine of mice treated with PLA-MPs (200 mg/kg/d) by oral gavage for 21 d (n = 6, biologically independent mice). (D) D-LA levels in gut microbiota, gut epithelial cells, blood, and urine of mice treated with PLA-MPs (200 mg/kg/d) for 21 d (n = 6, biologically independent mice). (E) PLA-MPs-induced damage to the intestinal villi of mice. (F and G) Food intake (F) and body weight (G) of mice during PLA-MPs treatment (n = 6, biologically independent mice). Data are expressed as mean ± SD. Statistical analysis is performed using one-way ANOVA for Fig. 5 C and D, and two-way ANOVA for Fig. 5 F and G, with comparisons against the control group. ns (not significant), *P < 0.05, **P < 0.01, and ***P < 0.001.

It is noteworthy that gut microbiota metabolized ¹³C-PLA-MPs into ¹³C-urate (SI Appendix, Fig. S13H), but minimal urate accumulated in the gut epithelium. A similar phenomenon was observed in both the gut microbiota and epithelium of PLA-MP-treated mice (Fig. 5C). Interestingly, urate levels in serum and urine of PLA-MP-treated mice increased 2.1-fold and 1.4-fold, respectively. Typically, serum urate is excreted via urine to maintain homeostasis (54). However, the significant rise suggests a mechanism retaining urate in serum. One possible mechanism involves D-LA, which inhibits urate excretion by acting as a substrate for renal organic anion transporters that mediate urate transport in exchange for monovalent organic anions like D-LA, thus impeding urate clearance (55). Our findings demonstrated significant increases in D-LA levels in the gut microbiota, serum, and urine by 3.2-fold, 1.3-fold, and 2.7-fold, respectively (Fig. 5D). Notably, while most D-LA in the serum is excreted through urine, the gut microbiota, as the sole endogenous source of D-LA in mammals (56), has the unique chiral fermentation ability to continuously supply D-LA by utilizing PLA-MPs as a carbon source. Collectively, these results indicate that gut microbiota can not only synthesize urate using PLA-MPs but also convert L-LA to D-LA through microbial chiral fermentation. It promotes the reabsorption of urate from urine, thus leading to an increased urate concentration in the serum.

We further investigated the biological consequence of orally administered PLA-MPs to the mice. Following oral gavage of 4-Chloro-7-nitrobenzofurazan-labeled PLA-MPs (200 mg/kg) for 21 d, no significant accumulation of PLA-MPs was observed in the remote organs including the liver, kidney, heart, lung, spleen, and brain of the mice. Nevertheless, these particles were retained in the colon (SI Appendix, Fig. S18). This suggests that 1) gut microbiota/epithelium-mediated the degradation and metabolism of PLA-MPs in the gut prevent their translocation to the distal tissues and 2) gut microbiota and epithelium are major factors in determining the final fate of PLA-MPs in biological hosts.

In addition, while the gut microbiota and epithelium were capable of metabolizing PLA-MPs into biological molecules, exposure of PLA-MPs still led to the damage of the gut epithelial barrier (Fig. 5E and SI Appendix, Fig. S19), and significantly reduced appetite and body weight (Fig. 5 F and G). However, it remains unclear whether these adverse effects arise from the physical influence of MPs (pristine and degraded fragments, etc.), the metabolites generated through PLA-MPs transformation, or a combination of both, which warrants further investigation. These findings collectively suggest that despite being considered environmentally friendly (57), PLA-MPs may have adverse effects on gut health.

Discussion

Although several studies have investigated the degradation of PLA-MPs in the gut and their interaction with gut microbiota (31, 58), they primarily focused on the phenomenon observation of PLA-MPs degradation and alteration of gut microbiota, and the underlying action mechanisms remain unclear. Our study found that PLA-MPs are primarily degraded by colon microbiota. Specifically, the esterase FrsA, produced by colon H. muridarum and B. viscericola, which are also present in humans (59), degraded PLA-MPs by breaking the ester bond of backbone. Interestingly, approximately 5% of PLA-MPs were degraded into L-LA monomers by colon microbiota in 72 h, with a microbiota/PLA-MPs incubation ratio of 1/50. This contrasts with a recent study showing that high concentrations of commercial gastric and small intestinal lipases degraded 5 to 10% of PLA-MPs into oligomers at an enzyme/PLA-MPs ratio of 1/5 (15). Our study found that orally administered PLA-MPs quickly moved from the stomach to the small intestine within 0.5 h and stayed in the colon for over 2/3 of the time (Fig. 1B). This suggests that PLA-MPs are mainly degraded in the colon due to its longer retention time and higher microbial esterase concentration compared to the stomach and small intestine. Our findings provide compelling evidence of colon microbiota-governed catabolism of PLA-MPs (31), thus enhancing our understanding of biodegradable plastic breakdown in the gut. Unlike PLA-MPs, petroleum-based plastics are typically inert (60), but recent studies show that PET, PE, and PVC can be degraded by environmental and insect bacterial strains (35, 36, 61). This implies that both anaerobic and aerobic bacteria may break down both petroleum-based and biodegradable plastics.

