Abstract

   The widespread use of agricultural plastic films has made micro- and
   nanoplastics (MNPs) and phthalate esters (PAEs) contaminants of
   emerging concern in agroecosystems. However, the interactive mechanisms
   underlying their combined pollution in soil–plant systems remain
   elusive. To fill this gap, this study investigated the interaction
   between submicron plastics (SMPs, 0.01% and 0.1% w/w) and
   di(2-ethylhexyl) phthalate (DEHP) in soil–lettuce systems. Contrary to
   the anticipated synergistic toxicity, DEHP significantly reduced SMP
   uptake into and by cracked surface cells of lettuce roots (with root
   concentration factors decreasing by 19%–64%), i.e., DEHP alleviated
   SMP-induced oxidative stress, as evidenced by reduced levels of
   reactive oxygen species (−26.8% and −66.7%) and antioxidant enzyme
   activities (−118% and −128%). Metabolomic profiling revealed that SMP
   exposure significantly dysregulated multiple metabolic pathways (amino
   acid, carbohydrate, energy, glycan, lipid, and nucleotide metabolism),
   while SMP ​+ ​DEHP co-exposure selectively attenuated these metabolic
   disturbances, showing enrichment only in glycan biosynthesis/metabolism
   and suppressing SMP-induced perturbations in other pathways
   (biosynthesis of secondary metabolites, energy metabolism, and signal
   transduction). Microbial community analysis showed that high-level SMP
   exposure significantly diminished bacterial α-diversity and amplicon
   sequence variant (ASV) richness, whereas DEHP supplementation enhanced
   those of Myxococcota in the soil, potentially counterbalancing
   SMP-induced microbial dysbiosis. These findings collectively
   demonstrate that co-contamination by MNPs and plastic additives may
   produce antagonistic interactions rather than uniformly synergistic
   effects, and provide a more comprehensive evaluation of the risks of
   PAEs and MNPs to food security, human health, and ecological
   environment.

   Keywords: Co-contamination, Micro- and nanoplastics, Phthalate esters,
   Plant metabolome, Soil microbiomes

Graphical abstract

   [41]Image 1
   [42]Open in a new tab

Highlights

     * •
       DEHP suppressed submicron plastic (SMP) uptake in lettuce roots by
       19%–64%.
     * •
       Co-exposure alleviated SMP-induced oxidative stress of lettuce.
     * •
       DEHP mitigated SMP-induced metabolic dysregulation in lettuce.
     * •
       DEHP attenuated high-level SMP effects on soil bacterial
       communities.

1. Introduction

   Agricultural plastic film mulching represents a primary source of
   micro- and nanoplastics (MNPs) in soils [[43]1]. Through mechanical
   abrasion and weathering, these films progressively degrade [[44]2],
   lose flexibility, and fragment into smaller particles, including
   microplastics (MPs, <5 ​mm) and submicron plastics (SMPs, <1 ​μm) or
   nanoplastics (NPs, <100 ​nm) [[45]3]. Importantly, plastic films and
   related products commonly incorporate phthalate esters (PAEs) as
   plasticizers [[46]4]. Since these additives are not covalently bonded
   to polymer matrices, they can easily leach into adjacent soils [[47]5].
   Among PAEs, di(2-ethylhexyl) phthalate (DEHP) is the most prevalent
   compound detected in agricultural plastic films, exhibiting a 70%
   detection frequency with maximum concentrations reaching 28.2% (w/w)
   [[48]6]. Substantial evidence indicates that exposure to MNPs induces
   multifaceted toxicological effects, including oxidative stress,
   cellular damage, compromised barrier integrity (intestinal, air-blood,
   and placental), immune dysregulation, endocrine disruption, and
   developmental toxicity [[49]7], with toxicity showing an inverse
   relationship to particle size [[50]8]. Furthermore, PAE plasticizers
   have been associated with potential carcinogenic, teratogenic, and
   reproductive health risks [[51]9].

   Global agroecosystems are subjected to substantial plastic pollution,
   with annual inputs estimated at 1.15–2.41 million tonnes of plastic
   waste [[52]10], inevitably leading to significant contamination by MNPs
   and PAEs. The agroecosystems of China alone used approximately 2.5
   million tonnes of plastic film in 2017, leaving over 0.46 million
   tonnes as residual waste that released 91.5 tonnes of PAEs during use
   and degradation, including 18.8 tonnes of dibutyl phthalate (DBP) and
   42.2 tonnes of DEHP [[53]11]. Soil investigation revealed substantial
   regional variability in PAE concentrations (0–1232 ​mg/kg), with the
   highest levels occurring in cotton fields of southern Xinjiang,
   followed by vegetable fields of Guangdong (0.20–33.6 ​mg/kg) and
   tobacco plantations of Guizhou (0.84–25.7 ​mg/kg) [[54]12]. Nationwide
   surveys showed average macroplastic residues of 103 ​kg/hm^2 and MP
   abundances of 4537 particles/kg (dry weight) across Chinese farmland
   [[55]13], with hotspot areas exhibiting MP concentrations up to 12,292
   particles/kg [[56]14]. The widespread contamination of agricultural
   soils by MNPs and PAEs has raised significant concerns regarding their
   combined ecological and health impacts.

