Materials, Surfaces, and Interfacial Phenomena in Nanoplastics Toxicology Research

Abstract

In response to the growing worldwide plastic pollution problem, the field of nanoplastics research is attempting to determine the risk of exposure to nanoparticles amidst their ever-increasing presence in the environment. Since little is known about the attributes of environmental nanoplastics (concentration, composition, morphology, and size) due to fundamental limitations in detection and quantification of smaller plastic particles, researchers often improvise by engineering nanoplastic particles with various surface modifications as models for laboratory toxicological testing. Polystyrene and other commercially available or easily synthesized polymer materials functionalized with surfactants or fluorophores are typically used for these studies. How surfactants, additives, fluorophores, the addition of surface functional groups for conjugation, or other changes to surface attributes alter toxicological profiles remains unclear. Additionally, the limited polymers used in laboratory models do not mimic the vast range of polymer types comprising environmental pollutants. Nanomaterials are tricky materials to investigate due to their high surface area, high surface energies, and their propensity to interact with molecules, proteins, and biological probes. These unique properties can often invalidate common laboratory assays. Extreme care must be taken to ensure that results are not artefactual. We have gathered zeta potential values for various polystyrene nanoparticles with different functionalization, in different solvents, from the reported literature. We also discuss the effects of surface engineering and solvent properties on interparticle interactions, agglomeration, particle-protein interactions, corona formation, nano-bio interfaces, and contemplate how these parameters might confound results. Various toxicological exemplars are critically reviewed, and the relevance and shortfalls of the most popular models used in nanoplastics toxicity studies published in the current literature are considered.

Graphical Abstract

1. Introduction

Commercially produced polymeric nanoparticles have undergone widespread use as models for environmental plastic pollutants in toxicology experiments. Microplastics and nanoplastics may enter the environment as small particles, or they may arise from weathering degradation of bulk plastic waste. After disposal, these plastics are fragmented by photodegradation, biodegradation, hydrolysis, photooxidation and mechanical degradation, over time, into micro and nanoscale plastic particles and fibers (Andrady, 2011; Murphy et al., 2016; Kögel et al., 2020). Their overall impact on the environment is unknown while the environmental load continues to increase. As of 2015, over eight billion metric tons of plastic had been produced (Geyer et al., 2017). Plastics production surpasses all other synthetic material production globally, with 400 million metric tons produced annually (Geyer et al., 2017). This corresponds to a billion kilograms being produced each day. Of all the plastic produced through 2015, only about 9% had been recycled and 12% incinerated, while the remaining 79% was accumulated in landfills or the natural environment (Geyer et al., 2017). It has been estimated that over 5 trillion plastic pieces, weighing over 250 thousand tons, is presently contaminating the world’s oceans (Eriksen et al., 2014). The continued accumulation of these materials in the environment constitutes a major worldwide ecological and health crisis and research into the toxic effects is of the utmost importance.

Micro- and nano-scale plastic particles have been shown to cause a range of health problems in bacterial species, plants, animals, and humans. Upon entering the bodies of organisms, the plastic particles can be transported in blood (Kashiwada et al., 2006), enter cells (Vogt et al., 2006), or even cross the blood-brain barrier (Kashiwada et al., 2006). They exert toxic effects on the immune systems of mussels (Brandts et al., 2018a) and fish (Greven et al., 2016; Brandts et al., 2018b) and bioaccumulate in the gills, brain, testes, liver, and kidneys of the medaka fish (Kashiwada et al. 2006; Gundersen, 2019), and in the gut, liver, and kidneys of mice (Deng et al 2017). Behavioral disorders associated with nanoplastics in the brains of fish have also been reported (Mattsson et al., 2017). More recent investigations suggest that protein misfolding and enzyme alteration may be induced by nanoplastics in biological systems (Gopinath 2019; Hollóczki 2019; Kihara 2019; Auclair 2020).

Understanding the toxicological effects and mechanisms of toxicity of these materials in biological systems is crucial for the development of feasible countermeasures. Plastic production does not appear to be decelerating, and we can only anticipate that the plastics pollution problem will greatly increase in the foreseeable future. Current toxicological inquiries into nanoplastic particles encompass only a narrow range of materials, sizes, and morphologies, while demonstrating broad-range systemic toxicities in a variety of different organisms. Meanwhile, the role of surface effects on interactions with neighboring particles, solvents, proteins, cells, and systems is also significant. It is unclear how adequate laboratory engineered particles and laboratory models mimic environmental nanoplastic exposure as many facets of engineered nanoparticles remain understudied or even unknown. Additionally, nanoparticles are known to interfere with standard toxicological assays, dyes, fluorophores, and pH indicators, confounding results. We have reviewed some of the most common and significant recent research findings and present potential shortfalls in the methodology and models representing the current state of the field.

2. Polymers Used in Toxicological Studies vs. Environmental Abundance

Microplastics have been discovered in deep-sea sediments (Van Cauwenberghe et al., 2013) collected off the coasts of Japan, Thailand, Malaysia, and South Africa, in which a variety of polymers including PE, PP, PS, polyethyleneterphthalates (PET), polyethylene-polypropylene copolyper (PEP), and polyacrylates (PAK), polyvinyl chloride (PVC), polyamide (PA), ethylvinyl acetate (EVA), and polycaprolactone (PCL) were reported (Matsuguma et al., 2017). PE and PP were also found to comprise between 63 and 89 % of the microplastics collected in sediments from Venice Lagoon in Italy (Vianello et al., 2013). In addition, even sea salts have been found to contain microplastics, with more than half being <200 μm, and the most common polymers being PET, PE and cellophane (Yang et al., 2015). Although the relative abundance of each polymer is unclear and will likely differ depending on location, it is important to note that a wide range of polymer types have been found. Therefore, micro and nano-plastic toxicological studies should encompass each of the polymer types that exist as environmental pollutants.

PS and poly (methyl methacrylate) (PMMA) are the most-studied polymers for nanoplastics toxicology models with PS and PMMA NPs making up approximately 82 % and 10 % of toxicology experiments compared to the 8 % attributed to all other polymers (Shen et al., 2019). The relative abundance of other polymer types (PE, PP, PVC, PS, PET, PA) in the environment is uncertain and theoretical or estimated values are used (Ronkay et al., 2021). In a study of larger, marine water plastic particles (1–5 mm), 90 % of the particles were composed of PE and 10 % were composed of PP. Smaller plastic particles (20–999 μm) exhibited more diversity, including PE (73 %), PP (13 %), PVC (8 %), PS (2 %), and PET (1 %) (Ter Halle et al., 2017). Nanoplastics ranging from 1–99 nm in size exhibited similar diversity in composition. The mean relative proportions of the anthropogenic pyrolytic fingerprints, as determined by pyrolysis GC-MS, were 70 % PVC, 17 % PET, 9 % PS, and 4 % PE (Ter Halle et al., 2017). Considering these reports in light of the numerous published studies on the toxicology of nanoplastics, the polymers of PVC, PET, PA, and PP appear to have been largely ignored by the toxicologists. Although some variation is expected in the relative abundance of polymers in samples collected from various locations, it is certain that all these polymer types exist at some level in the environment and are of interest for future toxicological studies. However, at present, the nanoplastics toxicology studies available in the published literature do not present a complete picture of the materials comprising environmental plastic pollution (Fig.1).

Fig. 1.

Reported relative abundance of (a) microplastics and (b) nanoplastics in the envrionment, compared to the published toxicological studies on nanoplastics by polymer type (c).

Since there does not currently exist standardized methods for extraction and quantification and assessment of environmental nanoplastics (Shen et al., 2019), it remains difficult to determine which polymer types dominate these size ranges, and what an environmentally relevant concentration for toxicology studies is. Dong et al. (2018), investigating 108 nm PS in a nematode model, justifies their exposure concentration by citing a prediction by Lenz et al., 2016 on microplastics concentrations. Lenz et al., (2016) discusses how studies often use microplastics concentrations that are orders of magnitude higher than those reported for field studies. However, microplastics concentrations do not necessarily correspond to nanoplastics concentrations. To overcome this present gap in knowledge, it is important to utilize concentration curves, so that the data obtained therein may be applied to current or projected environment relevant concentrations.

