Abstract

   Micro/nanoplastics (MNPs) are detected in human liver, and pose
   significant risks to human health. Oral exposure to MNPs derived from
   non-biodegradable plastics can induce toxicity in mouse liver.
   Similarly, nasal exposure to non-biodegradable plastics can cause
   airway dysbiosis in mice. However, the hepatotoxicity induced by
   foodborne and airborne biodegradable MNPs remains poorly understood.
   Here we show the hepatotoxic effects of biodegradable polylactic acid
   (PLA) MNPs through multi-omics analysis of various biological samples
   from mice, including gut, fecal, nasal, lung, liver, and blood samples.
   Our results show that both foodborne and airborne PLA MNPs compromise
   liver function, disrupt serum antioxidant activity, and cause liver
   pathology. Specifically, foodborne MNPs lead to gut microbial
   dysbiosis, metabolic alterations in the gut and serum, and liver
   transcriptomic changes. Airborne MNPs affect nasal and lung microbiota,
   alter lung and serum metabolites, and disrupt liver transcriptomics.
   The gut Lachnospiraceae_NK4A136_group is a potential biomarker for
   foodborne PLA MNP exposure, while nasal unclassified_Muribaculaceae and
   lung Klebsiella are potential biomarkers for airborne PLA MNP exposure.
   The relevant results suggest that foodborne PLA MNPs could affect the
   “gut microbiota-gut-liver” axis and induce hepatoxicity, while airborne
   PLA MNPs could disrupt the “airway microbiota-lung-liver” axis and
   cause hepatoxicity. These findings have implications for diagnosing PLA
   MNPs-induced hepatotoxicity and managing biodegradable materials in the
   environment. Our current study could be a starting point for
   biodegradable MNPs-induced hepatotoxicity. More research is needed to
   verify and inhibit the pathways that are crucial to MNPs-induced
   hepatotoxicity.

   Keywords: Micro/nanoplastics, Hepatotoxicity, Microbiota dysbiosis,
   Metabolic disruption, Transcriptomic dysregulation

Graphical abstract

   [33]Image 1
   [34]Open in a new tab

Highlights

     * •
       Foodborne PLA MNPs disrupt gut microbiota-gut-liver axis and cause
       hepatotoxicity.
     * •
       Airborne PLA MNPs affect airway microbiota-lung axis and induce
       hepatotoxicity.
     * •
       Foodborne and airborne PLA NPs and MPs induce distinctive
       alterations in mice.

1. Introduction

   Micro/nanoplastics (MNPs) have been found in the environment and living
   organisms [[35]1]. Polylactic acid (PLA), known for its high
   biocompatibility, is commonly utilized in food packaging materials
   [[36]2], and PLA nanoplastics (NPs) can be released from commercial
   teabags [[37]3]. PLA masks were found to release MNPs, which could
   enter the human airway [[38]4]. A person was estimated to ingest
   0.1–5 g MNPs weekly [[39]5]. Therefore, airborne and foodborne PLA MNPs
   will likely influence human health.

   NP and microplastics (MPs) could alter the gut microbiota in different
   living organisms. Fluorescent polystyrene (PS) MP exposure has been
   found to cause intestinal injury and gut microbiota dysbiosis (e.g.,
   increased Gordonia and decreased Aeromonas) in zebrafish [[40]6]. PS NP
   exposure could disrupt the gut microbiota in clams, e.g., enriched
   Vibrionaceae [[41]7]. Oral exposure to polyethylene (PE) MPs could
   enrich Staphylococcus and inhibit Parabacteroides in the gut of mice
   [[42]8]. The gut microbiota alternations induced by alternative common
   foodborne MNPs deserve further investigation.

   Airborne PS MNPs have been found to induce airway dysbiosis. Nasal
   exposure to fluorescent PS MPs could cause pulmonary inflammatory cell
   infiltration and bronchoalveolar macrophage aggregation in mice
   [[43]9]. In the mice nasally exposed to PS MNPs, the abundances of lung
   unclassified_Muribaculaceae and nasal Prevotella were augmented by PS
   NPs, while lung Roseburia and nasal Staphylococcus were enriched by PS
   MPs [[44]10]. PS NP exposure influenced the apoptosis, viability, and
   cell cycle of A549 human lung epithelial cells [[45]11]. However, the
   airway toxicity of airborne biodegradable PLA MNPs remains poorly
   understood.

   A few types of MNPs could induce or exacerbate hepatotoxicity. For
   instance, oral exposure to PS MPs was capable of causing inflammatory
   cell infiltration and hepatocyte edema in the liver of mice [[46]12].
   Polyethylene MP exposure could induce hepatocyte vacuolation and
   degeneration of nuclei in the liver of Nile tilapia [[47]13]. Foodborne
   PS MP exposure could worsen cyclophosphamide-induced hepatotoxicity in
   mice [[48]14]. Whether airborne MNPs could induce hepatotoxicity and
   foodborne biodegradable MNPs could cause hepatotoxicity is still
   unknown.

   PLA MPs were found to be toxic to some living organisms. For instance,
   oral exposure to PLA MPs caused gut damage and microbial dysbiosis in
   zebrafish [[49]15]. Oral exposure to PLA MPs could induce neurotoxic
   effects and REDOX imbalance in dragonfly larvae [[50]16]. The
   alternative toxicity of PLA MNPs needs to be further explored before
   more widespread application of PLA.

   It is hypothesized that foodborne PLA MNP exposure could induce
   alterations in gut microbiota, gut and serum metabolome and lead to
   liver transcriptomic changes and hepatoxicity, while airborne PLA MNP
   exposure could induce alterations in airway microbiota, lung and serum
   metabolome and lead to liver transcriptomic changes and liver toxicity.
   Therefore, in the current study, we aim to determine (1) whether
   foodborne biodegradable PLA MNPs could affect the “gut
   microbiota-gut-liver” axis and induce hepatotoxicity and (2) whether
   airborne biodegradable PLA MNPs could disrupt the “airway
   microbiota-lung-liver” axis and cause hepatotoxicity.

2. Materials and methods

2.1. Animal experiments

   Sixty male Institute of Cancer Research (ICR) mice (four-week-old,
   pathogen-free), PLA NP (50 nm), and MP (5 μm) were purchased and
   prepared for this study. Mice were allocated randomly to six groups,
   i.e., foodborne NP (FQ), foodborne MP (FR), foodborne control (FNC),
   airborne NP (AQ), airborne MP (AR), and airborne control (ANC) groups
   (n = 10 per group). They were kept at 22 °C under light/dark (12:12)
   cycles for seven days to adapt to the environment.

