Meike
van der Zande
*a,
Anita
Jemec Kokalj
b,
David J.
Spurgeon
c,
Susana
Loureiro
d,
Patrícia V.
Silva
d,
Zahra
Khodaparast
d,
Damjana
Drobne
b,
Nathaniel J.
Clark
e,
Nico W.
van den Brink
f,
Marta
Baccaro
f,
Cornelis A. M.
van Gestel
g,
Hans
Bouwmeester
af and
Richard D.
Handy
*e
aWageningen Food Safety Research part of Wageningen University & Research, Akkermaalsbos 2, 6708 WB, Wageningen, The Netherlands. E-mail: meike.vanderzande@wur.nl
bUniversity of Ljubljana, Biotechnical Faculty, Jamnikarjeva 101, 1000 Ljubljana, Slovenia
cUK Centre for Ecology and Hydrology, MacLean Building, Benson Lane, Wallingford, Oxon Ox10 8BB, UK
dUniversity of Aveiro, Department of Biology and CESAM- Centre for Environmental and Marine Studies, 3830-193 Aveiro, Portugal
eSchool of Biological and Marine Sciences, University of Plymouth, UK. E-mail: rhandy@plymouth.ac.uk
fDivision of Toxicology, Wageningen University, Stippeneng 4, 6708 WE, Wageningen, The Netherlands
gDepartment of Ecological Science, Vrije Universiteit, De Boelelaan 1085, 1018 HV Amsterdam, The Netherlands
First published on 27th April 2020
The structure of the gut barrier and luminal chemistry in non-mammalian vertebrates and invertebrates has been given little attention with respect to the dietary uptake of engineered nanomaterials (ENMs). This review compares the diversity of gut anatomy in selected species used for regulatory toxicity testing, especially in relation to gut lumen chemistry and the behaviour of ENMs, and the gut as a barrier to ENMs. High ionic strength, the presence of divalent ions and organic matter promote particle aggregation in the lumen. The redox chemistry of the gut offers reducing conditions for ENM transformation, and corona formation will depend on the gut contents. Areas of low pH in the gut lumen in several species will promote the dissolution of metallic ENMs. There is a protective unstirred layer over the surface of the epithelium that may concentrate ENMs. Some organisms, especially vertebrates, can slough mucus to remove this adsorbed nanomaterial and lower bioavailability. Invertebrates also have protective layers of cuticle or peritrophic membranes that will modulate ENM uptake. Paracellular uptake of ENMs is unlikely. Transcellular uptake via vesicular-dependent pathways remains the most likely route across the gut epithelium. Most species have receptor-mediated endocytosis pathways and/or macropinocytosis in the gut epithelium. Crucially, many invertebrates have another potential pathway via ‘intracellular digestion’ uptake routes leading into the gut epithelium, and with gut associated immune cells being a potential route for ENM translocation across the epithelium. The basal lamina provides another barrier prior to the internal compartments of many animals. The features of the gut lumen and epithelium can limit the uptake of ENMs across the gut barrier in vivo, although some ENMs are detected in the tissues. Invertebrates also have the ability for biogenic mineral formation at the nano scale inside tissues. In conclusion, despite the diverse structural anatomies of the gut barrier of animals, some common features in the gut lumen chemistry tend to promote particle aggregation and settling onto the gut surface. The functional anatomy ensures the gut remains a formidable barrier to ENMs, and with some potential novel uptake processes in invertebrates that are not present in vertebrate animals.
Environmental significanceThe bioaccumulation of engineered nanomaterials (ENMs) through aquatic or terrestrial food webs is a concern. However, the diverse structure of the gut barrier and luminal chemistry in the animal kingdom has been given little attention with respect to ENMs. A few key factors such as luminal pH, redox chemistry, ionic strength, and the organic matter in the gut enable some cross-species consideration of the hazard of ENMs. Differences in gut structure also achieve a functional physiology where the gut is likely to be a barrier to ENMs. Most species have receptor-mediated endocytosis pathways and/or macropinocytosis in the gut epithelium, but invertebrates are also at extra risk due to additional uptake routes involving cells that are used for ‘intracellular digestion.’ |
In the environment, ENMs are unlikely to remain in their pristine ‘as produced’ state, but will usually be subject to both chemical and physical transformations. These transformations can include alteration of the particle's composition, including the surface of the particle core, or any surface coatings on the ENMs; with subsequent effects on the agglomeration and aggregation behaviours of the materials, and/or dissolution of the particles, as well as changes in their chemical reactivity.3,7,8 Data suggests that these nanomaterial transformation processes are determined by both the physicochemical properties of the ENMs and the characteristics of their environment that are important in the behaviour of colloids (e.g., pH, ionic strength, the presence of natural organic matter).7,9,10 So, as is the case for traditional chemicals, it is expected that the environmental conditions will influence the fate and behaviour of ENMs in the ecosystem of concern. It has also been suggested that the chemical reactivity and colloid behaviours of ENMs in the environment will also influence their bioavailability to organisms, especially at critical external barriers such as the gills of fishes,11 or the human lung.12,13
However, the gut as a barrier has been given less attention, despite concerns that dietary exposure is likely a main route of exposure to wildlife. Aquatic mesocosm studies have shown that ENMs are deposited in sediments and biofilms resulting in contamination of the base of the food web.14,15 The trophic transfer of ENMs from primary producers to aquatic invertebrates has also been demonstrated in the laboratory.16,17 Fishes will also eat food contaminated with ENMs (TiO2,18 Ag,19,20 ZnO, carbon nanotubes,21). Similarly, in terrestrial ecosystems, invertebrates such as earthworms will ingest soil contaminated with ENMs,22–24 and potentially initiate a food chain hazard to predators.25 However, studies on the ecotoxicity and bioaccumulation of ENMs have been criticised for focussing on a few organisms that are used in regulatory toxicity tests.26,27 In contrast, the essence of legislation on environmental protection has always been to ‘protect most of the organisms most of the time’, and so some consideration of the vast biodiversity of wildlife is needed with respect to ENMs. There are far too many taxonomic groups and species to address concerns organism by organism. Instead, from the viewpoint of comparative physiology, biodiversity can be rationalised into a handful of basic body designs where form (structure) and biological function (physiology) are in harmony. The concept of ‘form and function’ is readily extended to toxicology, where adverse alterations in structure (pathology) and function (pathophysiology) informs on the hazard.