   Of particular concern is the food chain transfer dynamics, where the
   simultaneous presence of SMPs, NPs [[57]15] and PAEs [[58]16] creates
   synergistic exposure risks. Plants act as critical biovectors in this
   process [[59]17], not only absorbing and accumulating both contaminant
   classes but also potentially enhancing their bioavailability. This dual
   contamination pathway substantially amplifies human exposure risks
   through dietary intake, as crops may deliver complex mixtures of MNPs
   and PAEs directly into the food supply. Current research indicates MNPs
   may enhance PAE phytotoxicity. Hydroponic studies demonstrate MPs can
   intensify DBP toxicity in lettuce (Lactuca sativa) [[60]18], while
   foliar exposure experiments reveal that amino-functionalized
   polystyrene (PS) NPs facilitate PAE uptake in maize (Zea mays L.),
   leading to enhanced photosystem II impairment [[61]19]. Existing
   evidence suggests MNPs can amplify co-pollutant toxicity through the
   vector effect [[62][20], [63][21], [64][22]]. Antagonistic interactions
   have also been documented, exemplified by NPs reducing DBP cytotoxicity
   in Streptomyces coelicolor M145 through adsorption-mediated
   bioavailability reduction [[65]23]. These interactions are governed by
   multiple factors, including contaminant ratios, sorption properties,
   MNP characteristics (crystallinity, morphology), and subsequent PAE
   degradation dynamics [[66]24]. Notably, current studies primarily
   examine MNP effects on co-pollutants, while largely overlooking
   pollutant effects on MNP toxicity, leaving PAE-MNP reciprocal
   interactions in soil–plant systems poorly characterized.

   This study examines the combined effects of SMPs and PAEs on plant SMP
   uptake, physiological responses, and soil microbial communities, with
   particular focus on PAE-mediated modulation of SMP stress in soil–plant
   systems. We hypothesize that: (1) co-exposure induces greater
   phytotoxicity than individual contaminants, and (2) PAE facilitates SMP
   accumulation in plants while modifying soil microbiota, collectively
   influencing plant physiology through metabolic alterations. Given that
   PS constitutes a significant MP contaminant in agroecosystems due to
   its widespread use in packaging materials [[67]25], and considering
   DEHP represents one of the most prevalent PAEs in agricultural soils
   [[68]12], we chose DEHP and PS SMPs as model contaminants, and
   conducted controlled pot experiments, then analyzed PS-derived SMP
   accumulation patterns, growth parameters and photosynthetic pigment
   levels, antioxidant system responses, and metabolomic profiles of
   lettuce, and soil bacterial community composition to systematically
   evaluate their individual and combined phytotoxic effects through root
   exposure assays, thereby testing the aforementioned hypotheses
   regarding their environmental impact. This integrated approach
   elucidates mechanistic interactions between PS SMPs and DEHP, providing
   essential data for ecological risk assessment of contaminant mixtures
   in agricultural ecosystems.

2. Materials and methods

2.1. Microcosm experiment

   The source of SMPs in the experiment was monodisperse PS microspheres
   with a diameter of 100–200 ​nm ([69]Fig. 1A) at an initial
   concentration of 10 ​mg/mL (w/v). To analyze the size distribution of
   the purchased monodisperse PS microspheres, a laser particle size
   analyzer was utilized ([70]Fig. 1B). Standard of DEHP (CAS: 117-81-7)
   with a purity ≥99.1% was used. All instruments, chemicals, and assay
   kits used in the experiments can be found in [71]Tables S1 and [72]S2.
   The soil used for this study was sandy with less than 3% clay,
   collected from a farmland in Huai'an, Jiangsu Province, China, which
   mainly comprises yellow cinnamon soil [[73]26]. Soil samples collected
   at a depth of 0–20 ​cm were air-dried, ground to pass through a 2 ​mm
   nylon sieve, and stored at 25 ​°C. [74]Table S3 presents an overview of
   the basic physicochemical properties of the soil.

Fig. 1.

   [75]Fig. 1
   [76]Open in a new tab

   Scanning electron microscopy (SEM) images (A) and size distribution (B)
   of the added polystyrene (PS) submicron plastics (SMPs) in ultrapure
   water. SMP abundance (mg/g dry weight) in roots (C), SEM images of root
   surfaces (D) and stem cross sections (E) of lettuce grown in soil with
   different treatments. SMPs are marked with red ovals in the SEM images
   of the root surface and stem section. e, f, g, h, l, and k show
   enlargements of the areas indicated by the red squares of a, b, c, d,
   i, and j, respectively.

   Lettuce (Lactuca sativa) is a globally cultivated leafy vegetable
   commonly consumed raw and holds significant dietary importance across
   diverse geographical regions [[77]27]. The study design included a
   control group (blank soil), a DEHP exposure group (10 ​mg/kg), two SMP
   exposure groups (0.01% and 0.1% w/w), and two combined exposure groups
   (10 ​mg/kg DEHP ​+ ​0.01% w/w SMP and 10 ​mg/kg DEHP ​+ ​0.1% w/w SMP)
   ([78]Table 1). The exposure concentrations of SMPs and DEHP were
   selected based on: field survey data estimating macroplastic and MP
   loads [[79]13,[80]14] and DEHP residue levels [[81]12] in Chinese
   agricultural topsoils [[82]28] and assuming that for the worst-case
   scenario, all macroplastics and MPs present would convert at a certain
   time to SMPs; and (2) using established dose–response relationships
   from previous phytotoxicity studies on plant–MNP [[83]29,[84]30] and
   plant–DEHP [[85]31,[86]32] interactions, thus matching methodological
   consistency with existing research. In detail, methanol was used to
   dissolve DEHP standard solutions of varying volumes, which were then
   individually introduced into 600 ​g of soil. The mixture was thoroughly
   blended to ensure homogeneity. Initially, DEHP concentrations in the
   soil were set at 100 ​mg/kg. To achieve the desired final DEHP
   concentrations of 10 ​mg/kg, uncontaminated soil was subsequently added
   at a ratio of 9:1 to the mixture, then aged for three weeks in a dark
   environment at 25 ​°C with 20% relative humidity. To account for
   potential toxic effects of the methanol, control and the soils intended
   for treatment with SMPs were exposed to an identical methanol dosage
   and aging time as that administered to the soils contaminated with
   DEHP. SMP suspensions were homogenized using an ultrasonic vibration
   generator for 30 ​min in a water bath. The suspensions were prepared at
   concentrations of 0, 500, and 5000 ​mg/L prior to homogenization. Next,
   60 ​mL of SMP solutions (0, 500, and 5000 ​mg/L) were applied to the
   DEHP-contaminated soil and the control soil of 300 ​g in planters.