3. Complex Contributions to Toxicological Experimentation

While it is known that plastic particles can adsorb chemical toxins onto their surfaces and subsequently transfer them throughout the bodies of organisms, characterizing environmentally relevant toxins and contaminants in the laboratory is complicated by potentially toxic additives and nanoplastic surface modifications. Many plastics are produced using additives such as binders, plasticizers, stabilizers, and flame retardants, which may leach out and compound the toxic effects of the nanoplastics. Furthermore, subsequent weathering of plastic fragments in the environment and eco-corona formation alters the surface properties and roughness of the particles. This can affect acid-base interactions, surface charges, adsorption affinity, density, interactions with biological molecules and ultimately, cellular uptake. These changes influence interactions with cells combined with the toxic additives contribute to the complex toxicological profile of environmental nanoplastics.

3.1. Contaminant Adsorption and Transport

The ability of plastic particles to adsorb contaminants onto their surfaces and potentially transport them into the bodies of organisms has been established (Pittura et al., 2018) and is attributed to the low polarity, high degree of surface roughness, high surface area, and varying surface chemistries of the particles (Hodson et al., 2017; Koelmans et al., 2015). Polychlorinated biphenyls (PCBs), polycyclic aromatic hydrocarbons (PAHs), dichlorodiphenyltrichloroethane (DDT), polybrominated diphenyl ethers (PBDEs), and metals have been shown to adsorb onto the surface of plastic fragments (Mato et al., 2001; Endo et al., 2005; Ogata et al., 2009; Law and Thompson 2014; Fisner et al., 2013a; Fisner et al., 2013b; Van Cauwenberghe et al., 2015; Beckingham and Ghosh 2017; Wang et al., 2020). Toxins such as bisphenol-A (BPA) could be transported by nanoplastics in this way to the brain (Chen et al., 2017a). Plastic bag fragments have been shown to transport heavy metals such as zinc (Hodson et al., 2017), and zinc deposition was greater in synthetic earthworm digestive tracts exposed to zinc-adsorbed HDPE.

The types of adsorbed substances are not limited to persistent organic pollutants or heavy metals however, pathogens have also been found to adsorb to and be subsequently transmitted by plastic pollution. For example, virulent fish and human pathogens, some with antibiotic-resistant genes, have been found on marine plastic debris (Radisic et al., 2020), and multiple-drug resistance was observed in bacterial isolates harvested from environmental polystyrene (Lagana et al., 2019). Therefore, not only is it possible for plastic particles to transmit disease, but it is also possible that antibiotic resistance could be spread through marine environments. Although it remains unclear as to what extent these plastic particles can transport the various substances throughout waterways and in the water cycle, the distinct possibility still exists and should not be overlooked. These considerations extend to the transportation of toxins across the blood-brain barrier, into organs and tissues, into gametes, and even across the placenta, warranting further research into plastic-mediated delivery of organic compounds and heavy metals.

3.2. Chemical leaching of nanoplastics

The toxicity of leaching nanoplastics is another emerging subfield of nanoplastics research. Synthetic polymers often contain additional chemical components to alter their properties (da Costa et al., 2016). Plastic monomers and additives are released from plastics during their decomposition process while plastic additives such as phenolic antioxidants have been shown to be transferred to foodstuffs from different types of plastic food packaging materials such as low-density polyethylene (LDPE), high-density polyethylene (HDPE), PP, acrylonitrile- butadiene- styrene [(ABS) and high- impact polystyrene (SB) (Bieber et al., 1985). Other contaminants include polymerization catalysts, initiators, accelerators, plasticizers, pigments, dispersants, surfactants, antistatic agents, fibers, fillers, nanofibers, and nanoparticle additives (Sherman 2015 da Costa et al. 2016). The commonly used plasticizers include, but are not limited to, anti-UV radiation stabilizing agents, phthalates, flame retardants and BPA. Since phthalates are not covalently bound within the polymer matrices, they are likely to leach from the plastic pollutants (Casals-Casas and Desvergne 2011). These additives are known to have endocrine disrupting effects (Casals-Casas and Desvergne 2011; Meeker et al., 2009; Arbuckle et al., 2016). Models investigating the leaching of plastic additives as well as the toxicity of nanoplastics with common additives are necessary to determine how these additives affect organisms alone and in combination with plastic particles during both acute and chronic exposures.

3.3. The effects of weathering

Environmental weathering of polymeric materials results in embrittlement and fragmentation of particles (Barnes et al., 2009; Andrady, 2011; Song et al., 2017). It has also been reported that environmental modification to the surfaces of plastic particles can increase the affinity of contaminants to the particles (Fotopoulau and Karapanagioti 2012). UV radiation, which has been shown to enhance the adsorption capacity of plastic particles for heavy metal ions (Wang et al., 2020). Additionally, accumulative surface roughness, due to erosion, increases the surface area of the particle, potentially enhancing adsorption. Notably, virgin plastic particles did not demonstrate any acid-base behavior, whereas eroded PE at seawater pH obtained a negative charge due to surface ketone functional groups (Fotopoulau and Karapanagioti 2012). In this case, environmentally weathered particles exhibit significantly different interactions with contaminants as well as their behavior in the environment. These findings provide evidence that the behavior of environmentally weathered particles in bacteria and in the bodies of organisms may be significantly different from lab-created, unaltered particles.

3.4. Corona formation

In the environment or in biological fluids, proteins, enzymes, molecules and ions can adsorb to the surface of the nanoplastic, due to its high surface energy, forming a corona. In biological fluids, proteins bind to the surface of the nanoplastic particles in a dynamic way; the composition of this protein corona changes over time, undergoing fluctuations in protein adsorption and dissociation (Sahneh et al., 2013; Cedervall et al., 2007). Protein corona formation can enhance particle stability by forming a layer between the nanoparticle and the solvent, thus, interfering with interparticle interactions and reducing or inhibiting agglomeration.

Due to the way in which proteins can interact relatively strongly with nanoplastics, corona formation may have a profound effect on toxicity. It has been noted that eco-corona formation reduced toxicity and may passivate surfaces (Saavedra et al., 2019; Schultz et al., 2021), but little else is known. On the other hand, protein corona formation on the surface of nanoparticles can affect the mechanisms of internalization (Francia et al., 2019), particle biodistribution, and biocompatibility (Aggarwal et al., 2009), impart biological identity to nanoparticles (Monopoli et al., 2012) and modulate pathobiological effects (Tenzer et al., 2013), including cytotoxicity and immunotoxicity (Lee et al. 2015; Westmeier et al., 2016). Protein coronas have the capability to enhance or reduce immune responses thereby influencing the entire immunological cascade (Lee et al., 2015). Another protein-based interaction with measurable impact is the interaction with enzymes. Plastic particles have been shown to alter certain enzymatic activities, altering the response to additional stressors, and have been shown to alter the toxicity of human pharmaceuticals in fish cell lines (Almeida et al., 2019).

3.5. Interactions in media

As with cytokines, proteins, fluorophores, and dye molecules and other assay components, growth media components also interact with nanoparticles. Corona formation is well known to occur, but in media, it may adopt the surface characteristics of the media, concealing the identity of the polymer to some extent. Abundance of serum in media is proportional to the protein content of the corona and influences its attributes. The increase in hydrodynamic size and changes to the surface charges due to corona formation in serum may may either enhance or reduce cell uptake of nanoparticles (Wilhelm et al, 2003; Xing et al., 2005; Doak et al., 2009). It is important that the contribution of serum in media to the composition and properties of the protein corona are considered and not solely attributed to the polymeric particle.

The study published by Manfra et al. (2017) demonstrated a different toxicological profile for rotifers (Brachionus calyciflorus) grown in artificial media (RSW) vs. natural media (NSW). (Saavedra et al., 2019) also reported that the amidine PS nanospheres exhibited a reduced zeta potential value in the Daphnia magna and Thamnocephalus platyurus culture media, which they attribute to the ionic strength of the solvent compressing the particle electrical double layer. They also report a shift from a positive to a negative zeta potential for the amidine particles in B. calyciflorus culture media. Aggregate formation in media vs. ultrapure water resulted in increased hydrodynamic diameters of the particles as well (Saavedra et al., 2019). Since we have observed large alterations in the zeta potential when the solvent is changed (Fig. 3), it is reasonable to suspect that media influences the particles sufficiently that toxicological experiments for identical nanoparticles in different media and different water samples, throughout a range of salinities, and pHs would be necessary to better understand nanoplastic toxicity. Since the environmental water parameters vary by location and with weather and seasonality, reporting the water parameters used in the studies, and noting their effect(s) on zeta potential, hydrodynamic size and comparing the dependence of the toxicological profiles on these parameters, will improve the accuracy and environmental relevance of the research findings. As more information on environmental nanoplastics becomes available, and the mechanisms of toxicity are uncovered, models with better environmental relevance can be developed and standardized.

Fig. 3.