An important finding of this study is that we identified the entire carbon cycle of PLA-MPs as a carbon source in gut microbiota using isotope tracing (Fig. 3C). The process involved L-LA dehydrogenase converting PLA-MPs from lactate to downstream metabolites via a four-step transformation (¹³C-PLA-MPs—¹³C-lactate— ¹³C-phenylalanine—¹³C-cinnamate—¹³C-tyrosine). High levels of isovalerate and urate [a hazardous molecule (62)] were also generated. This suggests that PLA-MPs might impact human health through their transformed metabolites. Of note, 17 gram-negative bacteria (from Bacteroidetes and Proteobacteria phyla) were sensitive to PLA-MPs during degradation and fermentation process, likely due to their thinner cell walls compared to gram-positive bacteria (63). Additionally, we unraveled the whole carbon cycle of PLA-MPs in the gut epithelium. PLA-MPs were utilized as a major carbon source to generate amino acids (¹³C-alanine, ¹³C-serine, ¹³C-proline, ¹³C-ornithine, and ¹³C-arginine) and nucleotide components (¹³C-thymidine, ¹³C-AMP, and ¹³C-GMP) through the succinate pathway in the TCA cycle, revealing a distinct biological fate of PLA-MPs in the gut.

The final consequence of PLA-MPs metabolism as a carbon source is the alteration of metabolic phenotypes in gut microbiota. Particularly, PLA-MPs fermentation led to a decrease in straight-chain SCFAs but an increase in branched-chain SCFAs in gut microbiota. Since straight-chain SCFAs are the primary energy source for gut epithelial cells (30), their reduction poses a risk of energy depletion (64). In addition, gut microbiota convert L-LA from PLA-MPs into D-LA, which inhibits urate excretion and promotes its reabsorption into the blood, potentially associated with hyperuricemia (55). Moreover, elevated D-LA levels are also associated with diabetes (65) and memory impairment (66). Given these facts, PLA-MPs altered the metabolic pathways associated with life processes (such as carbon metabolism and energy metabolism) as a carbon source.

Although it is unclear how gut microbes distinguish PLA-MPs, one plausible reason is that frequent consumption of plastics through food and drinks has domesticated gut microbiota to recognize and break down these plastics. This domestication causes PLA-MPs to promote the growth of detrimental bacteria like H. muridarum (67), which then dominate and replace microbes that ferment conventional dietary fibers. These findings imply that, with the frequent consumption of MPs, gut microbiota may develop into a microbial population that relies on plastics instead of traditional diets. This shift in the gut microbiota structure may be harmful to the gut homeostasis, as the potentially hazardous molecule (like D-LA and urate) from microbial PLA-MPs catabolism may accumulate and pose health risks to hosts (46, 68). Notably, the PLA-induced metabolic alterations in gut microbiota exhibit persistence even after 21-d recovery, particularly within core pathways of fatty acid, nucleic acid, and amino acid metabolism. Complete restoration likely requires durations exceeding initial exposure periods, though precise recovery thresholds remain undetermined. Notably, given the inherent challenges in sustaining long-term avoidance of PLA-containing products, intermittent re-exposure may repeatedly reshape microbial metabolism. Prolonged cycles of exposure-recovery could ultimately establish a PLA-MPs-driven metabolic phenotype.

In addition, considering the limitation of the sample number used in this study (n ≤ 8), the universality of the biological fate and effects of PLA-MPs in the gut warrants further investigations. Overall, our study has unveiled the degradation and entire double carbon cycle of PLA-MPs in the gut microbiota and epithelium (Fig. 6), thus providing valuable insight into the ultimate fate and potential influence of biodegradable plastics. This highlights that the biodegradability of plastics does not necessarily equate to safety (22).

Fig. 6.

Proposed working model to elucidate the biological processes and carbon cycle of PLA-MPs in the gut. When PLA-MPs enter the gut lumen, gut microbiota secrete the esterase FrsA to degrade PLA-MPs, which then served as a carbon source to enter the carbon cycle in gut microbiota and epithelial cells, altering the gut metabolic phenotype. Ultimately, exposure of PLA-MPs disrupts the gut barrier and leads to significant reductions in appetite and body weight of mice.

Materials and Methods

We provide detailed information on materials and methods in SI Appendix, including PLA-MPs characterization, experimental animals, degradation of PLA-MPs in vitro, target protein and microbiota screening, metabolic flux analysis of ¹³C-labeled PLA-MPs, consistency analysis of gut microbiota in vitro and in vivo, computational simulations, esterase FrsA degradation verification, validation of PLA-MPs degradation by B. viscericola and H. muridarum, gut metabolic profiling, recovery of gut microbial metabolism, biosafety assessment, and statistical analysis.

Supplementary Material

Appendix 01 (PDF)

Data, Materials, and Software Availability

The raw data, metadata, animal experiment protocols, and other relevant data are accessible at Science Data Bank (https://cstr.cn/31253.11.sciencedb.20656) (69). All other data are included in the article and/or SI Appendix.