Table 1.

   Experimental design for the microcosm.
              Group            Abbreviation               Reagent
   Control check               Control
   DEHP                        D            10 ​mg/kg DEHP
   Low level of SMPs           P1           0.01% (w/w) SMPs
   High level of SMPs          P2           0.1% (w/w) SMPs
   Low level of SMPs ​+ ​DEHP  DP1          0.01% (w/w) SMPs ​+ ​10 ​mg/kg DEHP
   High level of SMPs ​+ ​DEHP DP2          0.1% (w/w) SMPs ​+ ​10 ​mg/kg DEHP
   [87]Open in a new tab

   Note: concentrations were measured by soil dry weight.

   DEHP, Di(2-ethylhexyl) phthalate; SMPs, submicron plastics.

   All treatments were replicated four times. The shape of the planter was
   a trapezoidal column with a base diameter of 9 ​cm and a height of 9
   ​cm. Details of plant culture are described in [88]SI Section 1. Leaves
   and roots from each parallel group were collected after 25 days, then
   washed, dried, and stored in the −80 ​°C refrigerator. The artificial
   climate chamber was set with light–dark cycle of 16:8 ​h, temperature
   cycle of 25:20 ​°C, and relative humidity cycle of 60%:30%. 10 ​mL of
   Hoagland nutrient solution was added each day for each planter. The
   formula of the modified Hoagland nutrient solution is shown in
   [89]Table S4.

2.2. Quantification and characterization of SMPs in lettuce

   The quantification method for PS SMPs in lettuce adopted the styrene
   dimer as the characteristic substance [[90]33]. In brief, the fresh
   leaf and root samples were washed, dried, crushed, ground, and
   extracted using a series of solvents, then analyzed using pyrolysis
   coupled with gas chromatography/mass spectrometry (Pyr-GC/MS). More
   details of quantification and characterization of SMPs and Pyr-GC/MS
   parameters are shown in [91]SI Section 2 and [92]Table S5. The washing
   steps likely removed all SMPs loosely attached to the roots; yet, it
   cannot be fully excluded that some of the SMPs remained stuck in or at
   disturbed surface cells, as also discovered by SEM ([93]Fig. 1D).

   The surface morphology of SMPs ([94]Fig. 1A) and the accumulation of
   SMPs on lettuce root and in lettuce stem cross sections were examined
   at an accelerating potential of 20 ​kV in high vacuum mode with
   backscatter detection, and observed by scanning electron microscopy
   (SEM).

2.3. Analysis of biomass, photosynthetic pigments, and antioxidants

   Fresh leaves and roots were washed in distilled water, dried, and then
   weighed. Since the fibrous roots of lettuce were still young and
   developing at the time of harvest, it was not possible to completely
   remove them from the soil. For this reason, the length and weight of
   the roots were not measured. The total amount of chlorophyll and
   carotenoids was determined by UV spectrophotometry. Details are
   described in [95]SI Section 3. Fresh leaf and root samples were
   promptly frozen using liquid nitrogen and then finely ground into
   powder. For each sample, 0.20 ​g of the powder was carefully
   transferred into a centrifuge tube with 2 ​mL of extract. Subsequently,
   the tubes underwent centrifugation at a speed of 8000×g for 10 ​min,
   all while maintaining a cool temperature of 4 ​°C. As a result of this
   process, the supernatants, which contained the desired components, were
   successfully separated and subsequently stored in 10 ​mL tubes. To
   determine the levels of reactive oxygen species [ROS, superoxide anion
   (O[2]^−) and hydrogen peroxide (H[2]O[2])], as well as to assess the
   activities of the enzymes superoxide dismutase (SOD), catalase (CAT),
   and peroxidase (POD), specific assay kits ([96]Table S2) were employed
   according to provided instructions.

2.4. Metabolomics analysis of lettuce leaves and roots

   The collected lettuce leaves were stored and processed to extract
   metabolites using various techniques, including washing, crushing,
   vortexing, and centrifugation. The resulting supernatant was analyzed
   using UPLC-MS/MS in both positive and negative modes to study the
   effects of DEHP and SMPs on lettuce metabolism. Further measurement and
   data analysis information can be found in [97]SI Section 4. The
   samples' raw sequence data has been uploaded to the MetaboLights
   ([98]https://www.ebi.ac.uk/metabolights) (MTBLS12878).