Reported zeta potential values for PS beads used in nanoplastics toxicological studies. Values are shown for aminated polystyrene (APS), carboxylated polystyrene (CPS), fluorescent polystyrene (FPS), or “unfunctionalized” polystyrene (PS) in freshwater (FW), saltwater (SW), or media.

4. Particles with Artificially Engineered Surfaces as Models for Environmental Nanoplastics Toxicity

Commercially available or synthetically produced nanoplastic particles are often used to model environmental nanoplastics in toxicological experiments (Canesi et al, 2016; Manfra et al., 2017; Dong et al., 2018; Liang et al., 2021; Muhammad et al., 2021). However, commercial nanoplastics demonstrate different surface properties and also differ from environmental pollutants in morphological attributes, size ranges, and aspect ratio when compared to naturally weathered plastic fragments (Lambert et al., 2017; Burns and Boxall 2018). For example, weathered particles are expected to have rough surfaces and come in a range of sizes and morphologies, whereas engineered particles tend to be smooth, spherical, and monodispersed. In addition to the physical attributes, commercially produced beads also have surface chemistries that may differ from environmental nanoplastics.

Commercial PS and PE beads designed for use in the laboratory are typically synthesized via a microemulsion method and coated with a surfactant that renders them water soluble. Non-commercial nanoplastic particles prepared in the laboratory also typically fall under this category. This modification is apparent since the PS or PE “latex beads” are dispersed in an aqueous colloidal suspension rather than flocculating. Unfortunately, it is often difficult to obtain information on these components from the suppliers because they are considered “proprietary.” In addition to the inert hydrocarbon composition, the lack of water-soluble functional groups such as ⁻OH, COOH, NH₂, and SO₃ also means the particles cannot be covalently conjugated to fluorescent proteins. Therefore, dyes or fluorescent molecules must be adsorbed onto the surface, but in a reversible, fluid and dynamic way. Alternatively, the plastic particles may be engineered with surface functional groups for conjugation via a much stronger covalent linkage. For this reason, many commercial nanoparticles are available with modified surfaces containing functional groups for further functionalization/conjugation to the desired moieties. However, these additional functional groups are not typically present on environmental nanoplastic particles, especially those arising from natural degradation of bulk waste. Since nanoparticle surface coatings cause changes in solubility and aggregation behavior and reactive oxygen species (ROS) generation (Li et al., 2013) as well as differences in cell uptake and overall toxicity via multiple mechanisms including ROS generation and triggering of apoptosis (Feng et al., 2018), toxicological studies using engineered surfaces must account for this.

4.1. Nanoparticles may invalidate assays

As interest in nanoparticle development increased from the early 2000s through the present, more data became available to suggest that nanoparticle toxicology experiments cannot be carried out using standard methodology. Obviously, fluorescent or dye-labeled particles in a colorimetric or fluorometric assay would be expected to spectrally overlap in many cases, resulting in signal enhancement, or shift, leading to misinterpreted results. Erroneous measurements have been reported even for non-fluorescent particles due to unanticipated, complex interactions between nanomaterials and laboratory reagents, and nanoparticle interactions with light (Doak et al., 2009). These interactions could be altered by the addition of surfactants on the surfaces of nanoparticles, which are known to complicate the surface chemistries (Belyanskaya et al., 2007). Carbon-based nanomaterials, particularly carbon nanotubes, interact with a wide range of the dyes and fluorophores present in various assays as well, including 3-(4,5-dimethylthiazole-2-yl)-2,5-biphenyl tetrazolium bromide (MTT) (Wörle-Knirsch et al. 2006; Casey et al. 2007), neutral red, Alamar blue, 2-(4-iodophenyl)-3-(4-nitrophenyl)-5-(2,4-disulfophenyl)-2H-tetrazolium, monosodium salt (WST-1) and Coomassie blue assays (Casey et al., 2007). Carbon-based nanoparticles can adsorb proinflammatory cytokines, thereby rendering them undetectable by assays, and subsequently leading to a false underestimation of the inflammatory response (Brown et al., 2010; Kocbach et al. 2008).

The reliability of assays used to investigate nanoparticle toxicity have been called into question (Doak et al., 2009; 2012; Monteiro-Riviere et al., 2009; Han et al., 2011; Love et al., 2012; Darolles et al., 2013). For example, methods for determination of ROS production by nanoplastics often rely on fluorescent dyes. Interference of certain nanomaterials with the fluorescent probes must be determined on a case-by-case basis to ensure valid results. Since nanoplastics are also carbon-based and have high surface energies, a wide variety of substances can adsorb onto their surfaces; therefore, polymeric nanoparticles (nanoplastics) may also to interact with fluorescent probes in ROS detection kits in a similar fashion. These interactions may invalidate the results of the test, so it is strongly suggested that cytotoxicity findings be verified with two or more independent testing systems, and assays should be specific to the type of material being investigated (Wörle-Knirsch et al. 2006; Monteiro-Riviere et al., 2009). Preliminary experiments using the nanomaterial and the assay components alone, in the absence of live cells, should be performed to rule out artefacts (Doak, et al., 2009). Considering carbon-based nanomaterials, clonogenic assays are an option for avoiding these potential problems. By including colony size as an endpoint, the effects on cell viability and cell proliferation can also be monitored (Herzog et al. 2007), while avoiding the use of colorimetric or fluorescence assay.

4.2. Surfactants

Although surfactant coating of nanoparticles to ensure their stability in aqueous solution may alter the toxicological properties, in the absence of surfactants, solvents or dispersant, the NPs will aggregate and flocculate, also resulting in the manifestation of different toxicological effects (Oberdörster et al., 2006; Smith et al., 2007). Sodium dodecyl sulfate (SDS) is a commonly used surfactant. It has been used to disperse carbon nanotubes in aqueous solution due to its low toxicity to fish (Abel, 1976; Ham et al., 2005; Smith et al., 2007). Lee et al. (2019) and Bergami et al. (2020) both reported that Tween20 was the surfactant used by the manufacturer and Bergami et al. (2020) details the surfactants Tween20 and SDS with sodium azide and their concentrations in the supplementary information section. They note that previous studies on the additives/stabilizers are likely negligible, citing works by (Della Torre et al. 2014; Bergami et al., 2017; Manfra et al. 2017; Bellingeri et al. 2019) in which particles with and without stabilizers can be compared, citing studies investigating the toxicity of SDS (Mariani et al. 2006; Libralato et al. 2016) and sodium azide (Heinlaan et al. 2020; and Sleet and Brendel, 1985). (Cole et al., 2020) used the procedure from (Cole and Galloway 2015) to remove preservatives such as azides from their stock solutions.

Despite these examples, many reports do not provide information regarding surfactant compositions or concentrations. A known surfactant can be tested for toxicological effects alone as an additional control. However, once again, the toxicity of “proprietary” or otherwise unknown solvents/surfactants/formulations cannot be independently determined. Our literature search revealed that over 85% of the papers reviewed used commercial nanoparticles with unknown or unreported surfactants for their experiments.

4.3. Surface functionalization and toxicity

PS beads with carboxyl (COO⁻), amine (N:), or sulfate (SO₄²⁻) functional groups are used alone or in combination with fluorescent proteins or dyes to track the movement of the plastic particles in vitro or in vivo. It has been reported that amine modified PS NPs exhibited high toxicity in macrophage (RAW 264.7) and human bronchial epithelium (BEAS-2B) cell lines via apoptotic (caspase activation) and necrotic pathways (Xia et al., 2008). Positively charged surfaces facilitate particle-cell surface electrostatic interactions (Nel et al., 2006). Thus, these cationic polymers can bind indiscriminately to cell surfaces and negatively charged DNA, where they form complexes with DNA (Nafee et al., 2009). Due to this toxic effect, cationic polymeric nanoparticles have limited use in biomedicine, however, this property is exploited for gene delivery applications (Xu and Szoka 1996; El Ouahabi et al., 1995). Amidine (CH(=NH) NH₂) functionalized PS particles have been reported to be approximately 3-times more toxic to D. magna than carboxy-PS was (Saavedra et al., 2019). The toxicity of amine-groups is also confirmed by (Agashe et al., 2006; Chauhan et al., 2009; Della Torre et al., 2014; Nasser and Lynch 2016), and the shielding of these groups has been shown to reduce toxicity (Dutta et al., 2008; McNerny et al., 2010).