The overall aim of the current review is to focus on the importance of the gut as a main biological barrier to ENM uptake by wildlife, and to address those concerns for a wide variety of organisms by also considering the ‘body plan’ which in traditional comparative physiology refers to the ‘structure’ of the body and the arrangement of any tissues or organs therein, and how the structure of the organism relates to function. The specific objectives are to: (i) explore bioavailability in the gut lumen from the view point of diverse luminal chemistries of organisms and colloid theory, (ii) highlight the concerns for ENM uptake across the gut epithelium for different designs of gut anatomy, (iii) consider the fate after crossing the gut barrier with respect to the different body plans of animals and the internal organs. The ‘gut barrier’ is often considered in terms of the physical barrier of the intact epithelium that separates the external environment from the underlying tissue and the internal body fluids or equivalent serosal compartment. Here we also take a physiological approach, where, functionally, the various extracellular matrices that form layers in the lumen (e.g., the cuticle of invertebrates, mucus secretions of vertebrates) contribute to the overall barrier properties and the characteristics of uptake of substances across the gut.
Fig. 1 An idealised diagram of the vertebrate gut epithelium showing the mechanisms of uptake for electrolytes, toxic metal ions (Me+), compared to nanoparticles (NPs, filled circles). Modified from Handy and Eddy (2004).102 The substances in the luminal fluid diffuse into an unstirred layer (USL) comprising of water/mucus secretions, prior to transfer across the gut epithelium. The upper portion shows this for solutes. Cations bind to strands of polyanionic mucus where the exclusion of free anions like Cl− contributes to the Donnan potential at the cell surface. Electrolytes and toxic metal ions move through the cell using ion transport pathways. The lower portion of the diagram is for nanomaterials. The NPs will diffuse into the USL, albeit at a slower rate than solutes, and may be influenced by natural organic matter (NOM), pH, NaCl and divalent ions in the media. NPs will bind to strands of mucus, either by electrostatic attraction and/or become entangled in the mucoproteins (steric hindrance). Nanomaterials are taken up by endocytosis-related mechanisms and trafficked through the cell. Lipophilic ENMs such as C60 might diffuse through the lipid bilayer. The Ca2+ and Mg2+ rich environment in the tight junctions suggest that NPs would aggregate rather than diffuse through the paracellular route. For clarity the serosal processing of ENMs in the blood is not shown. |
The first few steps leading up to the delivery of the ENM to the apical membrane of the gut epithelial cells (i.e., adsorption to any external protective layers and bioaccessibility to the cell surface) will be considered here. First consider the importance of colloid theory to the behaviour of ENMs in the gut lumen. Engineered nanomaterials generally form suspended dispersions or emulsions in water. They are not truly dissolved in the aqueous phase, except where atoms or ions are released from the particle surface by dissolution (see reviews by Handy et al.11,31). It is possible to make seemingly ‘stable’ dispersions of ENMs in liquid for applications in the food or drinks sector, dentistry and oral medicines. For example, the dispersion of colloidal silver products in sodium citrate solution. However, while such a liquid may appear homogeneous, it is still a dispersion, not a solution of ENMs. Consequently, in principle, the behaviour of the ENM in the liquid environment of the gut lumen will be partly explained by colloid chemistry as outlined in extended DLVO theory [after Derjaguin and Landau (1941), and Verwey and Overbeek (1948)].31–33 Engineered nanomaterials will move in the liquid phase by Brownian motion (i.e., diffusion) and periodically collide with each other in the dispersion. If there is enough energy in the dispersion, the particles may then separate and stay dispersed. Alternatively, if the attractive forces (e.g., van der Waals forces) acting on the particles are stronger than the repulsive forces (e.g., the surface charge of the particles), then the particles will tend to agglomerate, or aggregate and eventually settle out of the liquid. In colloid theory, agglomeration is considered a somewhat reversible process where the addition of some energy may re-disperse the material, while aggregation relates to stronger attraction between the particles that leads to settling (see Handy et al.31 for details of DLVO and the forces involved). Aggregation and settling can occur if the particles in the dispersion are all the same (homo-aggregation), or where the ENM is aggregating with other colloids in the dispersion (hetero-aggregation). For ENMs in the gut lumen hetero-agglomeration and hetero-aggregation with proteins, food particles, etc., is therefore the most likely processes in vivo.
According to DLVO theory, conditions in the gut lumen such as altered pH, high ionic strength, the presence of divalent ions, and the type and concentration of organic matter will play a crucial role in any aggregation behaviours. Table 1 shows the pH, ionic strength, divalent ion concentrations and likely natural organic matter of content and composition in the gut lumens of a range of invertebrate and vertebrate animals that are used in ecotoxicity testing. The phylogeny of the organisms, and the regulatory tests they are used in, is shown in ESI† Fig. S1. Inevitably, there are species differences in gut lumen chemistry that also vary along the digestive tract (Table 1). The complexity of those changes will depend on the anatomy of each organism, the feeding strategy and the type of digestion. The presence and types of extracellular matrices (e.g., mucus or other secretions), and the innate permeability and tightness of the gut epithelium (i.e., passive properties of the barrier), according to the anatomy of the animals might also affect the uptake of ENMs across the epithelium. However, knowledge on the bioavailability and mechanisms of uptake of ENMs has, so far, mostly been elucidated using vertebrate animal models and mammalian cell lines, and so some consideration of invertebrate species is warranted.