2.5. Analysis of soil microbial community structure

   The DNA from the soil was extracted using the PowerSoil DNA Isolation
   Kit following the provided instructions. Agarose gel electrophoresis
   was performed to assess the quality of the extracted DNA, while a
   NanoDrop 2000 UV-Vis spectrophotometer was used to determine DNA
   concentration and purity. The bacterial 16S rRNA gene V3–V4
   hypervariable region was amplified using universal primers 338F and
   806R. PCR master mix kit and the extracted DNA were used for the
   amplification process following the standard PCR protocol. The
   resulting PCR products were analyzed using agarose gel electrophoresis
   and sent for Illumina sequencing to Majorbio Bio-Pharm Technology Co.,
   Ltd. (Shanghai, China). Details of sequence data processing are
   provided in [99]SI Section 5. The samples' raw data have been deposited
   to National Center for Biotechnology Information (NCBI,
   [100]https://www.ncbi.nlm.nih.gov/) under the BioProject number
   PRJNA1308824.

2.6. Statistical analysis

   The root concentration factor (RCF) was the ratio of the abundance of
   SMPs in and on lettuce root relative to that in soil. Data was
   processed statistically by analysis of variance (ANOVA), and the least
   significant difference (LSD) post-hoc test was used to detect
   significant differences among mean values at p ​< ​0.05 using SPSS
   23.0. The figures were prepared using Origin 9.0.

3. Results and discussion

3.1. DEHP co-exposure inhibited SMP translocation and accumulation

   After root exposure to SMPs, large amounts of SMPs were detected in the
   lettuce roots of P1, P2, DP1, and DP2 treatments ([101]Fig. 1C), and
   were found attached to the surface of lettuce roots ([102]Fig. 1D).
   There was a significant difference between the treatment with low SMP
   additions (P1: 0.29 ​± ​0.03 ​mg/g) and the one with additionally added
   DEHP (DP1: 0.10 ​± ​0.05 ​mg/g); likewise, at high SMP loads, the
   concentration in the lettuce root of P2 (0.96 ​± ​0.05 ​mg/g) was
   significantly higher than in DP2 (0.78 ​± ​0.06 ​mg/g) ([103]Fig. 1C).
   P2 and DP2 treatments showed dense mats of SMPs attached to the xylem
   in the main veins of lettuce leaves, while the P1 and DP1 treatments
   did not ([104]Fig. 1E). When relating these results to the overall SMP
   added, DEHP thus decreased the RCF of SMPs from soil to lettuce roots
   from originally 0.96–2.88 without DEHP (P1, P2) to 0.78–1.03 after DEHP
   addition (DP1, DP2).

   Significant amounts of SMPs were detected in washed lettuce roots
   ([105]Fig. 1C), indicating their potential adherence to or embedding
   within the root surface ([106]Fig. 1D). However, the detection of SMPs
   in stem tissues ([107]Fig. 1E) suggests that despite their diameter
   ([108]Fig. 1B) exceeding the plant cell size exclusion limits (<20 ​nm)
   [[109]34] by up to five-fold, a portion of SMPs may have successfully
   penetrated the root system and translocated to aerial tissues. In the
   apoplastic pathway, the Casparian strip serves as a formidable physical
   obstacle [[110]35]. However, SMPs are capable of traversing the
   plasmodesmata (1–100 ​nm), despite being unable to penetrate the cell
   wall (<10 ​nm) [[111]36]. As a result of swelling or structural
   changes, the pore size of plasmodesmata increases, allowing larger
   particles to enter the cell [[112]37]. Once in the xylem, the particles
   may then be transported to above-ground tissues [[113]38]. However,
   likely due to the retention in soil [[114]39], SMPs accumulated in the
   above-ground parts of lettuce only at high SMP exposure (0.1% w/w)
   ([115]Fig. 1E). This is markedly different from the results obtained in
   hydroponic systems, where at an exposure level of 40 ​mg/L SMP, the
   root uptake of plastic particles and their translocation to the aerial
   parts of lettuce were observed under the influence of transpiration
   pull [[116]40].

   One possible way for lettuce roots to absorb SMPs in the soil or from
   hydroponic solution is through the root tip and the crack entry method,
   where the lateral roots appear [[117]15]. Crack entry of SMPs can be
   facilitated by root openings formed during plant aging, as well as
   after damage by soil-borne pathogens and physical destruction
   [[118]41]. These root openings can serve as entrances for SMPs in
   plants. The root apical meristem is highly porous due to active cell
   division in the root tip, and SMPs can be taken up through the root
   meristem and then penetrate the cell wall into the xylem, where they
   can move to the upper part of the plant under the root pressure
   [[119]15] and accumulate in the main veins of lettuce leaves ([120]Fig.
   1E). Upon irrigation, soil pores fill with water and induce buoyancy of
   SMPs, thus facilitating root uptake. When the density of PS SMPs (1.064
   ​g/cm^3) is less than Hoagland nutrient solution (2.19 ​g/cm^3), it is
   theoretically also possible that the floating SMPs migrate toward the
   soil surface and penetrate directly into the main leaf veins; yet,
   large numbers of SMPs were found in the cross-section of the main veins
   of lettuce leaves at locations close to the soil surface ([121]Fig.
   1D). Despite existing research efforts, it still remains uncertain
   whether SMPs have the capability to penetrate the stems and principal
   leaf veins of lettuce in a direct manner.