   In the foodborne MNP section, a 100-μL aliquot of sterile water with
   0.2 mg NPs, 0.2 mg MP, and 0 mg MNPs was administered orally to each
   mouse in the FQ, FR, and FNC groups. Oral gavage was conducted every
   day for six weeks before anesthetizing the mice to collect blood,
   liver, feces, and colon.

   In the airborne MNP section, a 10-μL aliquot of sterile saline with
   0.03 mg NPs, 0.03 mg MP, and 0 mg MNPs was intranasally administered
   for each mouse in the AQ, AR, and ANC groups, respectively [[51]9]. The
   nasal exposure was performed once every third day for 42 days before
   anesthetizing the mice to collect nasal tissue, lung, liver, and blood.
   The work was permitted by the Animal Care and Use Committee of Zhejiang
   University (2021-945).

2.2. Measurement of biochemical variables

   Serum extraction was performed via centrifugation. The liver function
   variables, i.e., aspartate aminotransferase (AST) and alanine
   aminotransferase (ALT), and the antioxidant biomarkers, i.e., total
   antioxidant capacity (T-AOC) and superoxide dismutase (SOD), in the
   serum of all the mice, were measured with commercial assay kits
   according to the manufacturer's protocols.

2.3. Histology analysis

   Formalin (10%) was used to fix the colon and liver of FQ, FR, and FNC
   groups and the lung and liver of AQ, AR, and ANC groups. Standard
   histological methods were used to process the fixed tissues. The
   stained samples were mounted on the microscope slides to assess
   pathological changes.

2.4. Molecular experiments and data processing

   The fecal, nasal, and lung samples were processed for DNA extraction
   using a DNA isolation kit. The 16S rDNA V3–V4 region of the extracted
   DNA was amplified by the bacterial primers (341F/785R) [[52]17].
   Polymerase chain reaction (PCR) products were purified before being
   sent to the sequencing lab for Illumina sequencing. Standard procedures
   were used to process the sequencing data, and sequences were clustered
   into different amplicon sequence variants (ASVs) [[53]18]. Taxonomy
   assignment was performed for the filtered ASVs against the Silva 138
   database [[54]19]. The details are provided in Supplementary Materials.

2.5. Microbiota composition analyses

   The main phyla and families in the gut microbiota of FQ, FR, and FNC
   groups, and in the nasal and lung microbiota of AQ, AR, and ANC groups,
   were determined. The alpha diversity indices of the gut, nasal, and
   lung microbiota were calculated. Permutational analysis of variance
   (PERMANOVA) was performed to compare the gut microbiota compositions of
   foodborne groups and to compare airborne groups for their lung and
   nasal microbiota compositions.

   Principal coordinate analysis (PCoA) and non-metric multidimensional
   scaling (nMDS) were carried out to demonstrate the gut microbiota of
   FQ, FR, and FNC groups and the nasal and lung microbiota of AQ, AR, and
   ANC groups. Similarity percentage (SIMPER) analysis was performed to
   investigate the gut microbiota similarities in foodborne groups and
   airway microbiota similarities in airborne groups. Linear discriminant
   analysis effect size (LEfSe) was used to compare foodborne groups for
   the gut microbiota and airborne groups for their airway microbiota.

2.6. Microbiota network analyses

   Co-occurrence network inference (CoNet) analysis was carried out to
   determine the microbiota networks in the gut of FQ, FR, and FNC groups,
   as well as those in the nasal tissue and lung of AQ, AR, and ANC
   groups. Network fragmentation analysis was conducted to explore the
   structural gatekeepers in the gut microbiota networks of FQ, FR, and
   FNC groups, and those in the nasal and lung microbiota networks of AQ,
   AR, and ANC groups. The details are provided in Supplementary
   Materials.

2.7. Microbial function analyses

   The gut microbial functional profiles of FQ, FR, and FNC groups, as
   well as the nasal and lung microbial functional profiles of AQ, AR, and
   ANC groups, were predicted using the PICRUSt2 pipeline. Statistical
   Analysis of Metagenomic Profiles (STAMP) was carried out to investigate
   (1) the gut microbial pathways upregulated in the FQ and FR groups, (2)
   the airway microbial pathways upregulated in the AQ and AR groups, and
   (3) the microbial pathways differed between MP and NP groups.

2.8. Metabolite extraction, profiling, and analyses

   Serum, colon, and lung were processed for untargeted metabolomic
   analysis (ten mice in each group). Principal component analysis (PCA)
   was carried out to demonstrate the metabolic profiles. Orthogonal
   partial least squares-discriminant analysis (OPLS-DA) was performed to
   explore the metabolites that differed between groups. Pathway
   enrichment analysis determined the enriched pathways in the foodborne
   and airborne MNP groups. The details are provided in Supplementary
   Materials.

2.9. Transcriptome analyses

   RNA was extracted from liver samples (ten mice in each group) and
   submitted for transcriptome sequencing on the Illumina platform.
   Principal component analysis (PCA) was used to visualize the gene
   expression profiles of different groups. Differential expression
   analysis was conducted to determine the differentially expressed genes
   (DEGs). Afterwards, DEGs were submitted for Gene Ontology (GO) and
   Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analyses. The
   details are provided in Supplementary Materials.

2.10. Pathway analyses

   Venny analysis was carried out to determine the pathways commonly
   enriched in the gut, serum, and/or liver of the FQ group and those of
   the FR group. The same strategy was performed to investigate the
   pathways commonly enhanced in the lung, serum, and/or liver of the AQ
   group and those of the AR group. The liver pathways commonly enriched
   in the FQ, FR, AQ, and AR groups were also explored.

2.11. Integrated analyses of altered microbiota, metabolome, and
transcriptome data

   The correlations between the altered components (i.e., microbes,
   metabolites, and transcripts) in the corresponding MNP groups were
   investigated using CoNet analysis and were demonstrated using
   Cytoscape. The sub-networks with only one type of component were
   filtered before the fragmentation analysis was conducted to determine
   the correlation gatekeepers in the correlation networks of the MNP
   groups. The key correlation gatekeeper in each group is defined as the
   component(s) associating with most alternative types of components. The
   details are provided in Supplementary Materials.

2.12. Statistical analysis

   Statistical analysis was performed in IBM SPSS statistics 25, and
   statistical significance was set at P < 0.05. Non-parametric and
   parametric methods were carried out as appropriate.

3. Results

3.1. MNPs affected liver function and antioxidant activity

   Both serum AST and ALT were at higher levels in the FQ and FR groups
   than in the FNC group ([55]Fig. 1a and b). Serum SOD was similar
   between the three groups ([56]Fig. 1c), while serum T-AOC was decreased
   in the FQ and FR groups ([57]Fig. 1d). In the nasal exposure section,
   serum AST and ALT were elevated in both AQ and AR groups ([58]Fig. 1e
   and f). SOD was reduced in the AQ group but was not significantly
   altered in the AR group ([59]Fig. 1g). By contrast, no significant
   difference was determined in serum T-AOC between AQ, AR, and ANC groups
   ([60]Fig. 1h).