Organism (species) | Chemistry aspect critical in DLVO theory | |||
---|---|---|---|---|
pH | Ionic strength | Divalent ions | NOM/other colloids | |
ND, no data available, either not reported in the literature, or not measured in a quantitative way with appropriate concentration units. FW, freshwater. SW, seawater. | ||||
Earthworms
(Lumbricus terrestris, Eisenia fetida, E. andrei) |
6.3–7.3 (ref. 173) | NaCl, ND |
Ca2+, 4 ≥ 0.5 mg g−1 (ref. 174)
Simulated gut fluid, 3.6 mmol L−1 Ca2+ (ref. 65). |
NOM from ingested soil. Also mucus, monosaccharides and glycoproteins.175 |
Aquatic polychaete worms
(Lumbriculus variegatus, Arenicola marina) |
Stomach: 5.4–6.0
Oesophagus: 6.5 Intestinal content: 7.0 (ref. 150) |
NaCl, ND |
Ca2+, ND
Mg2+, ND |
The gut content is 70–85% algae.176
Possible algal exudates and colloids from ingested water. Concentration of high molecular amino acids 58–215 mmol L−1, and high molecular organics, 85–233 mmol L−1 (ref. 177). Concentration of surfactants ∼13.3 mM (ref. 178). |
Roundworms
(Caenorhabditis elegans) |
Anterior pharynx, 5.96 ± 0.31;
posterior intestine, 3.59 ± 0.09 (ref. 179) |
NaCl, ND |
Ca2+, ND
Mg2+, ND |
ND |
Freshwater snails
(Physa acuta, Lymnea stagnalis) |
Oesophagus, 6.9–7.2;
gizzard/crop, 6.4; pylorus, 6.6; intestine, 7.1 (ref. 66) |
Surrogate artificial alimentary solution,66
NaCl, 35 mmol L−1. |
Surrogate artificial alimentary solution,66
Ca2+, 2.7 mmol L−1. |
Mucus and associated glycoproteins. |
Terrestrial isopod
(Porcellio scaber) |
Mid-gut, 6
Hindgut anterior part, 5.5–6.0; hindgut posterior part 6.0–6.5 (ref. 46 and 67) |
Fore- and mid-gut, 244.5 ± 6.1 mmol L−1 NaCl (ref. 180). |
Ca2+, ND
Mg2+, ND |
Around 10 mmol L−1 concentrations of surfactants in the hindgut.87 |
Water flea
(Daphnia magna) |
6.8–7.2 (ref. 181) | NaCl, ND |
Ca2+, ND
Mg2+, ND |
Filter feeder on algae. Possible algal exudates and colloids from ingested water. |
Springtail
(Folsomia candida) |
Anterior midgut and hindgut, 5.4–6.4
posterior midgut, 8.2–8.8 (ref. 182) |
NaCl, ND |
Ca2+, ND
Mg2+, ND |
Soil organism that grazes on fungal hyphae. Possible colloids from the food and soil in the gut. |
Mites
(Oppia nitens) |
Ventriculus and caeca, 5.4–6
Colon, 5.9–7.4 Post-colon, 6.5–8. In acaridid mites pH 4–7 from anterior to posterior.183 |
Large hindgut (in mmol L−1):
Na+, 35, K+, 98 Cl−, 146 (ref. 61). |
Ca2+, ND
Mg2+, ND |
Colloids from the liquid food (e.g., blood) of prey items. |
Honeybee
(Apis mellifera) |
Anterior, middle and posterior ventriculus: pH 6.0, 5.7 and 5.6 respectively.184
Large hindgut, 6.0–8.0 (ref. 61). |
No data on bees. Other insects (lepidopteran larvae)185
Na+, 1–1.3 mmol L−1 K+, 145–200 mmol L−1 |
No data on bees. Other insects (lepidopteran larvae).185
Mg2+, 8.6–27.4 mmol L−1 Ca2+, 11–19.6 mmol L−1 |
Numerous surfactants present in insect digestive fluid.62 |
Rainbow trout
(Oncorhynchus mykiss) |
Stomach; pH 2–5 (ref. 186)
Intestinal fluid187: pH 8.5 (FW), pH 8.1 (SW) |
Stomach (FW) in mmol L−1 (ref. 188):
Na+, 140–170 Cl−, 190–225 K+, 55–7 Intestinal fluid in mmol L−1 (ref. 187): Na+, 170 (FW), 20 (SW) K+, 4 (FW), 1 (SW) Cl−, 70 (FW), 50 (SW) |
Stomach (FW) in mmol L−1 (ref. 37):
Ca2+, 7–50 Mg2+, 12–40 Intestinal fluid in mmol L−1 (ref. 187): Mg2+, <1 (FW), 110 (SW) Ca2+, 2.1 (FW), 2.2 (SW) SO42−, <1 (FW), 110 (SW) |
Secreted mucus and organic matter from ingested food. |
Rat
(Rattus norvegicus) |
Stomach, pH 2.6–5.0 (ref. 189 and 190)
Intestine, pH 6.5–7.8 (ref. 189 and 190) |
Stomach in mmol L−1 (ref. 190 and 191):
Na+, 30.0–52.0 K+, 18.0–29.0 Cl−, 82.0–96.0 Small intestine in mmol L−1 (ref. 190 and 191): Na+, 113.0–153.0 K+, 6.0–52.0 Cl−, 60.0–100.0 |
Stomach in mmol L−1 (ref. 190 and 191):
Ca2+, 1.5–2.0 PO43−, 3.0 Small intestine in mmol L−1 (ref. 190 and 191): Ca2+, 0.25–8.0 PO43−, 23.0–24.0 SO42−, 3.4 |
Secreted mucus, and organic matter from ingested food. |
Chicken
(Gallus gallus domesticus) |
Proventriculus in chicken pH 2.1–3.8 (ref. 35 and 39)
Intestine of chicken, pH 6.4–7.7 (ref. 192) |
Intestine in mmol L−1 (ref. 192):
Na+, 67–83 K+, 19–27 |
Intestine in mmol L−1 (ref. 193):
filterable Ca2+, 17–11 |
In the intestine, the secreted mucus and organic matter from ingested food. |
Human
(Homo sapiens) |
Stomach: Fasted, pH 1.5–2; fed pH 3–7 (ref. 194)
Small intestines: Fasted, pH 4–8, typical value 6.5 in the upper small intestine.195 Fed, pH 3–7, typical value pH 5 in the upper small intestine.195 |
Simulated gastric fluid based on human in vivo data196 In mmol L−1:
Na+, 72.2 Cl−, 70.2 K+, 7.8 Simulated intestinal fluid based on human in vivo data.196 In mmol L−1: Na+, 123.4 Cl−, 55.5 K+, 7.6 |
Simulated gastric fluid based on human in vivo data.196 In mmol L−1:
Mg2+, 0.1 Ca2+, 0.15 Simulated intestinal fluid based on human in vivo data.196 In mmol L−1: Mg2+, 0.33 Ca2+, 0.6 |
Secreted mucus and organic matter from ingested food. |
The pH values in the stomach of carnivorous vertebrate animals with acid digestion can be as low as pH 2–3 (Table 1). Although, in the terrestrial vertebrate animals, the precise pH might vary with the age of the animals and type of food (e.g., in birds35,36). In contrast, for the terrestrial invertebrate animals such as earthworms, snails, isopods and honey bees, the gut lumen is less acidic with the lowest pH values around pH 5; with the possible exception of Caenorhabditis elegans where the posterior intestine can be around pH 3.5 (Table 1). Dissolution and/or the degradation of any surface coating on the ENM will also depend on the residence time of the ingested material in the acidic part of the gut. Gut transit time is influenced by many factors including body size, ration size, the type of food eaten, and body temperature.
The residence times of food in the stomach is typically a few hours in vertebrate animals (fish, <8 hours;37 rats, 6–9 hours;38 chicken, 0.5–1 hour in the proventiculus;39 humans, 0.5–5 hours40). Some species, like geese have relatively short gut transition times (several hours), with relatively low digestion efficiency,41 while vultures for instance may have gut residence times of >24 h.42 In contrast, gut transit times in invertebrate species are much less. Indeed, for invertebrate species the entire gut transit time can be from minutes (e.g., < 30 minutes in Daphnia,43 ∼35 min for springtails44) to several hours (e.g., in earthworms,45 isopods,46 aquatic worms47). At neutral pH in saline conditions the maximum dissolution rates of metal NPs are typically at the μg min−1 level,48 and even if acidity increased this one hundred fold, this would still only represent around milligram amounts of dissolved metal in the gut lumen. For nutritionally required metals such as Zn or Cu, where a few mg per day are needed for animal health, such releases would be of no consequence. However, in the case of a trace metals that are known to bioaccumulate or are toxic, such as Cd released from CdTe quantum dots,48 then a repeated dose of a few μg with each meal could present a long-term hazard. In contrast, some metal oxides are resistant to acidity and tend to show low dissolution in gut salines (e.g., TiO2,49). The higher temperature of the mammalian gut compared to cold blooded animals may also enhance dissolution of the ENMs, since in simple dispersions increased rates of dissolution with rising temperature can be demonstrated.50
Regardless of the dissolution mechanism, this highlights that the luminal chemistry may lead to a dissolved metal fraction that is available for uptake on solute transporters in the gut lumen (Fig. 1). For dissolved metals, bioavailability depends on the chemical speciation of the metal and the presence of competing cations in the media such as H+,51 Ca2+ or Mg2+ (e.g., as water hardness,52) and Na+ (ionic strength or salinity,53). These ideas have culminated in the biotic ligand model (BLM) which predicts metal exposure to the fish gill,54 and also to aquatic invertebrates.55 A gut BLM is not yet available and metal sources from ENMs are not currently included in the aquatic BLM model.