Many studies on the toxicity of PS beads have used amine-modified beads (Gonzáles-Fernández et al., 2018; Manfra et al., 2017). It is likely that the observed toxicity in these studies can be attributed to the surface amine groups and the positive charge rather than the polystyrene material itself. A recent study on PS in C. elegans reported the highest particle toxicity for positively charged (aminated) particles followed by uncharged (unfunctionalized) particles, and negatively charged (carboxy-modified) particles were the least toxic (Schultz et al., 2021). Therefore, it is possible that studies on carboxy-modified beads used as a model for environmental PS, may actually underestimate toxicity. Although nonphagocytic cells interact more strongly with positively charged particles, phagocytic cells exhibit a more enhanced interaction with negatively charged particles which has been attributed to charge similarities with bacterial cells (Fröhlich, 2012), which are also often negatively charged.

4.4. Particle Morphology

Thus far, all of the commercially available polymeric particles are spherical. Although particle morphology has been suggested to play a role in nanotoxicology (Peng et al., 2011), the vast majority of nanoplastics toxicology experiments still use spherical particles. Shape-dependent toxicity has been investigated for many nanoscale materials (Petersen and Nelson 2010; Ispas et al., 2009; Chithrani et al., 2006; Hamilton et al., 2009). High aspect ratio nanoparticles, like asbestos, have an elongated fibrous morphology that can be attributed to reduced particle clearance and carcinogenesis (Fubini et al., 2011). Asbestosis is a well-known example of a high aspect ratio particle inducing specific cytotoxicity (Fubini et al., 2011). However, a study on gold nanorods reported that surface modifications dramatically decreased their toxicity and found that coating was a bigger indicator of toxicity than aspect ratio (Wan et al., 2015). Therefore, both aspect ratio and surface attributes should be considered.

A study comparing polyethylene microplastics with different morphological features (rough and smooth surfaces), found that higher concentrations and rougher structures were associated with toxicity to immune and non-immune cells, triggered proinflammatory cytokine release and hemolysis, whereas the particles with smooth surfaces did not exhibit severe cytotoxicity (Choi et al., 2020). At present, studies on high-aspect ratio plastic particles and particles with porous or rough surfaces are significantly lacking. Of our literature survey, 93% of the nanoparticle toxicity studies used spherical particles. This is likely due to the commercial availability of spherical nanoplastic particles and the simplicity of synthesizing particles via microemulsion methods. It is highly unlikely that nanoparticles arising in the environment as a result of weathering will have a spherical morphology, or a smooth surface. However, it is possible that some synthetically produced spherical particles may enter the environment from improper disposal and cosmetic products, for example.

4.5. Fluorescence Leaching

Another attribute of the commercially available plastic nanoparticles is their availability in a variety of fluorescent colors and dyes. Although this may provide a convenience to researchers, additional complications may be introduced. While numerous studies use synthetic plastic nanoparticles conjugated to fluorescent molecules for tracking, some of these study results have been called into question due to the confirmation of fluorescence leaching from the particles (Schür et al., 2019). It is well known that adsorbed fluorescent or dye molecules can detach and result in artefacts (Kettiger et al., 2013; Rothen-Rutishauser et al., 2014). Many studies report biodistribution based on fluorescence tracking. However, the potential for fluorescence leaching is not commonly addressed, nor is the fluorescent dye typically confirmed to remain associated with the NPs.

Although it is possible to synthesize plastic with surface functional groups or perform surface modifications introducing surface functional groups for conjugation, most commercial sources for fluorescent beads do not provide information on their functionalization, conjugation, or adsorption schemes. For example, were the fluorescent molecules covalently bound, or are they simply adsorbed onto the surface? Unclear functionalization methodology invites an additional level of uncertainty regarding the stability of the fluorescently labeled plastics and the potential for leeching must be experimentally determined. It is also impossible to determine any role of the conjugation or adsorption scheme on toxicology when it has not been described. For example, polystyrene must be nitrated with nitric and sulfuric acid, for the addition of NO₂ to the ring (Booth 2012), followed by reduction to an NH₂ for conjugation. Other methods include the introduction of sulfate or carboxylate groups. These functional groups are not present on most polystyrene waste and how their addition alters the toxicology profiles remains unclear.

Bhargava et al. (2018) has developed a unique solution to fluorescence leaching by synthesizing PMMA particles and encapsulating hydrophobic perylene tetrabutyl ester (PTE) dye molecules within the PMMA composite. However, many studies, including the more recent ones, do not address this potential problem. In the absence of suitable electron microscopy techniques, fluorescent NPs are necessary for particle distribution studies. However, it is crucial that the particles, and not the detached or leeching fluorophores or dyes are being detected. For this task, computer biodistribution models may prove to be especially helpful. Other alternative models, such as plastic encapsulation of fluorescent or dye molecules and SEM or TEM equipped with energy dispersive x-ray spectroscopy (EDS/EDX) capabilities, are appropriate tools for verification of biodistribution.

4.6. Relevance of engineered particles

The fundamental shortfalls associated with using engineered particles as a model will not necessarily negate the previous studies performed using these commercial particles since some of the surface attributes (carboxyl groups, for example) could potentially mimic an environmental particle with biologicals adsorbed onto the surface. For example, it has been reported that, in natural river water, the surface charge of styrene nanoplastics changes from positive to negative, and the particles form large heteroaggregates at the isoelectric point (Oriekhova and Stoll 2018). It is also clear that the reported zeta potential measurements for “positively charged” (ie. Amine-modified) or “negatively charged” (ie. carboxyl-modified) particles are inconsistent because they depend on the characteristics of the coating (surface) materials and, not just the amine or carboxyl functional groups, in combination with the solvent interactions at the interface (see Table 1). It is important to identify trends in the data by comparing results in various solvents, at a range of pHs, temperatures, ionic strengths, protein concentrations, etc., and investigate the detailed interfacial phenomena directing interparticle, particle-solvent, particle-protein, and nano-bio interactions in dynamic systems.

Table 1.

Zeta potential values corresponding to coagulation and colloidal stability (Nimesh et al., 2017).

Average Zeta Potential (mV)*	Colloidal Stability/ Particle Behavior in Solution
0 to |5|	Maximum Rapid Coagulation or Flocculation
|10| to |30|	Incipient instability
|30| to |40|	Moderate Stability
|40| to |60|	Good Stability
above|60|	Excellent Stability

^*Zeta potential is presented as absolute value and could correspond to either a positive or negative value.

5. Zeta Potential

The zeta potential measurement, although widely used in the field of nanoparticle engineering, is less commonly used in the field of environmental science. Since there are multiple aspects regarding obtaining and interpreting the zeta potential measurements, it is important that environmental science researchers in the field of nanoplastics familiarize themselves with this type of nanoparticle characterization. Lowry et al (2016) prepared a detailed report on the use of zeta-potential measurements for environmental nanotoxicity which provided specific guidance on the influential parameters for obtaining accurate measurements and providing sufficient detail regarding obtaining and reporting repeatable measurements. Since plastic polymers are typically inert, it is likely that they are uncharged upon initial release into the environment, however, these particles may acquire charges over time, arising from the environmental conditions, interactions with components of water and soil, and biofouling. Additionally, the nanoparticles used for toxicology testing and biodistribution studies in the laboratory are engineered with surface functional groups, which will introduce surface charges. Changes in the surface chemistries determine how the particles will behave in solution, including their colloidal stabilities. For these reasons, zeta potential measurements in environmental nanotoxicity studies are necessary.

Zeta potential provides information regarding the potential at the hydrodynamic shear boundary, the layer that exists between the solid surface of the particle and the fluid in which velocity (fluid flow) gradients and shear stresses are present and is a measure of colloidal stability (Smith et al., 2017). In solution, an electrical double layer, composed of ions, surrounds charged particles. This layer is composed of two parts, the Stern layer and the diffuse layer (Fig. 2). The region closest to the particle (Stern layer) is where the ions are tightly adsorbed to the surface. The diffuse layer is the region where ions are less firmly associated with the surface. In the diffuse layer, a notional boundary exists, in which the particle behaves as a single entity. It is at this boundary is that zeta potential is measured. Zeta potential describes nanoparticle stability in different solvents (water, blood, saltwater, etc.), which in turn controls time in circulation, interactions with proteins, cell permeability to nanoparticles, and toxicity (Alexis et al., 2008; Albanese et al., 2012; Kamaly et al., 2012).

Fig. 2.

Zeta potential measurements and electrical double layer on surfaces of negatively charged (left) and positively charged (right) nanoparticles.