Fig. 1.

   [61]Fig. 1
   [62]Open in a new tab

   Serum biochemical parameters and histological changes in the PLA
   MNPs-exposed groups. a–d, Serum AST (a), ALT (b), SOD (c), and T-AOC
   (d) in foodborne NP (FQ), foodborne MP (FR), and foodborne control
   (FNC) groups. e–h, Serum AST (e), ALT (f), SOD (g), and T-AOC (h) in
   airborne NP (AQ), airborne MP (AR), and airborne control (ANC) groups.
   i, Liver histology in the FQ, FR, and FNC groups. j–k, Lung (j) and
   liver (k) histology in the AQ, AR, and ANC groups. Black arrows point
   at the histological changes in tissues.

3.2. MNPs caused liver and lung pathology

   Hepatocyte swelling, spotty necrosis, cell vacuolization, and
   karyopyknosis were determined in the liver of FQ and FR groups
   ([63]Fig. 1i). In contrast, no obvious pathological change was observed
   in the colon of the foodborne MNP and control groups (Supplementary
   Materials Fig. S1). In the nasal exposure section, hemorrhage,
   inflammatory cell infiltration, and exudates were observed in the lung
   of AQ and AR groups ([64]Fig. 1j). Hepatocyte swelling, cell
   vacuolization, and karyopyknosis were determined in the liver of AQ and
   AR groups ([65]Fig. 1k).

3.3. Foodborne MNPs altered gut microbiota

3.3.1. Foodborne MNPs changed gut microbiota composition

   Firmicutes was the most abundant bacterial phylum accounting for over
   60% of gut microbiota in the FQ, FR, and FNC groups ([66]Fig. 2a). The
   bacterial family Lachnospiraceae had the greatest abundance in both FQ
   and FR groups, while Lactobacillaceae was the most abundant bacterial
   family in FNC group ([67]Fig. 2b). Alpha diversity of gut microbiota
   was similar between FQ, FR, and FNC groups ([68]Fig. 2c–e). Average
   similarity within gut microbiota was lower in the FQ (42.88%) and FR
   (34.7%) groups compared to the FNC group (52.31%).

Fig. 2.

   [69]Fig. 2
   [70]Open in a new tab

   Main components and alpha diversity indices of gut microbiota in
   foodborne PLA MNP and control groups. a–b, Major gut bacterial phyla
   (a) and families (b). c–e, Gut microbial richness (c), diversity (d),
   and evenness (e) indices. f–g, PCoA (f) and nMDS (g) plots of gut
   microbiota. h, Differential gut bacteria between FQ and FNC groups. i,
   Differential gut bacteria between FR and FNC groups. j–k, Structural
   gatekeepers in the FQ (j) and FR (k) gut microbiota networks.

   A significant difference was determined between the gut microbiota
   compositions of FQ, FR, and FNC groups based on PERMANOVA results
   (R^2 = 0.18, P < 0.001) and was visualized by PCoA and nMDS plots
   ([71]Fig. 2f and g). The gut microbiota composition was greatly altered
   in both FQ and FR groups (R^2 > 0.17, P < 0.003). LEfSe determined
   multiple gut bacteria associated with FQ and FR groups, among which
   Lachnospiraceae_NK4A136_group, Lachnospiraceae_A2, and Helicobacter
   were most associated with FQ group ([72]Fig. 2h), while
   Lachnospiraceae_NK4A136_group, Roseburia, and Helicobacter were most
   associated with FR group ([73]Fig. 2i). Eight gut bacteria were
   commonly enriched in both FQ and FR groups, among which
   Lachnospiraceae_NK4A136_group was most enriched (Supplementary
   Materials Table S1).

3.3.2. Foodborne MNPs altered gut microbiota networks

   The five gut bacterial phylotypes with the most correlations in the
   networks of the three foodborne groups were determined (Supplementary
   Materials Table S2, Fig. S2), among which a
   Lachnospiraceae_NK4A136_group phylotype had the most correlations in
   both FQ and FR gut networks. Nine phylotypes assigned to
   Lachnospiraceae_NK4A136_group, Muribaculaceae, Lachnospiraceae, and
   Oscillospiraceae were determined as gatekeepers in the FQ gut network
   ([74]Fig. 2j). Eleven phylotypes assigned to Colidextribacter,
   Roseburia, Lachnospiraceae_ASF356, Lachnospiraceae_NK4A136_group,
   Lachnospiraceae, Oscillospiraceae, and Clostridia_vadinBB60_group were
   found as structural gatekeepers in the FR gut network ([75]Fig. 2k).

3.3.3. Foodborne MNPs altered gut microbial functional pathways

   A variety of gut microbial functional pathways were upregulated in the
   FQ and FR groups, among which “ABC transporters”, “flagellar assembly”,
   and “bacterial chemotaxis” were most upregulated in both of the two
   groups (Supplementary Materials Fig. S3).

3.4. Airborne MNPs changed nasal microbiota

3.4.1. Airborne MNPs altered nasal microbiota composition

   At the phylum level, Firmicutes was most abundant in the AQ, AR, and
   ANC groups, constituting over 50% of nasal microbiota in each of the
   three groups ([76]Fig. 3a). Staphylococcaceae was the most abundant
   nasal family in the AQ, AR, and ANC groups ([77]Fig. 3b). Nasal alpha
   diversity indices were all at higher levels in both the AQ and AR
   groups than in ANC group ([78]Fig. 3c–e). The average similarity within
   nasal microbiota was slightly higher in the AQ (18.36%) and AR (22.22%)
   groups compared to the ANC group (17.36%).

Fig. 3.

   [79]Fig. 3
   [80]Open in a new tab

   Main components and alpha diversity indices of nasal microbiota in
   airborne PLA MNP and control groups. a–b, Major nasal bacterial phyla
   (a) and families (b). c–e, Nasal microbial richness (c), diversity (d),
   and evenness (e) indices. f–g, PCoA (f) and nMDS (g) plots of nasal
   microbiota. h, Differential nasal bacteria between AQ and ANC groups.
   i, Differential nasal bacteria between AR and ANC groups. j–k,
   Structural gatekeepers in the AQ (j) and AR (k) nasal microbiota
   networks.