Alkaline pH values are also found in the intestines of fish and mammals, as well as parts of the gut in invertebrate species, with the intestinal fluid in fish reaching pH 8.5, and in posterior midgut of springtails reaching pH 8.8 (Table 1). Such alkaline conditions may preserve at least metal ENMs, since strong alkaline digestion methods are used to extract ‘intact’ ENMs from tissue (e.g., Ag NPs from fish liver,56). However, the extreme ranges of pH from acid to alkaline would suggest that at some point during the transit through the gut, the ENM will be at a pH value close to its zero point of charge, where particle settling due to aggregation may occur. Metal ENMs are often designed so that they disperse at pH 7 in water, and so for example, Ag NPs that have a point of zero charge around pH 3,57 might be expected to aggregate in the stomach of vertebrate animals.
The divalent ions in the gut lumen include Ca2+, Mg2+, SO42− and PO43− (Table 1). In birds, mammals, and freshwater fish, the dissolved Ca2+ and Mg2+ concentrations in the gut lumen are generally a few millimoles, but much higher in seawater adapted fish that drink the surrounding medium (Table 1). Some animals precipitate calcium and magnesium carbonates in the lumen of the intestine as a means of removing secreted HCO3− from the gut lumen as part of the animal's acid–base balance strategy (fish,63), or sometimes as phosphates (cows,64). Consequently, some caution is needed when interpreting calcium and magnesium measurements in the gut lumen as ‘dissolved’ metal. For invertebrates, not many data are available specifically for divalent ions. For earthworms65 and snails66 values of 3.6 mmol L−1 and 2.6 mmol L−1 Ca2+ respectively, were used to simulate gut fluid. Nonetheless, millimolar concentrations of other cations have been reported for some invertebrates (e.g. K+ in isopods;67 NH4+ in earthworms65) (Table 1), and this would at least contribute to particle aggregation along with the NaCl present. The enhanced charge screening of ENMs due to divalent cations is well known in DLVO theory (see Handy et al.31) and for metal ion adsorption to epithelial surfaces.54 On an equimolar basis, the higher charge density of divalent cations relative to monovalent ions such as Na+, will drive adsorption to the fixed negative charge of the particle, or the cell membrane in the case of epithelia. In reality, the ionic activity of all the competing cations in solution should be considered, and the mobility in water of the divalent metals relative to H+ which is the fastest diffusing ion in solution.53 As the mobile cations are attracted to the surface of fixed negative charge on the particle, the diffusible anions are excluded, and this contributes to the surface potential or zeta potential of the particles. Similarly, diffusible anion exclusion also contributes to the measurable Donnan potential (the voltage arising from the passive distribution of ions) on the surface of the gut mucosa (Fig. 1). In theory, these processes should also apply to divalent anions in the gut lumen being attracted to the surface of a material that has been manufactured with a positive coating or surface charge. So, one might expect phosphates and sulphates in the gut lumen to influence the behaviour of positively charged particles, leading to charge screening and eventually aggregation. However, these effects of anions appear not to have been investigated in biota for ENMs.
In vitro studies have also shown that the composition of the media in which nanomaterials are suspended will affect the composition of the ‘biomolecular corona' that spontaneously forms on the surface of the ENM, and subsequently how the nanomaterial interacts with the cell membrane to enable cellular (i.e. epithelial) uptake of the material.58,80–84 However, with so many colloids from the food and gut secretions present in the gut lumen, it is not yet possible to model how the corona could be modified as the ENM moves along the gut tract. However, empirical results on the behaviour of ENMs in the human digestive tract in vitro vary with the type of material, with the gut lumen conditions either increasing85 or decreasing86 the translocation of nanomaterials over the epithelial cell layer.
The presence of food during digestion in the gut has also been suggested to reduce agglomeration of the nanoparticles, perhaps by stabilising the nanomaterial dispersion with organic matter through corona acquisition and steric hinderance.86 Surfactants are also secreted into the guts of many animals to prevent the digestive enzymes from precipitating, or to improve the digestion of lipophilic nutrients.62,87 These are natural dispersing agents that might also improve the dispersion of certain types of ENMs in the gut lumen. For humans at least, there is some evidence that bile salts might influence the aggregation behaviour and dispersion of ENMs in the gut lumen,29 but almost nothing is known of these processes with ENMs in the gut of wildlife. Given the small size of invertebrate animals, sometimes only a few microliters of luminal fluid can be collected, albeit with uncertainty about contamination of the sample with sloughed cells, mucus, etc. Consequently, most of the information on the enzymes and proteins secreted by the gut of small invertebrates is derived by semi-quantitative methods such as immunohistochemistry of the gut epithelium. Nonetheless, it is likely that the luminal protein concentrations, digestive enzymes, presence of salts, divalent ions, etc., together exceed by far the critical concentrations for particle settling according to DLVO theory. The gut, by its very nature, will contain high levels of solid phase food components such as fibre that may adsorb ENMs, as is known for other chemicals, but alternatively, dissolved organic carbon throughout the digestive tract might aid dispersion. Further research is needed to resolve these issues for most ENMs and species of animals.
Limited research has been performed on how interactions between the gut microbiome and ENMs affects the ENM transformation, and thus bioaccessibility of the ENMs from the gut lumen matrix. The gut microbes are as a consortium of organisms involved in the digestion of the food and other processes. They can use carbon sources in the food directly, or the redox energy in the gut environment, to fuel their own energy metabolism. This might also include using the organic coatings on some ENMs as a carbon source (i.e., microbial degradation of the coating). Regardless, the microbial activity in the gut has the potential to alter the gut lumen chemistry, and certainly gut function.95 While the collection of data on the interaction between ENMs and gut microbes is at an early stage, there are some suggestions that ENMs might change the microbial community structure.30,96,97 Merrifield et al. demonstrated that the microbiome in the zebrafish gut following dietary Ag or Cu NP exposure varied, suggesting some effect relating to the chemical substance itself.98 However, the study also demonstrated that dietary CuSO4 resulted in a different microbial biodiversity to that of Cu NPs; indicating a ‘nano effect’ on the microbiome for the form of Cu presented to the fish. Whether or not such changes in the microbiome ultimately alter ENM bioavailability as well is currently unknown. The toxicological or nutrition consequences of such changing microbiology of the gut are also not clear. Important aspects, such as recolonization of the gut, or the evolution of a new steady state in the microbiome following ENM exposure have not been investigated.