We collected zeta potential values reported for a variety of polystyrene beads to compare their values in freshwater (FW), saltwater (SW), and media. We plotted those reports as a function of zeta potential and particle size (Fig. 3.). We have grouped the data according to surface functionalization as follows, aminated PS (APS), carboxylated PS (CPS), fluorescent PS (FPS, for all colors), and unfunctionalized (PS). The term “unfunctionalized” here refers only to the addition of functional groups and/or fluorophores and not surfactants. We obtained average zeta potential values and standard deviations (stdev) for PS as follows: APS in FW +34.23 mV (stdev = 23.96 mV), APS in SW +7.83 mV (stdev 15.14 mV), APS in media +18.80 mV (stdev 30.96 mV); for CPS in FW −40.70 mV (stdev 30.26 mV), for CPS in SW −13.80 mV (stdev 0.8 mV), for CPS in media −52.8 mV (stdev 5.74 mV); for FPS in SW −8.63 mV (stdev 5.12), FPS in FW −36.32 mV (12.93 mV), FPS in media −20.27 mV (12.58 mV); and PS in SW −25.27 mV (stdev 7.99 mV), for PS in FW −43.73 mV (stdev 24.56 mV), PS in media −29.65 mV (stdev 10.39). All fluorescent particles were combined into a single category and not separated by color or functional group used for conjugation, all “media” was combined into a single category which was not specific to a particular media type, the freshwater category includes fresh aquarium water, milliQ and deionized water. Despite these limitations, the zeta potential values are comparable, with a typical positive trend for APS in all solvents and an overall negative trend for all the other particle types. The highly negative zeta potential values and aqueous colloidal stability for “unfunctionalized’ PS is a result of the surfactant(s) on the particle surfaces. It is likely that the APS, CPS, and FPS particles also have a surfactant on their surfaces which may be contributing to these observed values.

Since zeta potential describes the solvent-solute (NP) interactions at the interface, multiple factors will contribute to zeta potential measurements. Temperature, salinity, hydrophobicity, conductivity/ionic strength, pH, viscosity, and protein concentrations all affect the measured values for zeta potential. Saavedra et al., (2019) reported zeta potential values from pH 3.0 to 11.0 and plotted them along with hydrodynamic size. They determined that the amine-nanospheres were stable and remained dispersed in the pH-range relevant to environmental values (7.0–9.0) (Saavedra et al., 2019). As zeta potential may also affect the formation and size distribution(s) of the aggregates, (Manfra et al., 2017) demonstrated that aggregation patterns and toxicity with varied differently with surface charges of PS beads. They reported the formation of microscale aggregates associated with the negative zeta potential particles and no mortality in rotifers, while positive zeta potential particles formed nanoscale aggregates leading to mortality of the rotifers. More work is needed to determine the roles the degree of aggregation in combination with pH and surface charges play in systemic toxicity.

6.1. Agglomeration

Surfactant coating of nanoparticles reduces agglomeration. However, hydrodynamic size measurements, prior to experimentation are often unreported in the literature. Electron microscope images do not capture the behavior of the particles in solution, rather they image what is typically a dry monolayer of stationary particles on a grid. Hydrodynamic size measurements are more accurate in determining what particle size the cells or organisms will truly encounter. It is well known that nanoparticles tend to agglomerate more at certain pH values, in high ionic strength (electrolyte content) solutions, and the presence of proteins can influence this behavior (French et al., 2009; Chambers et al., 2014; Bizmark and Ioannidis 2015; Lin et al., 2017), by stabilizing or destabilizing the nanoparticles through steric or electrostatic effects. Protein corona formation can counter ionic effects under physiological conditions and stabilize nanoparticles, thereby reducing agglomeration via steric interactions (Kittler et al., 2010). Therefore, a relevant solvent should be used when measuring hydrodynamic size.

Typical laboratory protocols for nanoparticle engineering and characterization call for sonication prior to size distribution measurements and imaging to reduce aggregation and obtain accurate measurements. However, many toxicology studies do not mention whether sonication is included in their protocols. Even nanoparticles specifically engineered with an exceptional colloidal stability in a particular solvent; having favorable interfacial interactions with the solvent molecules, can still form agglomerates. It is possible that many experiments, though reporting toxicity for particles in manufacturer’s reported size range, may actually be observing the toxicity of larger agglomerates, or collective toxicity of agglomerates and single particles. Vaz et al. (2021) reported a 2-fold reduction in acute toxicity to the D. magna after sonication of PS NPs vs non-sonicated NPs. Agglomeration can alter the toxicological profiles. Another study reported that the PS beads demonstrated aggregation patterns in high ionic strength media, with anionic beads forming larger aggregates (1000 nm) and cationic beads remaining small (~100 nm) (Manfra et al., 2017). Enhanced toxicity was observed for the particles that remained smaller.

Sonication will certainly reduce or eliminate agglomerates at the time of treatment, however, how long this last is still uncertain. Some researchers report sonication of their working suspensions of PS nanoparticles. (Sun et al., 2021a) reports sonication at 40 kHz (100 W) for 30 min. (Bergami et al., 2020) report sonication using CEIA CP316, at maximum power (600 W), nominal frequency 40 kHz, amplitude 90%, for 2 min. (Shao et al., 2019) reported a 30 min sonication time, but did not provide further details. Zhao et al. 2021 reported that, after sonication, aggregates did not form for at least two days. The authors do not disclose the properties of the solvent in which the size measurements were taken. Despite this, it is possible that agglomerates will still form over time. Another study measured agglomeration of the PS particles over time and reported size increases from 110 nm at t=0, 129 nm at 6 h, 240.4 nm at 24 h and 345.5 at t > 24 h (Brandts et al., 2018a).

Clear, time-dependent analysis of the agglomerate formation after sonication is necessary. An additional parameter to consider when performing hydrodynamic size measurements is the selection of the same solvent for the measurement as the experiments will be performed, if possible. For in vitro or cell work, the most accurate measurements will be obtained if the measurements should ideally be taken in the growth media. For fish or marine invertebrate, for example, the measurements should be performed in salt or freshwater with parameters identical to those used to expose the organisms.

7. Future Directions

Numerous reports have confirmed toxicological effects of nanoplastic particles to cells and organisms. Many studies investigate acute or shorter term chronic exposures and make observations such as overall mortality (Manfra et al., 2018; Liu et al., 2018; Lee et al., 2019), bioaccumulation of particles (Pitt et al., 2018; Kang et al., 2021; Sendra et al., 2021; Xu et al., 2021, effects on growth (Zhu et al., 2018) oxidative stress (Sun et al., 2012b; Estrela et al., 2021; Guimarães et al., 2021b), inflammation (Sun et al., 2012b; Peng et al., 2021), effects on reproduction (Kwak and An 2021; Sun et al., 2021a; Zhu et al., 2018), liver function (Lu et al., 2018), and neurotoxicity (Brandts et al., 2021a; Guimarães et al., 2021a). The effects of true chronic, low-dose exposures are unknown. Future work should focus on mechanistic studies and investigations into potential carcinogenicity.

More information on the complexity and the role of eco and protein coronas is rapidly becoming available, and corona formation has been shown to influence toxicity (Tenzer et al., 2013; Lee et al., 2015; Westmeier et al, 2016; Shultz et al., 2021). Future work should fully characterize corona formation over time, as it is a dynamic process that greatly impacts the toxicological effects. Since the corona can give synthetic materials a biological identity (Monopoli et al, 2012), changes in the geometry and size of and the introduction of steric hindrance to nanoparticles as a result of modifications made by adsorbed proteins and enzymes could be strong areas of future study. Additional investigations into the role of the surface charge on corona formation and fluidity are also necessary, especially since particles with a wide variety of charges have been used in experiments (Saavedra et al., 2019). Additionally, noting that the numerous contributions of surface interactions with proteins are significantly linked to the overall interactions at biological interfaces, some gaps in research become evident.

Since surface interactions between the inert or engineered particle can undoubtedly affect particle-protein interactions, and the subsequent configuration of the protein corona, the full scope of surface charges and toxicological observations cannot be realized without a thorough understanding of these interactions and their biological consequences. Differences in the ways in which engineered particles with chemically-modified surfaces vs. environmental nanoparticles, possibly enveloped by eco-coronas, interact with proteins and the potential to cause protein misfolding is unclear. Studies are necessary to determine the role of these coating agents on the end fates of nanoplastics in the environment and within the bodies of organisms.

Conclusions

As awareness increases regarding nanoscale plastics as environmental pollutants, toxicological investigations follow suit. To adequately model environmental nanoplastics, studies should ideally encompass the parameters relevant to environmental pollutants. Environmental weathering, surface roughness, eco-corona formation, and environmental fates of nanoplastics must be investigated further. There currently exists a wealth of literature on the toxicology of spherical polystyrene nanoparticles with engineered, or surfactant coated surfaces. Research on other pollutant polymers of interest, such as PE, PVC, PET, PA, and PP are significantly lacking. It is also concerning that polystyrene nanoplastics appear to demonstrate significantly more toxicity than all other nanoplastics (Yang and Nowack 2020).