   PERMANOVA determined the nasal microbiota composition was different
   between AQ, AR, and ANC groups (R^2 = 0.12, P < 0.001), and both AQ and
   AR groups were different from the ANC group (Both R^2 > 0.09, both
   P < 0.006). The nasal microbiota difference was visualized in PCoA and
   nMDS plots ([81]Fig. 3f and g). Multiple nasal bacteria were associated
   with AQ and AR groups ([82]Fig. 3h and i), of which
   unclassified_Muribaculaceae, Lachnospiraceae_NK4A136_group, and
   Acinetobacter were greatly upregulated in the AQ group ([83]Fig. 3h),
   while unclassified_Muribaculaceae, Lactobacillus, and Sporosarcina were
   largely upregulated in the AR group ([84]Fig. 3i). There were 20 nasal
   bacteria upregulated in both AQ and AR groups, among which
   unclassified_Muribaculaceae was the most enriched one (Supplementary
   Materials Table S3).

3.4.2. Airborne MNPs altered nasal microbiota networks

   The nasal phylotypes with the most correlations in the airborne groups
   were determined (Supplementary Materials Table S4, Fig. S4), among
   which two phylotypes assigned to Prevotella and Pelomonas had the most
   correlations in the AQ and AR nasal networks, respectively. Two
   Alistipes and Lachnoclostridium phylotypes were determined as the
   structural gatekeepers in the AQ nasal network ([85]Fig. 3j), while two
   structural gatekeepers assigned to Escherichia-Shigella and
   Sphingomonas were found in the AR nasal network ([86]Fig. 3k).

3.4.3. Airborne MNPs altered nasal microbial functional pathways

   Multiple functional pathways of nasal microbiota were upregulated in
   the AQ and AR groups, among which “flagellar assembly”, “bacterial
   chemotaxis”, and “cell cycle – caulobacter” were most upregulated in
   both the two groups (Supplementary Materials Fig. S5).

3.5. Airborne MNPs altered lung microbiota

3.5.1. Airborne MNPs changed lung microbiota composition

   At the phylum level, Firmicutes and Proteobacteria were most abundant
   in the lung microbiota of AQ and AR groups, while Firmicutes and
   Bacteroidota had the greatest abundances in the lung microbiota of ANC
   group ([87]Fig. 4a). Enterobacteriaceae, Muribaculaceae, and
   Lachnospiraceae constituted more than 30% of lung microbiota In the AQ
   and AR groups, while Muribaculaceae, Lachnospiraceae, and
   Staphylococcaceae were the three most abundant lung bacterial families
   in ANC group ([88]Fig. 4b). Lung bacterial observed species were lower
   in both AQ and AR groups compared to ANC group ([89]Fig. 4c). Reduced
   Shannon index was determined in the AR group ([90]Fig. 4d), while
   elevated Pielou index was found in the AQ group ([91]Fig. 4e). The
   average similarity of lung microbiota was decreased in the AQ and AR
   groups compared to ANC group, i.e., 8.02%, 9.81%, and 49.34%,
   respectively.

Fig. 4.

   [92]Fig. 4
   [93]Open in a new tab

   Main components and alpha diversity indices of lung microbiota in
   airborne PLA MNP and control groups. a–b, Major lung bacterial phyla
   (a) and families (b). c–e, Lung microbial richness (c), diversity (d),
   and evenness (e) indices. f–g, PCoA (f) and nMDS (g) plots of lung
   microbiota. h, Differential lung bacteria between AQ and ANC groups. i,
   Differential lung bacteria between AR and ANC groups. j, Differential
   lung microbial pathways between AQ and ANC groups. k, Differential lung
   microbial pathways between AR and ANC groups.

   PERMANOVA results indicated the lung microbiota composition was
   different between AQ, AR, and ANC groups (R^2 = 0.2, P < 0.001), which
   was visualized in PCoA and nMDS plots ([94]Fig. 4f and g). Compared
   with the ANC group, both AQ and AR groups had altered lung microbiota
   compositions (R^2 > 0.22, P < 0.001). Multiple lung bacteria were
   associated with AQ and AR groups ([95]Fig. 4h and i), and five were
   commonly enriched in both AQ and AR groups (Supplementary Materials
   Table S5). Klebsiella, Clostridium_sensu_stricto_1, and
   Escherichia-Shigella were most associated with both groups.

3.5.2. MNPs altered lung microbiota networks

   The five lung bacterial phylotypes with the most correlations in the
   three groups were determined (Supplementary Materials Table S6,
   Fig. S6), of which two phylotypes assigned to Klebsiella and
   Muribaculaceae had the most correlations in the AQ and AR lung
   networks, respectively. No structural gatekeeper was found in the AQ or
   AR lung networks.

3.5.3. Airborne MNPs altered lung microbial functional pathways

   Various lung microbial functional pathways were upregulated in the AQ
   and AR groups (Supplementary Materials Fig. S7a and b). “Flagellar
   assembly”, “porphyrin and chlorophyll metabolism”, and “phenylalanine
   metabolism” were most upregulated in the AQ group (Supplementary
   Materials Fig. S7a), while “flagellar assembly”, “bacterial
   chemotaxis”, and “phenylalanine metabolism” were most upregulated in
   the AR group (Supplementary Materials Fig. S7b).

3.6. Foodborne MNPs-induced gut and serum metabolic alterations

   Variations were found in the gut metabolic profile between FQ, FR, and
   FNC groups (Supplementary Materials Fig. S8a). There were 752 and 637
   altered gut metabolites in the FQ and FR groups, respectively
   ([96]Fig. 5a and b). Similarly, variations were determined between the
   serum metabolic profiles of FQ, FR, and FNC groups (Supplementary
   Materials Fig. S8b). A total of 832 and 753 serum metabolites were
   altered in the FQ and FR groups, respectively ([97]Fig. 5c and d).

Fig. 5.

   [98]Fig. 5
   [99]Open in a new tab

   Metabolic and transcriptomic alterations induced by foodborne PLA NPs
   and MPs. a–b, Alterations of gut metabolites in the FQ (a) and FR (b)
   groups. c–d, Alterations of serum metabolites in the FQ (c) and FR (d)
   groups. e–f, Enriched gut pathways (e) and serum pathways (f) in the FQ
   and FR groups. g–h, Altered liver transcripts in the FQ (g) and FR (h)
   groups. i–j, Enriched liver GO terms FQ (i) and FR (j) groups. k,
   Enriched liver pathways in the FQ and FR groups.

   The altered gut metabolites were mainly related to intensified pathways
   such as “steroid hormone biosynthesis”, “central carbon metabolism in
   cancer”, and “ABC transporters” in the FQ group; and “central carbon
   metabolism in cancer”, “protein digestion and absorption”, and
   “aminoacyl-tRNA biosynthesis” in the FR group ([100]Fig. 5e). A total
   of 18 gut pathways were enriched in both FQ and FR groups, among which
   “central carbon metabolism in cancer” was most upregulated
   ([101]Fig. 5e).