   Irrespective of the migration mechanism, and in contrast to our main
   hypothesis, DEHP reduced SMP uptake and subsequent plant performance.
   The log K[oc] value for DEHP is in the range of 4.9–5.7 [[122]42],
   i.e., the majority of DEHP added should stick to the soil. In
   principle, elevated additions of strongly adsorbing DEHP, could induce
   an aggregation of soil particles as observed for other soil organic
   matter constituents [[123]43,[124]44], yet, a high K[oc] value does not
   imply that sorption is irreversible. In fact, there might be
   interactions with dissolved organic matter. Also, PS SMPs might be a
   stronger bonding partner for DEHP than certain soil constituents, thus
   allowing a transfer of DEHP from soil surfaces to PS SMPs even at
   elevated log K[oc] values. Such interactions may then affect the uptake
   of SMPs by at least two mechanisms. On the one hand, surface-altered PS
   SMPs could show increased bonding to soil surfaces and thus enhanced
   soil retention. On the other hand, the SMPs could agglomerate, increase
   in size with different materials attached, and change their zeta
   potential, which are all critical parameters for the uptake of at least
   nanoparticles by plants, and thus also possibly of larger SMP
   aggregates [[125]36].

3.2. DEHP mitigates SMP-induced biomass reduction and attenuates oxidative
stress in lettuce

   Plant growth is the most intuitive indicator of toxicity, and previous
   studies have confirmed the negative effect of DEHP [[126]45,[127]46]
   and SMPs [[128]40,[129]47] on plant growth. In this study, as SMP
   concentrations increased, a detrimental effect on lettuce growth
   evolved, with larger effects on plant height and leaf weight ([130]Fig.
   2A–C). Both the exposure to SMP and DEHP also reduced the contents of
   chlorophyll and carotenoids in lettuce leaves ([131]Fig. 2D and E).
   Chlorophyll and carotenoids are crucial for photosynthesis and perform
   essential functions in this process [[132]48]. A decrease in plant leaf
   pigment content ([133]Fig. 2D and E) affected the photosynthetic rate,
   thus decreasing the biomass of lettuce ([134]Fig. 2A–C). Interestingly,
   the co-exposure of SMP and DEHP exhibited a significantly weaker
   inhibitory effect on lettuce growth compared to SMP treatment alone
   ([135]Fig. 2A–E), suggesting that the phytotoxic impacts of SMP and
   DEHP on lettuce photosynthesis were not synergistic. This finding
   contrasts with the results in a former report that di-n-octyl phthalate
   (DOP) (10 ​mg/kg) exacerbated the adverse effects of PS MPs (<13 ​μm,
   0.5%) on photosynthetic efficiency [[136]49]. This discrepancy may
   arise from the distinct physicochemical properties of phthalate
   compounds, which modulate their interactions with MNPs, ultimately
   leading to differences in the bioavailability effects of DOP and DEHP.

Fig. 2.

   [137]Fig. 2
   [138]Open in a new tab

   Plant height (A), number (B), fresh weight (C), total chlorophyll (D),
   and carotenoids (E) contents of lettuce leaves. The O[2]^− (F) and
   H[2]O[2] (G) contents, SOD (H), CAT (I), and POD (J) activities of
   lettuce leaves and roots under different treatments. Different
   lowercase letters indicate the significant difference among soils at p
   ​< ​0.05.

   ROS serve as pivotal signaling molecules and master regulators in
   plants, orchestrating diverse physiological processes spanning cellular
   metabolism, growth regulation, and developmental programming, while
   simultaneously mediating adaptive responses to both abiotic and biotic
   stressors [[139]50]. Under stress conditions, the disruption of
   cellular redox homeostasis through ROS accumulation triggers profound
   physiological shifts, wherein the altered redox status not only
   manifests as visible stress phenotypes but also directly correlates
   with the phytotoxic consequences of oxygen radical overproduction
   [[140]51]. Our results demonstrate a dose-dependent accumulation of ROS
   (O[2]^− and H[2]O[2]) in lettuce tissues under SMP exposure, with two
   key observations: (1) O[2]^− production peaked in SMP-only treatments
   without DEHP co-exposure ([141]Fig. 2F), and (2) root H[2]O[2] levels
   showed particular sensitivity to SMPs, increasing up to 237% compared
   to SMP ​+ ​DEHP combined treatments ([142]Fig. 2G). The elevated ROS
   content across all exposure groups (SMP, DEHP, and their combinations)
   triggered compensatory activation of antioxidant enzymes ([143]Fig.
   2H–J), reflecting that ROS serve both as oxidative damage markers and
   signaling molecules initiating defense mechanisms [[144]52]. These
   findings collectively indicate that SMP and DEHP exposure induces
   physiological stress in lettuce, disrupting redox homeostasis and
   activating the antioxidant system as a protective measure against
   xenobiotic-induced oxidative damage.