It is well-established in the literature that amine modified PS is more toxic than carboxy-modified or unfunctionalized beads. Often the introduction of these functional groups is performed to covalently conjugate a fluorophore for tracking; however, at least one study used unconjugated (non-fluorescent) amine-modified PS beads for their study (Manfra et al., 2017). The study provides strong evidence to establish the contribution of surface functionalization on the observed toxicity, which can be attributed to the presence of the amine functional groups rather than the fluorophores or the polystyrene itself.

Incorporation of chemical modifiers, the use of surfactants, the stability of fluorescence, conjugation schemes for fluorescent molecules, and the effect of fluorescent molecules on toxicological profiles remains unclear. It is possible that some aspects of toxicity could be attributed to the synthetic engineering method, the presence of chemical toxins, unreacted reagents, solvents, or other residuals from the synthetic process. It is important to identify potential laboratory contaminants, or chemical residues that may ensue due to chemical engineering and functionalization of these particles, which may or may not be present in actual environmental samples. In fact, it has been reported that some of the toxic effects originally attributed to nanoplastics were truly caused by additives, specifically sodium azide, and toxic synthesis residues (Pikuda et al., 2018). Another study performed additional exposures using the supernatant alone, with nanoplastic particles removed to determine whether the toxicity could be attributed to the presence of surfactants, synthesis residues, or other contaminants in the dispersion media (Schultz et al., 2021). Reproduction reduction was observed with nanoplastic exposure for commercial particles from Magsphere (Pasadena, CA, USA), but not with the supernatant alone, and therefore, they were able to confidently attribute the knockdown effects to the nanoplastic particles and not contaminants in the supernatant (Schultz et al., 2021). However, this simple additional control experiment is not routinely performed.

It is well-established in the literature that surface properties dictate the interactions at nano-bio interfaces, and the majority of in vitro and in vivo nanoplastics toxicology studies use nanoparticles with modified surfaces. There are, in fact, numerous examples that continue to support the idea that different functionalizations which impart different surface charges have a significant effect on particle toxicity. The materials used in the model with the toxicological findings, zeta potential (when measured), manufacturer’s reported size compared to the measured hydrodynamic size (when reported) are summarized in Table 2.

Table 2.