   Serum pathways such as “linoleic acid metabolism”, “arachidonic acid
   metabolism”, and “tryptophan metabolism” were enriched in the FQ group,
   while serum pathways such as “linoleic acid metabolism”, “drug
   metabolism – other enzymes”, and “tryptophan metabolism”
   ([102]Fig. 5f). Four serum pathways were enhanced in both FQ and FR
   groups, among which “linoleic acid metabolism” was the most enriched
   one ([103]Fig. 5f).

3.7. Foodborne MNPs-induced liver transcriptomic alterations

   Some variations were determined in the liver transcriptomic profile
   between FQ, FR, and FNC groups (Supplementary Materials Fig. S8c).
   There were 307 and 262 liver DEGs altered in the FQ and FR groups,
   respectively ([104]Fig. 5g and h). Three liver GO terms respectively
   belonging to biological process, cellular component, and molecular
   function, i.e., negative regulation of chromatin silencing,
   extracellular region, and nucleosomal DNA binding, were largely
   enriched in the FQ group ([105]Fig. 5i), while nucleosome assembly was
   most enriched ([106]Fig. 5j). Two different groups of nine liver
   pathways were enriched in the FQ and FR groups, among which “circadian
   rhythm” was most upregulated in both FQ and FR groups ([107]Fig. 5k).

3.8. Airborne MNPs-induced lung and serum metabolic alterations

   PCA showed variations in the lung metabolic profile between AQ, AR, and
   ANC groups (Supplementary Materials Fig. S9a). There were 864 and 596
   lung metabolites respectively elevated in the AQ and AR groups
   ([108]Fig. 6a and b). The serum metabolic profile differed in the AQ,
   AR, and ANC groups (Supplementary Materials Fig. S9b). There were 503
   and 664 serum metabolites altered in the AQ and AR groups, respectively
   ([109]Fig. 6c and d).

Fig. 6.

   [110]Fig. 6
   [111]Open in a new tab

   Metabolic and transcriptomic alterations induced by airborne PLA NPs
   and MPs. a–b, show the altered lung metabolites in the AQ (a) and AR
   (b) groups. c–d, Alterations of serum metabolites in the AQ (c) and AR
   (d) groups. e–f, Enriched lung pathways (e) and serum pathways (f) in
   the AQ and AR groups. g–h, show altered liver transcripts in the AQ (g)
   and AR (h) groups. i–j, show enriched liver GO terms in the AQ (i) and
   AR (j) groups. k, Enriched liver pathways in the AQ and AR groups.

   The altered lung metabolites were mainly related to pathways such as
   “steroid hormone biosynthesis”, “arachidonic acid metabolism”, and
   “purine metabolism” in the AQ group; and “arachidonic acid metabolism”,
   “steroid hormone biosynthesis”, and “choline metabolism in cancer” in
   the AR group ([112]Fig. 6e). Nine lung pathways were enriched in both
   AQ and AR groups, of which “arachidonic acid metabolism” was most
   upregulated ([113]Fig. 6e).

   Serum pathways such as “central carbon metabolism in cancer”, “alanine,
   aspartate and glutamate metabolism”, “butanoate metabolism”, and
   “protein digestion and absorption” were upregulated in the AQ group,
   while serum pathways such as “protein digestion and absorption”,
   “tryptophan metabolism”, “phenylalanine, tyrosine and tryptophan
   biosynthesis”, and “central carbon metabolism in cancer” were enriched
   in the AR group ([114]Fig. 6f). There were 11 serum pathways enriched
   in both AQ and AR groups, of which “central carbon metabolism in
   cancer” and “protein digestion and absorption” were largely enhanced
   ([115]Fig. 6f).

3.9. Airborne MNPs-induced liver transcriptomic alterations

   Great variations were found in the liver transcriptomic profile between
   AQ, AR, and ANC groups (Supplementary Materials Fig. S9c). Two groups
   of 303 and 289 liver DEGs were changed in the AQ and AR groups,
   respectively ([116]Fig. 6g and h). “Nucleosome assembly” was the most
   upregulated liver GO term in both AQ and AR groups ([117]Fig. 6i and
   j). Multiple liver pathways were upregulated in the AQ and AR groups,
   among which “alcoholism”, “systemic lupus erythematosus”, and
   “neutrophil extracellular trap formation” were largely upregulated in
   both groups ([118]Fig. 6k).

3.10. Pathway analysis

   “Arginine biosynthesis” and “arachidonic acid metabolism” were enriched
   in both the gut and serum of the FQ group (Supplementary Materials
   Fig. S10a), while “antifolate resistance” was enhanced in both gut and
   serum of the FR group (Supplementary Materials Fig. S10b). In the AQ
   group, “biosynthesis of unsaturated fatty acids” was enriched in both
   lung and liver, while “mineral absorption” was upregulated in both
   serum and liver (Supplementary Materials Fig. S10c). Similarly, in the
   AR group, the “mTOR signaling pathway” was intensified in both lung and
   serum, while the “FoxO signaling pathway” was enhanced in both lung and
   liver (Supplementary Materials Fig. S10d).

   “Circadian rhythm” was commonly enriched in the liver of FQ, FR, AQ,
   and AR groups (Supplementary Materials Fig. S10e). Three enriched DEGs,
   i.e., Arntl, Cry1, and Npas2, were commonly upregulated in “circadian
   rhythm” in the four groups (Supplementary Materials Table S7).

3.11. Correlation networks in MNP groups

   The four correlation networks composed of altered microbes,
   metabolites, and transcripts in the MNP groups were determined
   ([119]Fig. 7a–d) and the multiple corresponding correlation gatekeepers
   (Supplementary Materials Tables S8–S11). The DEG Gm1673 was the key
   correlation gatekeeper correlating with 23 components belonging to
   alternative types in the FQ group ([120]Fig. 7e). The DEGs cyp2b9 and
   klhdc7b, and serum metabolite bacterioruberin diglucoside were the key
   correlation gatekeepers in the FR group ([121]Fig. 7f). The DEG zfp804b
   and gut metabolite 11-cis-3,4-didehydroretinol were the key correlation
   gatekeepers in the AQ and AR groups, respectively ([122]Fig. 7g and h).

Fig. 7.

   [123]Fig. 7
   [124]Open in a new tab

   Correlation networks and the correlation gatekeepers in PLA
   MNPs-exposed groups. a–d, Correlation networks of altered microbes,
   metabolites, and liver transcripts in the FQ (a), FR (b), AQ (c), and
   AR (d) groups. e–h, Key correlation gatekeepers (larger-sized) and the
   correlated alternative types of components (smaller-sized) in the FQ
   (e), FR (f), AQ (g), and AR (h) groups.