   SOD serves as the frontline enzymatic defense, catalyzing the
   disproportionation of O[2]^− into H[2]O[2] and O[2], playing a pivotal
   role in detoxifying ROS [[145]53]. Subsequently, H[2]O[2] generated as
   a product of SOD activity is effectively decomposed by both POD and
   CAT. Within plant systems, POD serves as a crucial H[2]O[2]-scavenging
   enzyme, exhibiting plant-specific functionality that is vital for
   safeguarding chloroplasts and other cellular components against
   oxidative damage induced by hydrogen peroxide and its derived hydroxyl
   radicals [[146]54]. Similarly, CAT acts as an essential metabolic sink
   for H[2]O[2], playing an indispensable role in plant stress defense
   mechanisms [[147]55]. Collectively, SOD [[148]56], CAT [[149]55], and
   POD [[150]54] form an integrated antioxidant network that mitigates
   oxidative stress by efficiently removing excess O[2]^− and H[2]O[2],
   thereby protecting plants against various biotic and abiotic stressors.
   Overall, the ROS levels and antioxidant enzyme activities in lettuce
   leaves and roots across different treatments exhibited the following
   trend: Control ​< ​D ​< ​DP1 ​< ​P1 ​< ​DP2 ​< ​P2 ([151]Fig. 2F–J),
   indicating that the combined toxicity of DEHP and SMPs was lower than
   that of SMPs alone. Comprehensive comparative analysis revealed that
   DP1 treatment significantly attenuated oxidative stress in lettuce,
   reducing total ROS (O[2]^− and H[2]O[2]) accumulation in roots and
   leaves by 26.8% relative to P1, while decreasing antioxidant enzyme
   (SOD, POD, and CAT) activity by 118%. Similarly, DP2 treatment
   demonstrated more pronounced effects, showing 66.7% reduction in total
   ROS content and 128% decline in antioxidant enzyme activity compared to
   P2. These findings suggest that DEHP addition mitigated the adverse
   effects of SMPs on lettuce, particularly under high SMP exposure
   ([152]Fig. 2). These findings differ from previous investigations
   examining: (1) lettuce responses to MP-DBP co-exposure [[153]18], where
   MPs (about 23 ​μm, 250 ​mg/L, 500 ​mg/L, and 1000 ​mg/L) were shown to
   amplify DBP (5 ​mg/L) phytotoxicity and impair photosynthetic
   function—consistent with our observations, though Gao et al. [[154]18]
   further demonstrated supra-additive toxicity in combined treatments;
   and (2) algal systems exposed to MP fibers (<20 ​μm) with DBP (1 ​mg/L)
   [[155]57], which exhibited concentration-dependent interactions ranging
   from synergistic toxicity at high MP concentrations (100 ​mg/L) to
   antagonistic effects at lower doses (1 ​mg/L). The observed
   discrepancies in experimental outcomes may be attributed to multiple
   factors, including variations in contaminant concentrations and MNP
   particle sizes. However, it is crucial to emphasize the distinct
   phytotoxic responses between hydroponic and soil–plant systems, with
   the latter representing a more complex and ecologically relevant
   matrix. Our findings specifically address this knowledge gap by
   elucidating how the intricate interactions within the soil environment
   modulate the combined phytotoxic effects of these contaminants.

3.3. Metabolic reprogramming in lettuce under DEHP and SMP co-exposure

   Through orthogonal partial least-squares discriminant analysis
   (OPLS-DA) screening (VIP > 1, p ​< ​0.05), treatment-specific
   differential metabolites were identified in both lettuce roots and
   leaves ([156]Fig. 3). Notably, root tissues exhibited a significantly
   higher abundance of differential metabolites than leaves ([157]Fig. 3).
   Metabolic profiling revealed a concentration-dependent response, with
   elevated SMP exposure leading to a marked increase in differential
   metabolite accumulation ([158]Fig. 3). Interestingly, DEHP co-exposure
   attenuated these metabolic perturbations, yielding fewer differential
   metabolites with diminished expression changes compared to SMP-only
   treatments ([159]Fig. 3). These metabolic alterations were functionally
   correlated with physiological modifications in lettuce. Multivariate
   analysis, coupling both principal component analysis (PCA) and OPLS-DA,
   elucidated treatment-specific metabolic patterns. PCA score plots
   demonstrated that SMP stress induced more pronounced metabolic shifts
   in roots ([160]Fig. S1A) than in leaves ([161]Fig. S1B), where
   treatment clusters showed considerable overlap, suggesting limited
   treatment effects on foliar metabolites. In contrast, supervised
   OPLS-DA achieved superior group discrimination ([162]Figs. S2, S3),
   with tight clustering of replicates within treatments and clear
   separation between treatments. This distinct metabolic fingerprinting
   indicates fundamentally different physiological impacts between
   SMP-only exposure and combined SMP-DEHP treatments on lettuce growth.

Fig. 3.

   [163]Fig. 3
   [164]Open in a new tab

   Volcano map of differential metabolites in the lettuce roots and leaves
   under P1, P2, D, DP1, and DP2 treatments compared to control. The
   horizontal axis represents the fold change (FC) in metabolite
   expression between the two groups, while the vertical axis represents
   the statistical significance of the differential expression. Higher
   values indicate more significant differential expression. Both the
   horizontal and vertical coordinates are logarithmically transformed.
   Each point in the plot represents a specific metabolite, with the size
   of the point indicating the Variable Importance in Projection (VIP)
   value. VIP value indicates the impact and explanatory power of
   metabolite accumulation differences on sample group classification.
   Orange points denote significantly upregulated metabolites, green
   points denote significantly downregulated metabolites, and gray points
   represent metabolites with non-significant differences.

   In this study, the differential metabolites screened mainly mapped to
   the following Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways.
   Under SMP treatment (P1, P2), the enriched metabolism pathways included
   amino acid metabolism, biosynthesis of other secondary metabolites,
   carbohydrate metabolism, energy metabolism, global and overview maps,
   glycan biosynthesis and metabolism, lipid metabolism, membrane
   transport, metabolism of other amino acids, nucleotide metabolism,
   signal transduction, and translation ([165]Fig. 4A). It illustrated
   that SMP exposure significantly altered metabolite expression of amino
   acid, carbohydrate, energy, glycan, lipid, and nucleotide, causing the
   original dysfunction of those regulatory pathways. SMP ​+ ​DEHP
   co-exposure resulted in distinct metabolic responses, failing to induce
   enrichment in specific pathways: biosynthesis of other secondary
   metabolites, energy metabolism, global and overview maps, metabolism of
   other amino acids, signal transduction, and translation, while inducing
   the enrichment of glycan biosynthesis and metabolism ([166]Fig. 4A).
   The results showed that DEHP had an obvious inhibitory effect on SMP
   stress-induced metabolic disorders within lettuce roots and leaves.