The mucus layer secreted by the epithelium is integral to the USL and its sol–gel properties. The hydrated mucus of vertebrate animals is typically around 97% water and contains mucoproteins, which are made of a peptide backbone with numerous polysaccharide side chains. The composition of mucus is very highly conserved across species, with the mucus produced by diverse organisms such as anemones, jellyfishes, molluscs, rainbow trout skin, pig and human gut, showing similarity in structure (see review on mucus103). From the viewpoint of metal ions, mucus is a polyanionic matrix that attracts metals according to their charge density and ionic mobility in solution relative to H+.53 Thus trivalent metals such as Al3+ are attracted into mucus more than say Cu2+ ions, and divalent ions much more than Na+ on an equimolar basis. Engineered nanomaterials that have a net positive surface charge, would in theory, be electrostatically attracted to the USL and the mucoproteins therein. Negatively charged particles might show some exclusion. In mammalian mucus at least, neutral or slightly negatively charged ENMs have been shown to interact minimally with the mucous layer, enabling quick access to the intestinal cells.104
Alternatively, since ENMs are much larger than solutes, they may physically tangle with the strands of mucus (steric hindrance). Decoration of nanomaterials with longer surface chains has been shown to decrease mucous penetration, most likely caused by entanglement of the surface groups in the mucous mesh. The average pore size of the mucous mesh in mammalian preparations is 100 nm,105–107 indicating potential access to the cell layer for nanomaterials <100 nm, the size cut-off value is suggested to lie around 500 nm.108 Whatever the mechanism, the precipitation of mucus with ENMs in gut preparations has been observed (TiO2,49 Ag NPs,94), such that typically two thirds of the exposure dose is sloughed from the epithelium. When this is coupled with aggregation of ENMs in the gut saline, the bioavailable fraction that is taken up into the tissue is often only a few percent of the initial dose in vivo.19 It is of course, one of the functions of the gut mucus to protect the underlying epithelium from chemical insult in the gut lumen, and this would seem to be the case also for ENMs.
Fig. 2 A diagram of invertebrate gut barriers showing different types of extracellular matrix on the apical side (facing gut lumen) of gut epithelium and with the external cuticles and/or peritrophic membranes. Mucus secretion into the lumen is possible in some species (not shown). The basal lamina and muscle layer on basal side of the epithelial cell provide some protection from uptake into the internal body compartment. Invertebrate gut is typically composed of three regions (fore-, mid-, and hind-gut) each with distinct functions and type of extracellular matrix. Potential uptake across epithelium is shown via receptor-driven endocytosis, phagocytosis or pinocytosis. Also, the protrusion of phagocytic cells to the gut lumen as part of the immune function is possible and being a potential pathway for the uptake of ENMs. The subsequent fate of ENMs after passing the epithelial barrier is shown depending on the absence of a body cavity (acoelomates), and presence of a pseudo cavity (pseudocoelomates) and with a true body cavity (coelomates) (after Sadava et al.228) |
Organism and plan of the gut | Structures in the gut barrier | Evidence of endocytosis-related mechanisms in the gut epithelium | Biomineralisation of metal granules | |||
---|---|---|---|---|---|---|
Peritrophic membranes | Cuticle | Mucus secretion | Epithelium and underlying tissue | |||
Earthworms (Eisenia fetida and Lumbriculus sp). Foregut, midgut and hindgut. The foregut contains the buccal chamber, pharynx and anterior portion of oesophagus. | Over the midgut. The peritrophic membrane in some species also contain chitin.197,198 | The foregut and hindgut are lined with non-chitinous cuticle.197 | The pharyngeal glands secrete mucus.199 | Epithelial lining of glandular cells, and non-glandular ciliated cells.199 Muscular contractions of the gizzard mechanically grind up the food. Circular and longitudinal smooth muscle under the epithelium. | Digestion mostly extracellular with no evidence of phagocytosis or pinocytosis.151 | The calciferous glands have type D granules composed of Ca. Chloragocytes contain two types of metal granules; the ‘chloragosomes’ which are type A granules containing P, Ca, Zn, Mg, Fe., and ‘debris vesicles’ which contain S.200 |
Aquatic polychaete worms (Arenicola marina): Basic annelid plan of foregut, midgut and hindgut, similar to earthworms. | Absent in the midgut, but in Arenicola, a peritrophic membrane-like structure is formed around the faeces.198 | The foregut and hindgut are lined with a cuticle of sclerotized proteins without chitin.112 | Oesophageal pouches secrete mucus destined for the stomach, goblet cells and mucus in the intestines.150 | Epithelial cells are ciliated in parts of the gut. Circular and longitudinal smooth muscle under the epithelium. The buccal cavity and foregut secrete enzymes involved in digestion. The mid- and hindgut are restricted to absorption. | In the stomach, some epithelial cells take up food particles by phagocytosis. Some food components are digested by wandering amoebocytes.150 | Ca and Mg rich phosphate granules in the proventriculus and muscle tissue.201 |
Roundworms the gut of Caenorhabditis elegans comprises pharynx and buccal cavity, and a long tube of intestine, and short hindgut/rectum.202 | Active glycocalyx protects the intestine, rather than peritrophic membranes.202 | The lumen of the pharynx and buccal cavity are lined with cuticle,203 and is composed of cross-linked collagens, cuticle proteins or ‘cuticlins’, glycoproteins and lipids.204 | Mucus is secreted in the pharynx.203 | The intestine has epithelial cells with a brush border and specialization cells involved in digestion. The pharynx is lined with muscle cells, but elsewhere the serosal surface is less protected.203 | Phagocytosis occurs in the gut epithelial cells.205 Intracellular digestion is present in nematodes.141 | Formation of Pb phosphate granules during metal exposures, crystalline pyromorphite.206 |
Freshwater snails: in Lymnaea stagnalis comprises of mouth and pharynx (buccal mass), the oesophagus, stomach, digestive gland, intestine, and anus. | A mucous and viscous material surrounds the fecal string in the gut appearing arranged like a peritrophic membrane which has no chitin and no microfibrils.198 | Mollusc lack hard chitinous cuticle. It was only found in the buccal cavity and in some cases the epithelium of the stomach in gastropods is covered by a cuticle called the gastric shield.109 | Mucus secreting cells in the stomach.207 | The entire digestive system, with the exception of the gizzard and parts of the buccal cavity, is ciliated A muscularis supports the epithelium. Food is ground with sand in the gizzard by muscular movements in the stomach.66 Sorting of food particles takes place in a complex system of ciliated passages in the pylorus. | Amebocytes were found in digestive gland.208 Only particles with a diameter of less than 400 nm enter in the digestive gland, where intracellular digestion via phagocytosis occurs.208 | At least three types of metal storage granules are observed for metal exposure.209 |
Terrestrial isopod: In Porcellio scaber, a straight tube consisting of a small foregut (oesophagus, and stomach or proventriculus), a junction with two pairs of tubular midgut glands, and a large hindgut (80–90% of the total length). | Fine mesh filters (mesh size 40–50 nm) prevents access of particles to digestive gland lumen.114 | The foregut and hindgut are lined with a thick chitinous cuticle which is permeable to digestive products up to 1.9 nm.210 | No evidence of mucus production in gut. | Food is mechanically degraded in the foregut and digested in the hindgut. Adsorption of nutrients takes place in midgut digestive gland. | No evidence of phagocytosis in terrestrial isopods. Pinocytotic vesicles were found in digestive gland of predatory marine isopod,211 and in hindgut of terrestrial isopod Armadillidium vulgare.212 | Cu-, sulfur- and Fe- rich granules in cells of the digestive gland (type B and C granules).200 |