3.12. Multi-omics analysis determined the dissimilarities between NP and MP
exposures

3.12.1. Dissimilarities between foodborne NP and MP groups

   There were 11 gut bacteria and 17 gut microbial pathways
   differentiating FQ and FR groups, among which Lachnospiraceae_A2 and
   “ABC transporters” could largely distinguish the FQ group from the FR
   group, while Roseburia and “oxidative phosphorylation” could greatly
   differentiate FR group from FQ group (Supplementary Materials Fig. S11a
   and b). A total of 273 gut metabolites differed between the FQ and FR
   groups, among which 174 were more elevated in the FQ group, while 99
   were at higher levels in the FR group (Supplementary Materials
   Fig. S11c). Six gut pathways were at different levels between the FQ
   and FR groups, all of which were more intensified in the FQ group
   (e.g., “glycolysis/gluconeogenesis”) (Supplementary Materials
   Fig. S11d). A total of 267 serum metabolites differed between FQ and FR
   groups, among which 175 were more elevated in the FQ group, and 92 were
   more elevated in the FR group (Supplementary Materials Fig. S11e).
   These differential serum metabolites were mainly involved in serum
   pathways such as “glycerophospholipid metabolism” and “arachidonic acid
   metabolism” more upregulated in the FQ group (Supplementary Materials
   Fig. S11f), and serum pathways such as “choline metabolism in cancer”
   more upregulated in the FR group (Supplementary Materials Fig. S11g)).
   Multiple liver DEGs and differential GO terms were determined between
   FQ and FR groups (Supplementary Materials Fig. S11h and i), and they
   were mainly related to 12 differential liver pathways, among which
   “acute myeloid leukemia” and “retinol metabolism” were greatly
   intensified in the FQ and FR groups, respectively (Supplementary
   Materials Fig. S11j and k).

3.12.2. Dissimilarities between airborne NP and MP groups

   Multiple differential nasal bacteria and related microbial pathways
   were determined between AQ and AR groups (Supplementary Materials
   Fig. S12a and b). Among them, Rikenellaceae_RC9_gut_group and “vitamin
   B6 metabolism” could largely differentiate the AQ group from the AR
   group, while Ileibacterium and “ABC transporters” could greatly
   distinguish the AR group from the AQ group (Supplementary Materials
   Fig. S12a and b). In contrast, only two differential lung bacteria,
   i.e., Hungatella and Lachnospiraceae_A2, were more enriched in the AQ
   and AR groups, respectively (Supplementary Materials Fig. S12c), and no
   lung microbial pathway differed between the two groups.

   Among the 776 lung metabolites that differed between AQ and AR groups,
   422 were more elevated in the AQ group, while 354 were at higher levels
   in the AR group (Supplementary Materials Fig. S12d). The two groups of
   lung metabolites were mainly involved in the upregulated pathways such
   as “arachidonic acid metabolism”, “steroid hormone biosynthesis”, and
   “ABC transporters” in the AQ group (Supplementary Materials Fig. S12e),
   and the upregulated pathways such as “purine metabolism” and “ABC
   transporters” in the AR group (Supplementary Materials Fig. S12f).
   Similarly, the serum metabolites differed between the AQ and AR groups;
   66 were at higher levels in the AQ group, while 135 were more enriched
   in the AR group (Supplementary Materials Fig. S12g). These differential
   serum metabolites were mainly related to the more intensified serum
   pathways, such as “ABC transporters” and “purine metabolism” in the AQ
   group (Supplementary Materials Fig. S12h), and the more intensified
   serum pathways, such as “d-amino acid metabolism”, “pentose phosphate
   pathway”, and “central carbon metabolism in cancer” in the AR group
   (Supplementary Materials Fig. S12i). Multiple DEGs and differential GO
   terms were determined between AQ and AR groups (Supplementary Materials
   Fig. S12j and k), and they mainly contributed to the more enriched
   liver pathways “hematopoietic cell lineage” and “parathyroid hormone
   synthesis, secretion and action” in the AQ and AR groups, respectively
   (Supplementary Materials Fig. S12l and m).

4. Discussion

   Eco-friendly materials have been used to decrease environmental
   pollution [[125]20]. However, some of them can be toxic to human
   beings. Although foodborne and airborne MNPs can cause abnormal health
   conditions in multiple organisms [[126]4,[127]21], their toxic effects
   are still largely underestimated [[128]22,[129]23]. There is a lack of
   research about the amount and concentrations of PLA MNPs in the human
   living environment, which needs further exploration. An individual was
   estimated to ingest up to 5 g MNPs per week [[130]5], and the MNP doses
   for the animal experiments in this study were below the amount.
   However, this amount could be largely underestimated due to inefficient
   MNP collection and quantification techniques [[131]24,[132]25]. This
   study found foodborne and airborne biodegradable PLA NPs and MPs to
   disrupt different microbiota-involved axes and induce hepatotoxicity.

   In this study, foodborne and airborne PLA NPs and MPs were found to
   increase the serum AST and ALT and affect the serum SOD and T-AOC,
   suggesting the PLA MNP exposure could affect the liver function and
   serum antioxidant activity. Higher levels of AST and ALT and lower
   levels of SOD and T-AOC were determined in the serum of mice orally
   exposed to PS MPs [[133]26]. Liver pathological changes (e.g., cell
   vacuolization and hepatocyte swelling) were found in the FQ, FR, AQ,
   and AR groups, suggesting that both foodborne and airborne PLA MNPs
   could impair the liver. Oral exposure to PS MPs could induce liver
   pathological changes such as hepatocyte edema, enlarged nuclei, and
   portal inflammation in mice [[134]27]. Lung pathological changes, i.e.,
   hemorrhage, inflammatory cell infiltration, and exudates, were
   determined in the AQ and AR groups, suggesting airborne PLA MNPs could
   impair the lung. PS MP nasal exposure was capable of causing
   inflammatory cell infiltration in the mouse liver [[135]9].

   FQ and FR groups differed from the FNC group in terms of the gut
   microbiota composition, suggesting that foodborne PLA NPs and MPs could
   alter the gut microbiota composition. PS MNP oral exposure was found to
   cause marked gut microbial dysbiosis [[136]28]. The alpha diversity of
   gut microbiota was not significantly altered in the FQ and FR groups,
   suggesting PLA MNP oral exposure did not alter the gut microbial alpha
   diversity. This is different from previous studies, which determined
   that PE and PS MNP oral exposures can alter the alpha diversity of gut
   microbiota in mice [[137]8,[138]29]. Among the gut bacteria elevated in
   the FQ and FR groups, Lachnospiraceae_NK4A136_group,
   Lachnospiraceae_A2, and Helicobacter were greatly enriched in the FQ
   group, while Lachnospiraceae_NK4A136_group, Roseburia, and Helicobacter
   were largely enriched in the FR group. Gut
   Lachnospiraceae_NK4A136_group was more abundant in the mice orally
   exposed to PS MNPs compared to the control group [[139]30]. Reduced
   Lachnospiraceae_A2 was found in the mice orally exposed to
   polychlorinated biphenyls [[140]31]. Helicobacter pylori could
   cooperate with PE MPs to cause gastric injury in mice [[141]32]. Less
   gut Roseburia was determined in the PS MNPs orally exposed mice
   compared to the control group [[142]28]. Gut
   Lachnospiraceae_NK4A136_group was most upregulated in both FQ and FR
   groups, suggesting it is a potential biomarker of PLA MNP oral
   exposure.