Fig. 4.

   [167]Fig. 4
   [168]Open in a new tab

   Enrichment analysis of the differential metabolites in the lettuce
   roots and leaves under P1, P2, DP1, and DP2 treatments. The bubble size
   and color represent the number of metabolites and the second category
   of metabolic pathway, respectively. (A) The abscissa coordinate
   represents the enrichment factor (EF) of metabolic pathway, EF
   represents the enrichment rate of metabolic pathways. (B) Differential
   abundance score represents the overall changes in all metabolites
   within a metabolic pathway. A positive value indicates an upward trend
   in the expression of all annotated differential metabolites in the
   pathway, while a negative value indicates a downward trend. The length
   of the line segment corresponds to the absolute value of the
   differential abundance score. Text in red indicates metabolic pathways
   that appeared in both the P1 and P2 treatments, while text in blue
   represents pathways present in both P2 and DP2. ∗, ∗∗, and ∗∗∗ indicate
   the significant meaning of the enrichment results at p ​< ​0.05, p ​<
   ​0.01, and p ​< ​0.001, respectively.

   KEGG pathway enrichment analysis identified distinct tissue-specific
   regulation patterns in glycerophospholipid metabolism, with significant
   downregulation observed in root tissues under P1 and DP2 treatments,
   while leaf tissues exhibited pronounced upregulation under P2 and DP1
   exposure conditions ([169]Fig. 4B), demonstrating opposing metabolic
   responses between belowground and aerial plant organs to differential
   nutrient treatments. Glycerophospholipids are fundamental components of
   cell membranes, crucial for cell structure, signaling, and function
   [[170]58]. Modulation of membrane lipid composition under varying
   environmental conditions is an important part of plant stress
   adaptation [[171]59]. Glycerophospholipid metabolism is a metabolic
   pathway that is commonly altered in response to stress in plants, which
   was observed in the response of Lactuca sativa leaves to SMP stress
   [[172]40], brassica juncea roots to cadmium stress [[173]60], Cucumis
   sativus seedling leaves and cotton (Gossypium barbadense and Gossypium
   hirsutum) [[174]61] to drought stress, and Brassica campestris to cold
   stress [[175]62]. Another metabolic pathway that plays an important
   role in regulating plant growth, development and tolerance to
   environmental stresses [[176]63] is the ABC transporters, which was
   also simultaneously enriched both in lettuce roots and leaves under P1
   and P2 treatments, while only in lettuce leaves under DP2 treatment
   ([177]Fig. 4B). The range of processes in which members of the various
   subclasses of plant ABC transporters have been implicated encompasses
   polar auxin transport, lipid catabolism, xenobiotic detoxification,
   disease resistance, and stomatal function [[178]64]. Notably, the ABC
   transporter pathway exhibited upregulation in lettuce roots under the
   P2 treatment and in leaves under both P1 and P2 treatments. DEHP
   mitigated the disruption of ABC transporter function induced by SMPs in
   lettuce roots and leaves at low exposure levels, as well as in roots
   under high SMP exposure ([179]Fig. 4B). ABC transporters play a pivotal
   role in the spatiotemporal regulation of secondary metabolites that
   confer critical protective functions in plant physiological processes
   [[180]65], potentially explaining their observed upregulation under SMP
   exposure.

   Notably, we observed distinct differences in metabolic pathway
   enrichment patterns between P1 and DP1 treatments ([181]Fig. 4B),
   suggesting that DEHP suppresses low-level SMP-induced metabolic
   modification. Under high SMP exposure conditions (P2, DP2), four
   metabolic pathways showed consistent enrichment in lettuce roots:
   phenylalanine, tyrosine, and tryptophan biosynthesis, amino sugar and
   nucleotide sugar metabolism, galactose metabolism, and starch and
   sucrose metabolism. In leaves, only purine metabolism was concurrently
   enriched in P2 and DP2 ([182]Fig. 4B). Comparative analysis revealed
   that P2 treatment produced more differential metabolites and more
   significant metabolic alterations in leaves compared to DP2 treatment,
   while the opposite pattern was observed in roots ([183]Fig. 4B). These
   results demonstrate that DEHP plays a dual regulatory role: mitigating
   high-level SMP-induced metabolic dysregulation in leaves while
   stimulating another metabolic pathway enrichment in root tissues.

3.4. Rhizosphere bacterial communities under DEHP and SMP co-exposure

   MNPs can significantly disrupt mutualistic interactions between soil
   bacteria and plant root systems [[184]66]. Notably, bacterial
   communities exhibit greater sensitivity to DEHP contamination compared
   to fungal or archaeal populations [[185]67]. This differential
   responsiveness, coupled with bacteria's pivotal role in nutrient
   cycling and soil ecosystem functioning, makes them optimal
   bioindicators for investigating the individual and synergistic impacts
   of emerging contaminants on soil ecological health. [186]Fig. 5A
   displays the alpha diversity of the bacterial community, which was
   assessed using the Shannon, Simpson, and Chao1 diversity index under
   SMP, DEHP, and their combined exposure. Our analysis revealed that the
   addition of SMP had a significant impact (p ​< ​0.05) on these indices,
   particularly at high SMP levels (P2, p ​< ​0.01). The sequences were
   clustered into 4343 bacterial amplicon sequence variants (ASVs). With
   the increase in SMP levels, the number of ASVs in the soil
   substantially decreased ([187]Fig. 5B). Among the analyzed bacterial
   groups, only the facultative Armatimonadota bacteria showed significant
   responses to both SMP and DEHP exposure. Several phyla, including
   Proteobacteria, Firmicutes, Bacteroidota, Cyanobacteria, and
   Sumerlaeot, were significantly influenced by the SMP treatments.
   Additionally, Gemmatimonadota, Myxococcota, Armatimonadota,
   Dependentiae, Nitrospirota, and GAL15 exhibited significant alterations
   specifically under DEHP exposure ([188]Fig. 5C).