Water flea (Daphnia magna). The alimentary canal of Daphnia consists of a tube-shaped foregut, a midgut, and a short hindgut. | The tubular peritrophic membrane surrounds the food and extends through the midgut and hindgut.213 The peritrophic membrane is typically 280 nm thick, made of chitin containing microfibrils, embedded in a matrix of proteins, glycoproteins and proteoglycans. The peritrophic membranes are permeable to 31 and 130 nm nanoparticles but not to 327 nm (ref. 214). | The foregut and hindgut are lined with a thick cuticle (up to 1–2 μm in D. pulex).215,216 | No evidence of mucus production in gut. | The midgut consists of three parts: one dedicated to produce peritrophic membranes, a pair of small diverticula or hepatic caeca for secretion of digestive enzymes, and the third part for adsorption of substances. The inner surface of the midgut is folded and densely covered with numerous microvilli.213 The thin gut muscularis encircles the entire length of the midgut and caeca. | No evidence of phagocytosis. Pinocytotic vesicles were described in epithelial cells of other crustaceans.141 | In case of cadmium exposure calcium-contain granules are formed in the midgut.217 |
Springtail: (Folsomia candida) a foregut, an enlarged sac-like midgut, and a small tubular hindgut. | The midgut is protected by peritrophic membranes. Polysaccharides, glycoproteins and carbohydrate components have been demonstrated on the surface of microvilli.218 | The fore- and hindgut are lined by a cuticle. | No evidence of mucus production in gut. | Nutrients are absorbed from the lumen by the midgut epithelium that consists of a single layer of simple columnar or cuboidal cells which bear numerous filament-like microvilli.219 | No evidence of phagocytosis or pinocytosis. | Midgut cells contain numerous ‘type A' granules with Ca, Mg and K.200 Mineral aggregates (spherites) in the gut epithelium during metal exposure.219 |
Mites (Oppia nitens): a foregut consisting of pharynx and oesophagus, a midgut composed of ventriculus, paired caeca, colon, intercolon and postcolon, and a hindgut or anal atrium.117 | The midgut in several Acari species is lined by a peritrophic membrane.198 Peritrophic membrane can be up to 1 μm thick.117 | The foregut and hindgut are lined by a cuticle.117 The foregut cuticle thickness is ∼500 nm thick, hindgut up to up to 3 μm. Foregut cuticle it is perforated by numerous pore canals (70–150 nm). | No evidence of mucus production in gut. | Digestion begins externally using enzymes that are pumped into the body of prey or a plant cell, and continues in the midgut and gastric caeca.220 In most mite species, the caeca and ventriculus compose a large portion of the gut, and therefore probably play a key role in food digestion.183,221 The gut epithelium is as simple squamous epithelium, or cuboid in places, and a digestive cells have numerous microvilli.117 The midgut is surrounded by two layers of muscles. | Some mites that feed on decomposed prey have intracellular digestion processes implying phagocytosis.149Pinocytotic vesicles were detected in digestive glands of mite Acarus siro no specific data for Oppia sp. | There are granules in the cells of the caeca granules and seem to play a role in the storage of Ca and metals.222 |
Honey bee (Apis mellifera): a tube containing a foregut (a mouth cavity, long thin oesophagus and the crop or ‘honey stomach’), a midgut (ventriculus or stomach), and a hindgut followed by rectum).223 | The midgut secretes the peritrophic envelope or membrane,.61,113,224 Made of proteins (peritrophins) and chitin. There is a particle diameter limitation to pass the envelope in insects, ranging from 4.5–35 nm depending on species.61,113 | The foregut and hindgut are lined by a cuticle (e.g., 10 μm thick in cockroaches)61,224 cuticle is impermeable to polysaccharides, and inulin (3 nm).225 | No evidence of mucus production in gut. | The gut is lined with an epithelium consisting of a single layer of cells with microvilli. | No evidence of phagocytosis. Generally, the digestion of insects is considered exclusively extracellular.226 | Midgut and fat body cells contain granules with Fe, P, Ca, K and also other metals.227 |
The midgut epithelium is the absorptive epithelium and lacks a cuticle, instead it is separated from the lumen contents by peritrophic membranes (Table 2). While these peritrophic membranes have no associated cuticle, they may be with or without chitin (in Arthropoda and some annelids and nematodes, respectively). In some cases the peritrophic membrane may be permeated by microscopic pores (for example up to 35 nm for honey bee, up to 130 nm for daphnids).61,113 In Crustacea, the peritrophic membrane is present in some species (e.g., Daphnia magna), but completely absent in others (terrestrial isopods e.g. Porcellio scaber). In the latter organisms, access of particles to the midgut cells is likely limited by a fine mesh filter positioned at the entrance to digestive glands. The mesh size of these filters was described as 40–50 nm,114 but, wolfram oxide fiber-like ENMs (mean diameter below 100 nm, their length was on the millimetre scale) were found inside the digestive gland lumen and attached on the cells.115
Clearly, the cuticle and/or any associated peritrophic membranes will vary in composition, thickness and the size of pores; and these factors are likely to be important in the physical access of ENMs to the apical surface of the gut epithelial cells. The structure of the cuticle also varies significantly in different regions of the gut. The isopod cuticle can be 1.5–3 μm thick and allows the passage of 0.7–1.9 nm particles,116 while 70–150 nm pore canals were found in the cuticle of some gut regions in mites with the cuticle thickness up to 0.8–2.5 μm.117 Beneath the epithelial cells typically lays the basal lamina, which can have a thickness from 100–300 nm in isopods (being regarded as outstandingly thick for invertebrates).116 The basal lamina is supposed to act like a charged sieve, in which the passage of macromolecules depends on its charge and porosity. In insects, for example, it has been shown that the basal lamina of the midgut prevented the passage of 6–15 nm gold nanoparticles.118
Teleost fishes can have six major anatomical sections to the digestive tract.119,120 These are the buccal cavity (mouth), oesophagus, stomach, pyloris (anterior intestine), mid and hind intestine. However, with some 38000 species of fishes it is perhaps no surprise that there are some very diverse anatomies.121 In some herbivores (e.g. carp species) the stomach pouch may be less distinct or absent with the stomach being simply a continuous tube with the rest of the intestine. In carnivores that use acid digestion (e.g. trout), the stomach is usually well defined, while omnivores may show an intermediate anatomy. Notably, the volume of lymphatic drainage in fishes and some amphibians far exceeds anything that a mammal could achieve. These former animals may therefore take up dispersed materials faster than anticipated by their body temperature/metabolic rate. In many lower vertebrates, the gut anatomy can be transient or seasonal with the food supply. Only birds and mammals tend to maintain the gut anatomy in a constant state of readiness to absorb substances from the gut lumen. Thus, feeding status will be a critical factor in the passive gut permeability of some fishes, amphibians and especially reptiles.