   Nasal microbiota composition differed between the AQ/AR and ANC groups,
   suggesting that airborne PLA NPs and MPs could cause nasal microbiota
   alterations. Airborne PS NPs and MPs were found to alter the nasal
   microbiota compositions in the exposed mice [[143]10]. The nasal alpha
   diversity indices were all greater in the AQ and AR groups compared to
   the ANC group, suggesting PLA MNP nasal exposure could alter the nasal
   microbial alpha diversity, which is different from our previous study
   determining PS MNP nasal exposure not capable of altering nasal
   microbial alpha diversity in mice [[144]10]. Among the multiple
   differential nasal bacteria between AQ and AR groups,
   unclassified_Muribaculaceae, Acinetobacter, and
   Lachnospiraceae_NK4A136_group were largely enriched in the AQ group,
   while unclassified_Muribaculaceae, Lactobacillus, and Sporosarcina were
   largely enriched in the AR group. Unclassified_Muribaculaceae was
   largely upregulated in the lung of mice nasally exposed to PS NPs
   [[145]10]. Acinetobacter was at a greater level in the intestine of
   mice orally exposed to PS MPs compared to the control group [[146]33].
   The abundance of gut Lachnospiraceae_NK4A136_group could be augmented
   in the mice orally exposed to PS NPs and MPs [[147]34]. Gut
   Lactobacillus was enriched in the foodborne PE MPs-exposed mice
   [[148]35]. Increased gut Sporosarcina was determined in the PS
   MPs-exposed marine medaka [[149]36]. Nasal unclassified_Muribaculaceae
   was most upregulated in both AQ and AR groups, suggesting it is a
   potential biomarker for PLA MNP nasal exposure.

   The AQ and AR groups altered the lung microbiota composition compared
   to the ANC group, suggesting that airborne PLA NPs and MPs could induce
   lung microbiota alterations. Nasal exposures to PS NPs and MPs were
   found to change the lung microbiota compositions in mice [[150]10]. As
   for the lung bacterial alpha diversity, reduced observed species in the
   AQ and AR groups, decreased Shannon index in the AR group, and elevated
   Pielou index in the AQ group was determined, suggesting that airborne
   PLA MNPs could alter the lung microbial diversity. Shannon index and
   observed species of lung microbiota were found at lower levels in the
   mice exposed to airborne PS MPs compared to the control group
   [[151]10]. Multiple bacteria were enriched in the lungs of AQ and AR
   groups, among which Klebsiella, Clostridium_sensu_stricto_1, and
   Escherichia-Shigella were most associated with both the two groups.
   Klebsiella pneumoniae was determined to have the ability to degrade
   submicroplastics [[152]37]. Clostridium_sensu_stricto_1 was reduced in
   the gut of PS NPs-exposed largemouth bass [[153]38]. In the human
   placenta, Escherichia-Shigella was negatively correlated with the PVC
   concentration but positively correlated with the
   polytetrafluoroethylene concentration [[154]39]. Lung Klebsiella was
   most enriched in both AQ and AR groups, suggesting it is a potential
   biomarker for PLA MNP nasal exposure.

   Multiple structural gatekeepers were found in the gut microbiota
   networks of FQ and FR groups, some of which (e.g., Lachnospiraceae and
   Oscillospiraceae) were found in both of the two networks. Reduced
   Lachnospiraceae was found in the intestine of PE MPs-exposed crayfish
   [[155]40]. The abundance of gut Oscillospiraceae was augmented in the
   intestine of mice exposed to foodborne PS MPs [[156]41]. Likewise, some
   nasal phylotypes assigned to bacteria, such as Alistipes and
   Sphingomonas, were determined as structural gatekeepers in the AQ and
   AR groups, respectively. Gut Alistipes was positively correlated with
   gastrointestinal dysfunction in the individuals consuming food in
   disposable plastic containers [[157]42]. Nasal Sphingomonas was
   increased in the airborne PS MPs-treated mice [[158]10].

   Multiple gut/serum metabolic and liver transcriptomic alterations were
   determined in the FQ and FR groups, suggesting both foodborne PLA NPs
   and MPs could cause metabolic disruption and transcriptomic
   dysregulation in mice. Two groups of gut metabolic pathways were
   enriched in the FQ and FR groups, among which “central carbon
   metabolism in cancer” was largely enriched in both groups. “Central
   carbon metabolism in cancer” was enriched in the intestines of rats
   orally exposed to PS NPs [[159]43]. Likewise, multiple serum metabolic
   pathways were enriched in the FQ and FR groups, of which “linoleic acid
   metabolism” was greatly enriched in both the two groups. “Linoleic acid
   metabolism” was enriched in the zebrafish embryos exposed to PS NPs and
   MPs [[160]44]. Multiple liver pathways with enriched DEGs were found in
   the FQ and FR groups, the “circadian rhythm” of which was greatly
   enriched in both groups. PS NP oral exposure was found to induce
   neurotoxicity in mice by affecting pathways related to circadian rhythm
   [[161]45]. The above three pathways, i.e., gut pathway “central carbon
   metabolism in cancer”, serum pathway “linoleic acid metabolism”, and
   liver pathway “circadian rhythm”, could be vital to the liver toxicity
   of foodborne PLA MNPs. We acknowledge they need to be further
   investigated.

   Various lung/serum metabolites and liver transcripts were altered in
   the AQ and AR groups, suggesting both airborne PLA NPs and MPs could
   induce metabolic disruption and transcriptomic dysregulation in mice.
   “Arachidonic acid metabolism” was the lung pathway most enriched in
   both AQ and AR groups, while “protein digestion and absorption” and
   “central carbon metabolism in cancer” were the two serum pathways
   greatly enriched in both AQ and AR groups. Abnormal “arachidonic acid
   metabolism” was determined in the PS NPs-exposed earthworms [[162]46].
   “Protein digestion and absorption” was enriched in Ietalurus punetaus
   larvae exposed to PS NPs [[163]47]. “Central carbon metabolism in
   cancer” was upregulated in PE MPs-exposed Daphnia magna [[164]48].
   Among the multiple liver pathways enriched in the AQ and AR groups,
   “alcoholism”, “systemic lupus erythematosus”, and “neutrophil
   extracellular trap formation” were largely enriched in both groups.
   “MAPK signaling pathway” was involved in chronic ethanol-induced
   alcoholism, which was associated with the reproductive toxicity caused
   by PS MP oral exposure [[165]49]. MPs could be a particular matter in
   the air and induce autoimmune diseases such as lupus erythematosus
   [[166]50]. PS NP exposure could induce neutrophil extracellular traps
   in mouse neutrophils [[167]51]. Lung pathway “arachidonic acid
   metabolism”, serum pathways “central carbon metabolism in cancer” and
   “protein digestion and absorption”, and liver pathway “alcoholism” were
   largely enriched in both AQ and AR groups, suggesting they could be
   important to the hepatotoxicity of airborne PLA MNPs. We acknowledge
   they deserve further exploration.