Summary of Some Notable, Recent Toxicological Studies on Plastic Nanoparticles

NPs†	Manufacturer	Measured Size (nm)	Animal Model	Concentration	Zeta Potential (mV)	Observations	Reference
APS, 50 nm	Bangs Lab. Inc.	in mQ: 50, in ASW: 200.3	Mussel (Mytilus galloprovincialis) hemocytes	1–50 mg/mL	43±1 in mQ, +14.2 in ASW	increased cellular and lysosomal damage, ROS production, protein corona observed and governs NP behavior in organism	Canesi et al., 2016
GF CPS, 45 nm	Bangs Lab. Inc.	in water: 58.48±1.8, in NSW media: 998±67, in RSW media: 1109±128	Rotifer (Brachionus plicatilis)	0.5–50 mg/L	in mQW: −51.3±1.17, in NSW: −12.4±2.5, in RSW: −10.7±3.1	no toxicity	Manfra et al., 2017
APS, 50 nm	Bangs Lab. Inc.	in water: 54.02±1.41, in NSW media: 108±13, in RSW media: 93.99±11	Rotifer (Brachionus plicatilis)	0.5–50 mg/L	in mQW: +57.6±2.0, in NSW: +22.2±2.0, in RSW: +16.8±2.11	high mortality	Manfra et al., 2017
aqPS, 100 nm + TiO2 NPs	PS from Janus New-Materials Co. (Nanjing, China)	108.2±4.5 nm (PS) Solvent parameters unk	Nematode (Caenorhabditis elegans)	0.01–1 μg	Unk solvent: −9.698±0.966	enhanced toxicity of TiO2, oxidative stress, ROS production, decreased locomotion	Dong et al., 2018
aqPS, 75 nm	BaseLine Chromtech Research Centre, Tianjin, China	in FW, by TEM: 71.78, by DLS 85.86	Water Flea (Daphnia pulex)	0.1–1 mg/mL		LC50 varied with age group, lowest LC50 was 80.02 mg/mL, decreased energy	Liu et al., 2018
GF CPS, 100 nm	Micromod Labs (Germany)	in mQW: 131.82 ± 0.23, in FSW: 140.11 ± 0.92	Oyster (Crassostrea gigas) gametes	0.1–100 mg/mL	in FSW: −2.89 ± 7.21 mV, in OM: −8.17 ± 4.66, in SM: −13.60 ± 2.92	dose-response increase in ROS production, NPs adhere to spermatozoa	Gonzales-Fernandez et al., 2018
APS, 100 nm	Micromod Labs (Germany)	in mQW: 130.83 ± 0.32, in FSW: 141.13 ± 1.50	Oyster (Crassostrea gigas) gametes	0.1–100 mg/mL	in FSW: −5.69 ± 3.65, in OM: −8.26±3.20, in SM: −6.95 ± 1.37	adhesion to spermatozoa, increase in cell size and complexity	Gonzales-Fernandez et al., 2018
aqPS 5–5.9 μm and 50–100 nm	MACKLIN® Shanghai, China	by SEM/TEM: 5.69±1.71 μm and 93.19±5.463 nm; by DLS: 106.3 nm (nano PS in mQ w/son.)	Silkworm (Bombyx mori)	10 μg/mL was applied to leaves (food)		accumulation in the gut, immunosuppression, oxidative stress	Muhammad et al., 2021
PS, 110 nm	prepared by ME	by TEM 110 ± 6.9 nm, by DLS in SW 129 nm, at 6h 340.4 ± 73.6 nm, at 24h 345.5 ± 49.4 nm	Mussel (Mytilus galloprovincialis)	0.05 mg/L PS with 6.3 μg/L carbamazepine	−43.0 mV, in SW −16.2 mV	Decreased enzymatic activity, effects on neurotransmission, lipid peroxidation, inhibition of cholinesterase	Brandts et al., 2018a
PMMA, 45 nm**	prepared by ME	by TEM: 45 nm, by DLS: 40 nm, in ASW: 58.6, 97.3 at 24h, 120.3 at t>34h	Sea Bass (Dicentrarchus labrax)	0–20 mg/L for 96h	in UP: −26.4 mV	altered in liver, molecular signaling	Brandts et al., 2018b
FPS, 50 nm	Polysciences Co. (Warrington, PA, USA)	in UP: 47.0±0.2	Zebrafish (Danio rerio)	1 mg/mL for 3d	in UP: −30 mV	Nanoplastics enhanced BPA uptake, neurotoxic biomarker expressions upregulated	Chen et al., 2017a
PTE-loaded F PMMA, 185 nm***	prepared by Nppt	in DI: 185±3	Acorn barnacles (Amphibalanus amphitrite)	25 ppm for 3h, 1 ppm throughout growth cycle	in DI: −35.1±0.4 mV	PMMA nontoxic up to 25 ppm particles were ingested and retained in the body for all stages of development	Bhargava et al., 2018
PS 35 nm and APS 35 nm	synthesized at Shandong Univ. China	by DLS: 34.8±3.3 (PS) and 35.3±2.8 (APS)	Nematode (Caenorhabditis elegans)		−19.7 ± 1.78 (PS) and −25.2 ± 0.97 mV (NPS)	transgenerational reproductive toxicity, enhanced toxicity of APS, germline apoptosis	Sun et al., 2021a
FCPS, 0.03 μm	Sigma Aldrich	by EM: 23.03 ± 0.266	Grass Carp (Ctenopharyngodon idella)	0.04 ng/L to 34 μg/L		formation of micronuclei and other nuclear abnormalities, DNA damage, oxidative stress, and reduced antioxidant defenses	Guimarães et al., 2021b
PS, 50 nm and 60 nm	Magsphere, Pasadena, CA, USA	PS by TEM: 50±12, 64±18, DLS 51.8±0.3, 69.3±0.8 (in exposure media);	Nematode (Caenorhabditis elegans)	0, 1, 2.8, 7.1, 18.8, and 50 mg/L	in media: −22.3±29.4, −37.0±14	EC50 18.1 m/L	Schultz et al., 2021
APS, 50 nm and 60 nm	Magsphere, Pasadena, CA, USA	TEM: 50±10, 61±9, DLS: 51.0±4, 59.8±0.6),	Nematode (Caenorhabditis elegans)	0, 1, 2.8, 7.1, 18.8, and 50 mg/L	in media: +38.6±18.4, +51.8±17.0	eco-corona reduced toxicity, EC50 2.6 mg/L	Schultz et al., 2021
CPS, 50 and 60 nm	Magsphere, Pasadena, CA, USA	by TEM: 49±7, 63±20, DLS: 50.8±0.3, 64.9±0.7	Nematode (Caenorhabditis elegans)	0, 1, 2.8, 7.1, 18.8, and 50 mg/L	In media: −59.4 ± 17.8, −49.0 ± 11.1	minimal toxicity even at the highest tested concentration (200 mg/L)	Schultz et al., 2021
APS, 200 nm	Invitrogen (Thermofisher scientific, USA)	in UP at pH 8: 226.4±8.6	Water flea (Daphnia magna), Rotifer (Bachionus calyciflorus), and beaver-tail fairy shrimp (Thamneocephalus platyurus)	10 to 400 mg L−1	in UP at pH 8: +50 mV	APS is more toxic than CPS, eco-corona reduced toxicity	Saavedra et al., 2019
CPS, 200 nm	Invitrogen (Thermofisher scientific, USA)	in UP at pH 8: 220.1±9.1	Water flea (Daphnia magna), Rotifer (Bachionus calyciflorus), and beaver-tail fairy shrimp (Thamneocephalus platyurus)	11 to 400 mg L−1	in UP at pH 8: −50 mV	APS is more toxic than CPS, eco-corona reduced toxicity	Saavedra et al., 2019
F CPS, 0.03 μm	Sigma Aldrich, USA. Product n. L5155	by EM: 23.03±0.266	Grass Carp (Ctenopharyngodon idella)	760 μg/L for 3d		affected response to mirror test, inactivity towards alarm substances and DNA damage, oxidative stress	Estrela et al., 2021
RF PS, 100 nm	Janus New-Materials Co. (Nanjing, China)	by DLS: 102.55 ± 3.8 nm	Nematode (C. elegans)	1–100 μg/L	PS-NPs: −9.143 ± 0.258 mV	alteration in expressions of four intestinal long non-coding (lncRNAs), demonstrated the role of lncRNAs in intestinal barrier to mediate a protective response to PS-NPs exposure	Zhao et al., 2021
GF and RF PS, 50 nm and 500 nm and 5000 nm	Magsphere, Pasadena, CA, USA	by SEM: 50.7, 503.6, and 5047.0 nm	Mouse	2.5–500 mg/kg body weight		combining particle sizes increased toxicity, intestinal barrier dysfunction by reactive oxygen species (ROS) mediated epithelial cell apoptosis	Liang et al., 2021
GF PS NPs, 0.03 μm	Sigma Aldrich L5155		Dragonfly (Aphylla williamsoni)	34 μg/L in exposure water for 48 h		PS NPs cause REDOX imbalance and neurotoxic effect	Guimarães et al., 2021a
PS NPs and F PS NPs, 100 nm and 200 nm and 500 nm and 700 nm	Goose Technology Co, Ltd. Tianjin, China		Cucumber plants	50 mg L−1		distribution throughout entire plant, including fruits, size dependent effects, protein content increased and Mg, Ca, Fe decreased	Li et al., 2021
PS, 100 nm	Thermo Fisher Scientific, USA	no notable change in size and polydispersity	Tilapia (Oreochromis mossambicus)	20 mg/L for 7 days		PS NPs cause abnormal glycolipid metabolism, possible inflammatory response	Pang et al., 2021
GF PS and CPS and APS, 100 nm	Tianjin Baseline ChromTech Research Centre (Tianjin, China)		Mouse	100 μL PS, CPS and APS (10 mg/mL) by gavage once a day (1 mg/day)		systemic organ accumulation after oral exposure, hematological system injury and lipid metabolism disorder	Xu et al., 2021
PE in liquid suspension	Capital Medical University	in UP: 191.10 ± 3.13 nm	Zebrafish (danio rerio)	25, 50, 100, 200, 400, 600, 800 and 1000 μg/mL	in UP: −48.98±3.25	pericardial edeema and cardiac toxicity in embryos, endothelial damage and hemodynamics alterations, oxidative stress, systemic inflammation	Sun et al. 2021b (chemosphere)
PS, 45 μm and 50 nm	Polysciences Co. (Warrington, PA, USA)	in UP: 41,011 ± 420 nm and 47 ± 0.2 nm	Zebrafish (danio rerio)	1 mg/L	in UP: −45 ± 0.2 mV and −30 ± 2 mV	developmental neurotoxicity of nanoplastics, lowered by co-exposure with microplastics	Chen et al., 2017b
APS, 52–330 nm	Bang laboratories (Fisher, IN, USA)		Algae (Scenedesmus sp.), Daphnia magna, Crucian carp, Carassius carassius	0.005 g/L to 0.150 g/L		trophic transfer observed, particles in brain of fish, changes in brain structure	Mattsson et al., 2017
PS NPs and F PS NPs, 42 nm	Bangs Laboratories, Inc. (Fishers, IN, USA)	PS 30.67 ± 8.97 nm, F PS 34.5 ± 10.8 nm	Zebrafish (danio rerio)	90, 45, and 120 mg/mL	PS −24.7 ± 2.93 mV, F PS −21.1 ± 2.47 mV	bradycardia, NPs translocated GI, liver, pancreas, uninflated swim bladders	Pitt et al., 2018
PS and FPS, 50 nm and 200 nm and 500 nm	Polyscience Inc. (USA)		Zebrafish (danio rerio)	0.1 mg ml−1	In DI: −62±8 (50 nm), −77 (200 nm), −60 (500 nm); in EW: −30 (50 nm), −45 (200 nm), −50 (500 nm)	exacerbated mortality, deformation, reduced hatching rate, subcellular damages, increased toxicity	Lee et al., 2019
NPS, 44 nm	Bangs Laboratories, Inc. (Fishers, IN, USA)	44.73 nm (hydrodynamic diameter), without 5% ERSE 129.9 nm	Zebrafish (danio rerio)	1 ppm, 10 ppm	−38.0 mV, without 5% ERSE −8.39mV		Trevisan et al., 2019
GF PS (internally dyed), 25 nm	Thermo Fisher Scientific (Waltham, U.S.), JRC Nanomaterials Repository	19 nm	Zebrafish (Danio rerio)	20 mg L−1 PSNPs for 2 days, 10 to 100 mg L−1		reduction in insulin expression domain size, hyperactivity dark challenge phase, decreased glucose levels,	Brun et al., 2019
PS and F PS, 50 nm and 1 μm	Polyscience, Inc,	in UP: 50 nm 57.5±0.15, 50 nm F 71.2±0.4, 1 μm 106 4±19.1, 1 μmF 1016±27.7	Zebrafish (Danio rerio)	10 mg·L−1 PS NPs for 5 h	in UP: 50 nm −31.9±3.2, 50 nm F −48.3±0.7, 1 μm −48.9±0.4, 1 μmF −37.7±0.07	Immunodeficient and infected larvae decreased survival, translocated from gut lumen to intestinal cells	Sendra et al., 2021
PS and F PS, 50 nm and 45 μm	Polysciences (Warrington, PA, USA)		zebrafish (D. Rerio); European sea bass (Dicentrarchus labrax); Oryzias melastigma, fed with brine shrimp (Artemia salina)	10 μg/mL for 24 hours		detected in gut; NPs stronger toxicity than MPs, may be species specific,	Kang et al., 2021
aq PS, 0.05 to 0.1 μm	Klamar Co., Shanghai, China		Enchytraeus crypticus	oatmeal-PS mixture was used, 0, 0.025, 0.5 and 10%		few PS entering worms; weight reduced; increased reproduction; decreased alpha diversity; stimulation in cocoon production; inhibition of growth	Zhu et al., 2018
PE and PLA and PPC	Zoomlion Plasticizing Ltd. (Changsha, China)		Earthworm (Eisenia fetida)	50% soil dry weight (0, 0.125, 1.25, 12.5, 125, 250 and 500 g kg−1)		Increased rates of earthworm avoidance; residual cow dung increased; biomass increased; less cocoons.	Ding et al., 2021
OF 180 PE and 250 PE, 180–212 μm and 250–300 μm	Cospheric (Santa Barbara, CA, USA)		Worm (Eisenia andrei)	1000 mg/kg dry soil		damaged male reproductive organs; decreased sperm density; fragmented nanoparticles; impaired spermatogenesis, arrangement of sperm bundles in seminal vesicles; inhibited cellular viability of coelomocytes	Kwak and An 2021
GF microbeads, 27–32 μm diameter	Cospheric LLC CA, USA—UVPMS-BG-1.025	31.5 μm (±7.6 standard deviation, S.D)	Antarctic krill (Euphausia superba)	402 or 1606 μg L−1		Antarctic krill were physically fragmenting beads after ingestion, whole beads were found in the stomach and midgut content, as well as fecal pellets.	Dawson et al., 2018
CPS NP, 62 and FPS 60 nm, APS 50 nm	Bangs Laboratories Inc.		Antarctic krill (Euphausia superba)	2.5 μg/mL	CPS −19.3 ± 1.62, APS +16.4 ± 3.47	active swimming behavior and a typical escape response to disturbances	Bergami et al., 2020
PS, 50 nm, and 500 nm and 2 μm; CPS and APS 50 nm	Polysciences/Bangs Laboratories	UW 2-μm 2681.0 ± 50.5, 500 nm 774.3 ± 29.3, COOH-50 55.9 ± 0.4, Plain-50 49 ± 0.4, NH2-50 53.3 ± 2.3	Pacific oyster (Crassostrea gigas)	100 μg/mL at T0 and T24h	UW PS 2-μm: −44.8 ± 0.9, PS 500 −67.8 ± 7.0, CPS-50 −62.1 ± 0.4, PS-50 −70.1 ± 1.4, APS 50 44.0 ± 1.5 In FSW: PS 2 μm −30.5±1.5, 500 nm −28.3±0.6, CPS 50 −13.8±0.8, PS 50 −31.3±4.4, APS 50 15.6±2.7	impairment of biological membranes, sub-cellular toxicity, or physical blockages,	Tallec et al., 2018
FPS 50 nm	Magsphere, Pasadena, CA, USA	(HDD) of 52.65 ± 2.88 nm (DLS), mean particle size was 34 ± 8.6 nm (TEM)	Mussels (Mytilus spp.)	500 ng mL−1 of 20 μm PS 50 nm PS for 24 h or 7 days	−30.9 ± 1.9 mV	Both PS-MP and PA-MF in digestive gland, induced lipid peroxidation in the digestive glands	Cole et al., 2020
PS, 0.5 and 50 μm	Microspheres-Nanospheres (New York, USA)		Mouse	100 and 1000 μg/L		decrease in body weight, including relative liver and fat weights, decrease in levels of serum TG, hepatic TG and TCG decreased, hepatic PYR increased, mucin secretion decreased,	Lu et al., 2018
PS, 75 nm	BaseLine Chromtech Research Centre (Tianjin, China)	71.18 nm (TEM), 85.86 nm (DLS)	Prawn (Macrobrachium nipponense)	0, 20, 40, 80, 160, 320, 640, or 1280 mg/L, for 96 -h and 0, 5, 10, 20, and 40 mg/L for 28d		Survival rate decreased with increasing NP concentration. Molting behavior was also affected. Activation of ROS defenses.	Li et al., 2020