With respect to mammals, the laboratory rat has been used as a model for the dietary uptake of pollutants for many years (e.g. metals122 and organic chemicals123). The gross anatomy is similar to human anatomy with the mouth, oesophagus, stomach, an absorptive small intestine (duodenum, jejunum, ileum), the large intestine (caecum leading to colon), and the rectum. However, there are many functional differences in the rat gut physiology compared to humans.124 The topography and spatial arrangement of the vertebrate gut can influence the transit time of particulate materials; but in the small intestine, solutes and particulate materials may be absorbed. There is some controversy on the total surface area of the small intestine in humans, as it is estimated at 250–300 m2, but more recent at ∼30 m2.125 Microfold cells (commonly referred to as M-cells) are associated with the Peyers' patches in humans and rats.126 M-cells are involved in the active transport of ENMs over the gut barrier.127–129 Mammals as terrestrial animals must conserve water by reabsorbing any fluids secreted into the lumen during digestion as well as osmotically drying the faeces. There is generally a net secretion of watery fluid into the lumen of the stomach and small intestine of mammals, but it is removed in the large intestine. Consequently, like freshwater fishes, the prospect of passive net uptake of nanomaterials by solvent drag processes in the small intestine of mammals is not likely. However, solvent drag would be theoretically possible in the large intestine where water is reabsorbed.
Birds comprise a wide range of species and their digestive tract is very adjustable, changing according to food availability,130 but also, for instance, in relation to wintering conditions,131 migration132 and moulting.133 Thus the readiness of the absorptive epithelium to take up ENMs will likely vary with these conditions, although it has not been specifically investigated. The digestive tract of birds consists of different regions, with specific functions: oesophagus, crop, proventriculus, ventriculus or gizzard, duodenum, jejunum, ileum, caeca and colon. It is known that the grain size of the food particles can alter the nutritional performance of poultry.134 However, the absorption efficiency of ENMs or how much they are taken up in the different regions of the gut in birds has not been investigated. The oesophagus contains a fold of tissue called the crop where food can be temporarily stored, before being released onwards to the stomach or gizzard. Whether or not ENMs incidentally stored in the crop are modified is unknown. Mucus is also secreted into the oesophagus in order to transport the food to the gizzard, and the concerns for ENMs in mucus (above) also apply to the gut of birds. The oesophagus widens into the proventriculus which secretes enzymes and acids for the digestion of the food. Following this, the food will go to the ventriculus/gizzard which is the part of the stomach that will mechanically grind and mix the food. This grinding can be relatively harsh in order to break up the hard grains of wheat seeds, etc., that birds may eat and so there is a prospect also of mechanical and chemical erosion of ENMs in this stage of the gut. In the small intestines (duodenum, jejunum and ileum) further digestion takes place of the food. In the duodenum digestive enzymes are being released and the bile duct also enters the gastrointestinal tract here. The caeca are blind pouches at the end of the small intestines where further digestion of the food can take place, and where water is reabsorbed.135 Similar to mammals, the net fluid secretion into the small intestine would prevent the passive uptake of ENMs by solvent drag, but in the lower reaches of the intestine and colon some absorption might be possible. The colon, which can be relatively short in birds, is where further water is reabsorbed. The net fluid influx would promote ENM movement onto the epithelium. Unfortunately, dietary studies on birds, so far, have tended not to characterise the ENMs in the food, or include metal salt or bulk material controls in the study design, so drawing nano-specific conclusions is problematic. Nonetheless, there are some tentative suggestions that some ingested ENMs might alter the weight gain and food conversion ratio of young poultry (e.g., chromium-containing ENMs,136).
If paracellular uptake is excluded or limited, then uptake through the gut epithelial cells is the most likely route to the serosal compartment. In vertebrates, extracellular digestion is prevalent where secreted enzymes degrade food particles in the gut lumen and smaller molecules (sugars, amino acids, etc.,) are then taken up by cells. Therefore, vertebrate enterocytes are regarded as ‘non-phagocytic’.141 Nonetheless, the possible active mechanisms of vesicular uptake of ENMs by the gut of vertebrate epithelial cells includes macropinocytosis, clathrin-mediated endocytosis and caveolae-mediated endocytosis (reviews,107,142). In some special cases, such as bacterial infection in the gut, immune dendritic cells can protrude to the apical surface of the gut between two adjacent enterocytes as demonstrated with monolayers of Caco-2 cells.143 While the extent of this latter phenomena is unclear for vertebrate animals, the phagocytosis of ENMs is theoretically possible via this route, and especially in some invertebrates (see below).
Uptake of ENMs on solute transporters in the cell membrane of gut epithelial cells is excluded because the ENMs are far too big.11 For spherical lipophilic materials, such as pristine C60 particles with a size of ∼0.7 nm, it has been shown that they translocate through the artificial lipid bilayer in a manner consistent with diffusion.144 However, pristine C60 might be a special case, and for example, engineered liposomes (∼100–200 nm) are often designed to fuse with the apical membrane to release their contents, rather than be transported as intact liposomes. Regardless, caution is needed when extrapolating the idea of lipophilic diffusion to other materials, or from in vitro models with artificial media to the in vivo situation. However, most of the effort on mechanisms of ENM uptake comes from studies on rodent gut,145 fish gut,49 or gut cell lines such as the Caco-2 cell culture and co-culture variants of this model.146,147 For example, nystatin, a putative caveolae-mediated endocytosis inhibitor, blocks Ti uptake during TiO2 exposures in perfused intestine preparations from trout.49 Similar pharmacological studies have been done with Caco-2 cells along with electron microscopy to capture vesical formation at the cell membrane.146 However, one should remember that most of the mammalian cell lines are derived from carcinoma tissue,147 and may not show ‘comparable’ endocytosis processes to primary cell cultures or in vivo. Nonetheless, endocytosis is an active, energy-dependent, receptor-driven process and is implicated in the uptake of different ENMs (reviewed in detail by des Rieux et al.).148 Albeit, with some uncertainty about how ENMs might interact with the membrane receptors involved.