Fig. 5.

   [189]Fig. 5
   [190]Open in a new tab

   (A) Alpha diversity index, and (B) Venn diagram of soil bacterial
   communities on ASV level under different treatments. (C) Heatmap of the
   relationship of SMP abundance/DEHP concentration and soil bacterial
   phylum. Different lowercase letters indicate the significant difference
   among soils at p ​< ​0.05. ∗, ∗∗, and ∗∗∗ indicate significant
   difference between different comparison groups at p ​< ​0.05, p ​<
   ​0.01, and p ​< ​0.001, respectively.

   The presence of SMPs can modify the size distribution of water-stable
   soil aggregates [[191]68], consequently influencing soil bacterial
   communities [[192]69]. Our results showed that SMPs significantly
   enhanced the relative abundance of Proteobacteria (p ​< ​0.001,
   [193]Fig. 5C). This finding aligns with previous observations in paddy
   soils, sediments, and seawater, suggesting that SMPs may consistently
   promote favorable conditions for Proteobacteria across various
   environments. Conversely, SMP exposure significantly reduced the
   relative abundances of Firmicutes, Bacteroidota, Cyanobacteria, and
   Sumerlaeot in a dose-dependent manner (p ​< ​0.01, [194]Fig. 5C).
   Notably, both Firmicutes [[195]70] and Sumerlaeota [[196]71] are
   recognized for their stress-resistant properties. The phylum
   Bacteroidota is a major bacterial group in the plant rhizosphere
   [[197]72], and Cyanobacteria is an important phylum of bacteria that
   increase plant tolerance because of their direct action in increasing
   the availability of water in the soil, improving its physicochemical
   conditions faced with different abiotic stresses [[198]73]. These
   results indicate that SMP exposure negatively impacts both soil
   bacterial stress adaptation mechanisms. Given the intimate relationship
   between soil microbial communities and plant growth in soil–plant
   systems [[199]74], such SMP-induced microbial shifts may have important
   implications for plant development.

   DEHP exposure significantly reduced the relative abundance of
   Gemmatimonadota, Nitrospirota, and GAL15 (p ​< ​0.05, [200]Fig. 5C).
   Gemmatimonadota, commonly found in plant-associated environments and
   the rhizosphere, plays important ecological roles in organic matter
   degradation through its ability to utilize diverse organic compounds as
   carbon and energy sources, thereby contributing to ecosystem stability
   [[201]75]. Nitrospirota is particularly notable for its essential
   function in soil nitrification processes [[202]76]. Conversely, DEHP
   increased the relative abundance of Myxococcota, Armatimonadota, and
   Dependentiae (p ​< ​0.05, [203]Fig. 5C). Notably, Myxococcota, as apex
   predators in soil microbial food webs, possesses unique predatory
   capabilities and strong soil adaptability that enable it to regulate
   microbial ecological networks and suppress plant disease [[204]77].
   However, the enrichment of Myxococcota suggests a potential role in
   mitigating SMP stress, though further genomic analysis is needed to
   confirm their functional contributions. These results suggest that
   while DEHP exposure negatively affects the rhizosphere environment and
   nutrient cycling processes, it may simultaneously alleviate SMP
   phytotoxicity through specific alterations in bacterial community
   composition, particularly by enhancing beneficial bacterial phyla.
   Supporting these findings, previous research has demonstrated that DEHP
   accumulation in soil can reduce metabolic activity and virulence
   factors in the plant pathogen Acidovorax citrulli [[205]78].

4. Conclusion

   MNPs and PAEs, as unavoidable emerging contaminants in agricultural
   soils, exhibit complex interactions that defy traditional assumptions
   of additive or synergistic toxicity. Our study demonstrated that DEHP
   unexpectedly mitigated the phytotoxicity of SMPs by (1) inhibiting SMP
   uptake by lettuce roots, (2) alleviating metabolic dysregulation in
   lettuce, and (3) stabilizing soil microbial community structure. These
   findings challenge conventional contaminant risk assessment paradigms
   and suggest that the phytotoxicity of SMPs and DEHP is not synergistic
   but potentially antagonistic. This antagonistic effect underscores the
   necessity of incorporating contaminant interaction dynamics into
   ecological risk assessment frameworks, particularly for plastics and
   their derived pollutants. Future research should focus on evaluating
   contamination risks under conditions where multiple emerging pollutants
   coexist and interact, thereby enabling a comprehensive assessment of
   their ecological and human health risks.

CRediT authorship contribution statement

   Yu Wang: Writing – review & editing, Writing – original draft,
   Visualization, Methodology, Funding acquisition, Formal analysis, Data
   curation, Conceptualization. Fang Wang: Writing – review & editing,
   Project administration, Funding acquisition, Conceptualization. Leilei
   Xiang: Writing – review & editing, Methodology. Maoyuan Liao:
   Methodology. Mingyi Wang: Methodology. Yongrong Bian: Methodology. Xin
   Jiang: Project administration. Ravi Naidu: Writing – review & editing.
   Matthias C. Rillig: Writing – review & editing, Funding acquisition.
   Wulf Amelung: Writing – review & editing, Funding acquisition.

Declaration of competing interests

   The authors declare no competing financial interest.

Acknowledgments