   Multiple correlations were determined between the altered components
   (i.e., microbes, metabolites, and transcripts) in the MNP groups, and
   some components played vital roles in the correlation networks: DEG
   Gm1673 in the FQ group; DEGs cyp2b9 and klhdc7b, and serum metabolite
   bacterioruberin diglucoside in the FR group; DEG zfp804b in the AQ
   group; and gut metabolite 11-cis-3,4-didehydroretinol in the AR group.
   Gm1673 was found over-expressed in the nucleus accumbens of
   nitroglycerin-treated mice [[168]52]. Cyp2b9 could be increased in the
   liver of mice orally exposed to PVC MPs [[169]53]. Klhdc7b was enriched
   in breast cancer tissue and cells [[170]54]. Zfp804b encoded zinc
   finger proteins could be increased by maternal fluoxetine exposure
   [[171]55]. Cyp27c1 was found to converse 11-cis retinal to
   11-cis-3,4-didehydroretinal in the eye of zebrafish [[172]56]. Some
   pathways were enriched in more than one organ and/or serum of the FQ,
   FR, AQ, or AR groups, and “circadian rhythm” was commonly enriched in
   the liver of the four groups. These pathways were likely to play vital
   roles in the liver toxicity induced by foodborne and airborne PLA NPs
   and MPs, and we acknowledge that they need to be further investigated
   in future work.

   Multiple factors could differentiate foodborne PLA NP and MP exposures.
   The two differential gut bacteria between the FQ and FR groups, i.e.,
   Lachnospiraceae_A2 and Roseburia, were greatly more enriched in the FQ
   and FR groups, respectively. They have been mentioned earlier in this
   Discussion section for their associations with MNPs [[173]28,[174]31].
   Six gut pathways were at different levels between the FQ and FR groups,
   and all of them were more enriched in the FQ group, among which
   “glycolysis/gluconeogenesis” was most enriched. Maternal PS MP exposure
   could disturb “glycolysis/gluconeogenesis” in the placenta of pregnant
   mice [[175]57]. Among the multiple differential serum and liver
   pathways between FQ and FR groups, “glycerophospholipid metabolism” in
   serum and “acute myeloid leukemia” in the liver were greatly more
   enriched in the FQ group, while serum pathway “choline metabolism in
   cancer” and liver pathway “retinol metabolism” were largely more
   enriched in the FR group. Disrupted “glycerophospholipid metabolism”
   was found in the gut of zebrafish exposed to a higher dose of
   polypropylene (PP) MPs (100 μg L^−1) [[176]58]. Increased occurrence of
   acute myeloid leukemia has been found in workers exposed to high levels
   of styrene [[177]59]. “Choline metabolism in cancer” could be enriched
   in the Daphnia magna exposed to PE MPs [[178]48]. “Retinol metabolism”
   in the mouse liver was disrupted by the PS MP oral exposure [[179]12].

   Likewise, multiple factors could differentiate airborne PLA NP and MP
   exposures. Among the differential airway bacteria between the AQ and AR
   groups, nasal Rikenellaceae_RC9_gut_group and lung Hungatella were
   greatly increased in the AQ group, while nasal Ileibacterium and lung
   Lachnospiraceae_A2 were largely elevated in the AP group.
   Rikenellaceae_RC9_gut_group was overrepresented in the intestines of
   zebrafish exposed to commercial PS MPs [[180]60]. Hungatella was at a
   higher level in the gut of PLA MPs-exposed silkworm than the control
   group [[181]61]. Gut Ileibacterium was enriched in the foodborne PS
   MPs-exposed mice [[182]41]. A shifted light-dark cycle could cause
   circadian disruption-induced metabolic syndrome via modulating the gut
   microbiota in mice, e.g., downregulation of Lachnospiraceae_A2
   [[183]62]. A variety of differential lung, serum, and liver pathways
   were found between the AQ and AR groups. Some of these differential
   pathways were more enriched in the AQ group, e.g., lung and serum
   pathway “ABC transporters” and liver pathway “hematopoietic cell
   lineage”, while some alternative differential pathways were more
   enriched in the AR group, e.g., lung pathway “purine metabolism”, serum
   pathway “d-amino acid metabolism”, and liver pathway “parathyroid
   hormone synthesis, secretion, and action”. PS NP exposure could enhance
   arsenic toxicity to AGS cells by disturbing the activity of ABC
   transporters [[184]63], and it could induce genotoxicity in the
   hematopoietic cell lines [[185]64]. Exposure to PS NPs could upregulate
   “purine metabolism” in the muscle of razor clams [[186]65]. PE MP oral
   exposure could enrich “d-amino acid metabolism” in the chicken's
   intestine [[187]66]. Copper oxide NP exposure enriched “parathyroid
   hormone synthesis, secretion, and action” in the human placenta
   [[188]67].

5. Conclusion

   In summary, foodborne and airborne PLA NPs and MPs could cause
   microbiota alteration, metabolic disruption, and transcriptomic
   dysregulation in mice. PLA MNP oral exposure could disrupt the “gut
   microbiota-gut-liver” axis and induce hepatotoxicity, while PLA MNP
   nasal exposure could influence the “airway microbiota-lung-liver” axis
   and cause hepatotoxicity. The relevant findings could benefit the
   clinical diagnosis of hepatotoxicity induced by foodborne and airborne
   PLA MNPs and the proper use of biodegradable materials in the
   environment.

CRediT authorship contribution statement

   Hua Zha: Conceptualization, Formal Analysis, Funding Acquisition,
   Investigation, Methodology, Writing - Original Draft, Writing - Review
   & Editing. Shengyi Han: Data Curation, Investigation. Ruiqi Tang:
   Investigation, Methodology. Dan Cao: Investigation. Kevin Chang:
   Writing - Review & Editing. Lanjuan Li: Funding Acquisition, Project
   Administration, Resources, Supervision, Writing - Original Draft,
   Writing - Review & Editing.

Declaration of competing interest

   The authors declare that they have no known competing financial
   interests or personal relationships that could have appeared to
   influence the work reported in this paper.

Acknowledgements