APS: aminated polystyrene, ASW: artificial saltwater, aq: aqueous, CPS: carboxylated polystyrene, DI: deionized water, DLS: dynamic light scattering, EW: egg water, GF: green fluorescent, EM: electron microscopy, F: fluorescent, FW: freshwater, HDD: hydrodynamic diameter ME: microemulsion, mQ:milliQ water, NP: nanoparticle, Nppt: nanoprecipitation, NSW: natural seawater, OF: orange fluorescent, OM: oocyte media, RF: red fluorescent, ROS: reactive oxygen species, RSW: standardized reconstituted seawater, SM: spermatozoa media, son: sonication, SW: seawater, TEM: transmission electron microscopy, unk: unknown, UP: ultrapure water

Modified beads are always identifiable by the functional groups carboxy (COO⁻), amine (N:), sulfate (SO₄²⁻), or the addition of fluorescent molecules, proteins, or dyes. The presence of a surfactant or coating on the particles is not always obvious, nor is it typically reported. However, the presence of a surfactant can be inferred when hydrophobic polymeric particles appear as a colloidally stable, aqueous suspension. Lastly, the surface properties affect particle-protein interactions, subsequently influencing their ultimate fates.

The zeta potential of the particle in aqueous suspension can also provide information on the surface state of the particle and colloidal stability. Unmodified plastics, such as PS are not water soluble, will not form a stable colloid in water but will coagulate and flocculate. Rossi et al. (2014) published a simulation in which polystyrene chains, forming 7 nm agglomerates, readily penetrate lipid membranes. It is possible that surfactant modification of commercial PS may negate this transfixing hydrophobic interaction upon rendering these materials water-soluble. Another potential shortfall, apparent when comparing the experimental data, is limited information regarding detailed solvent parameters in which measurements were obtained. As described in Section 5, favorable or unfavorable solvent-solute interactions contribute to stability. Even though the hydrodynamic sizes are occasionally reported, the full solvent parameters, concentration, ionic strength, temperature, and time are not consistently described, which makes it difficult to determine whether agglomerates are present when the organisms or cells are exposed. Since exposure to multiple nanoparticle sizes simultaneously has been shown to alter cell uptake (Li et al., 2019), more investigations are necessary to determine to what extent the presence of agglomerates and single particles also influences uptake.

It is generally accepted that the toxicity and environmental effects (Klaine et al., 2008) of polymeric nanoparticles are highly dependent on several parameters such as size, morphology, aspect ratio (Poland et al., 2008; Takagi et al., 2008), hydrophobicity, surface area, porosity, surface charge (Schaeublin et al., 2011; Baek et al., 2011a; Baek et al., 2011b), material, concentration, and composition. These parameters correspond to reactive oxygen species (ROS) production capabilities, protein interactions (Hollóczki and Gehrke 2019), receptor interactions, dispersion, aggregation, phagocytic cell uptake, adsorption of molecules and compounds, interactions with DNA, and more. It is also necessary to determine whether interactions with cytokines, assay components, or other molecules, chemicals or light may invalidate results. Carbon-based nanomaterials are known to “hoax” scientists in assays (Wörle-Knirsch et al., 2006). Also, the combination of multiple particle sizes appears to alter cell-uptake compared to single size exposures alone (Li et al., 2019) and merits further investigation as it is unlikely that an environmental exposure will consist of only a single size range of nanoplastics.

Even if plastics production was halted immediately, it would still be necessary to develop countermeasures for the million tons of plastic waste that is already present in our environment. This plastic waste is constantly being degraded into nanoplastics. It is well established that nanomaterials have different properties from bulk materials. We know that nanoplastics can enter organisms and go systemic, potentially disrupting every organ system in the body and enter virtually every cell type. There is still no way to remove them from our water sources. The future implications of plastic pollution must govern policy before it is too late.

Highlights.

• Discussed the types of nanoplastics used in studies of environmental toxicology.

• Summarized reported zeta potential values for several polystyrene nanoplastics.

• Highlighted the limitations of each type of engineered nanoplastic particles.

• Reviewed the animal models for nanoplastic toxicity assessment.

• Summarized the previous studies on environmental toxicology of nanoparticles.

Abbreviations

PVC

polyvinylchloride

PE

polyethylene

PP

polypropylene

PS

polystyrene

PET

polyethylene terephthalate

PEP

polyethylene-polypropylene copolymer

PAK

polyacrylates

PA

polyamide

EVA

ethylvinyl acetate

PCL

polycaprolactone

PMMA

poly(methyl methacrylate)

CEA

carcinoma embryonic antigen

PCB

polychlorinated biphenyls

PAH

polycyclic aromatic hydrocarbons

DDT

dichlorodiphenyltrichloroethane

PBDE

polybrominated diphenyl ethers

BPA

bisphenol A

LDPE

low-density polyethylene

HDPE

high-density polyethylene

ABS

acrylonitrile-butadiene-styrene

SB

high-impact polystyrene

ROS

reactive oxygen species

SDS

sodium dodecyl sulfate

FW

freshwater

SW

saltwater/seawater

APS

aminated polystyrene

CPS

carboxylated polystyrene

FPS

fluorescent polystyrene

RSW

artificial media/standardized reconstituted seawater

NSW

natural media/natural seawater

EDS/EDX

energy-dispersive x-ray spectroscopy

ASW

artificial saltwater

DI

deionized water

DLS

dynamic light scattering

EW

egg water

GF

green fluorescent

EM

electron microscopy

ME

microemulsion

NP

nanoparticle

Nppt

nanoprecipitation

OF

orange fluorescent

OM

oocyte media

RF

red fluorescent

SM

spermatozoa media

TEM

transmission electron microscopy

UP

ultrapure water