In contrast to vertebrates, there are few pharmacological or kinetic studies on the gut uptake mechanisms of ENMs in invertebrate species. This is partly because their small size makes techniques such as gut perfusions challenging, but also because cell cultures of gut cells from invertebrate species are not routinely available. Nonetheless, several invertebrate species possess phagocytosis or pinocytosis in the gut epithelium (Table 2). These organisms can also have ‘intracellular digestion’ where food material is only partly digested in the gut lumen and the resulting food particles are taken up by phagocytosis or pinocytosis and digested intracellularly by cells in the lumen. Intracellular digestion primarily occurs in primitive unicellular animals (like Protozoa, e.g., Amoeba) as well as some organisms without a true gut (Porifera, e.g., sponges).149 From the viewpoint of evolution, intracellular digestion was first assisted by amoeboid cells in the gut lumen, then gradually evolved to be entirely replaced by extracellular digestion in the gut lumen. Some invertebrate groups, however, have retained some mechanisms of intracellular digestion in co-existence with extracellular digestion. The cells involved in intracellular digestion can be “wandering” amoeboid cells,150 or gut epithelial cells that phagocytose food.112,151 In a recent review on the phagocytosis in enteric cells, Hartenstein et al.141 refers to two types of phagocytic cells differing in terms of the size of particles they take up: (i) cells that perform phagocytosis of large particles (even particles almost the size of a cell) that are sequestered in large phagosomes, and (ii) cells that are involved mainly in uptake of smaller particles ending up in endocytic vesicles. The intracellular digestion has been related most to the type of food collected, being more commonly observed in organisms that have fine sorting systems of food and those that suck partly-digested food. Some examples of organisms with evident presence of phagocytosis were found in Rotifera (e.g., Brachionus), Platyhelminthes (e.g., planarians), many classes of Mollusca (especially Bivalvia and Gastropoda), Cnidaria (e.g., jellyfish), Ctenophora (e.g., comb jellies), and some blood sucking Nematoda species.112,149,152,153 Annelids predominately have extracellular digestion, but some members like marine polychaetes phagocytose food particles.150,151 In the majority of Arthropoda (crustaceans, insects), except for Arachnida (spiders, mites), the digestion is considered predominately extracellular but also here some evidence of pinocytosis does exist. Clearly, where phagocytosis or pinocytosis mechanisms exist in the gut of invertebrates, ENMs may follow the same route of intracellular uptake as food particles. The extent to which this is possible will depend on the type of phagocytosis and the predominant type of digestion. However, this has yet to be demonstrated in the gut of most invertebrates for ENMs.
For the invertebrates, when the ENMs have transversed the epithelial cells via endocytosis, or similar mechanisms, they most probably do not move freely in the extracellular fluid, but are taken up again by cells (e.g., haemocytes) responsible for recognition of ‘non-self material.’ These cells are broadly analogous in function to the macrophages in the immune system of vertebrate animals. This has been evidenced in vivo in the case of haemocytes in insect larvae158 and mussels.159 There are also numerous in vitro data reporting ENM uptake by the haemocytes of invertebrates.160,161 The fate of ENMs in extracellular fluid will partly depend on the local immune response, and any congestion of haemocytes (or similar cells) in the tissue to remove foreign material. Similar arguments could be applied to vertebrate animals when the ENMs arrive in the blood supply or lymphatics, since, for example, macrophages can engulf ENMs. A detailed discussion of immune systems is beyond the scope here, but there are substantial differences in the components that make up the immune systems of vertebrates and invertebrates. For example, invertebrates do not have the range of functions found in the white blood cells of mammals, and they do not have antibodies per se.162 The immune response of different organisms to chemicals is also dependent on this phylogeny of the immune system functions (e.g., with pesticides163), but the importance of differences in how the immune systems of animals enables ENM uptake across the gut and distribution in the tissues requires further investigation.
The concerns raised regarding pH, ionic strength, divalent ions and organic matter in the gut lumen (Table 1), also apply to the extracellular fluid or haemolymph. The pH values of extracellular fluids are usually in the neutral range (e.g., pH 7.4 in humans, pH 7.8 in freshwater trout and crayfish), and therefore less likely for any organic surface coating of ENMs to be ionised, i.e., perhaps close to the point of zero charge where aggregation can occur. For invertebrates, the NaCl concentrations typically exceed 100 mmol L−1 (i.e., high ionic strength), except insects which have high KCl concentrations in their body fluids, and with millimolar concentrations of divalent ions or higher also present.164 Similarly, vertebrate blood has around 140 mmol L−1 as NaCl concentrations. Consequently, ENMs, when present, are likely to aggregate in the extracellular fluid of both invertebrate and vertebrate animals.
The distribution of ENMs around the body and to the different internal organs will depend on the body design of the animal and the presence and/or type of circulatory system. Some groups (acoelomate, for example flatworms and sponges) lack an enclosed fluid-filled body cavity and do not have a circulatory system. In this case ENMs that potentially cross the gut barrier are more probably retained locally at the site of entry and entrapped in phagocytic cells, such as in the macrophage-like cells and granule-containing cells of sponges.165 On the other hand, in organisms with a pseudo body cavity (pseudocoelomate, e.g., roundworms) and true body cavity (coelomate, e.g., earthworms), the extracellular fluid could transmit cells with ENMs to other sites internally (Fig. 2). For animals with open circulatory systems (e.g., in molluscs and arthropods) the ENM is likely to be moved around more rapidly by the bulk transport of the body fluids. In animals with closed circulations, where the arterial and venous systems are joined in a continuous circuit (as in earthworms, squid, octopus and vertebrate animals), it implies that the blood vessels provide an additional barrier. However, the volume of distribution for animals exposed in vivo to ENMs has mostly not been measured. Engineered nanomaterials can cause damage to the vascular system in fish,166 or accumulate at the site of injection in dosimetry investigations,167 but there are no large scale defects in cardiovascular functions reported so far. There is also a concern that the aggregation of ENMs in the body fluids will result in aggregates that adsorb to the surface of the blood cells, with the latter subsequently controlling the fate of ENMs in the circulation. However, the precise fraction of ‘free’ ENMs compared to adsorbed ENMs in the body fluids has not been determined in vivo in most animals. Also, any damaged blood cells with ENMs adsorbed to their surface might be removed by the spleen of vertebrate animals (e.g., Cu NPs in fish168), and perhaps by melanisation and formation of nodules in invertebrates. Thus, the distribution of ENMs to the internal organs may be a limited portion of the original exposure dose.
Furthermore, many invertebrate species have the ability to store dissolved metals as granules by biomineralisation processes (Table 2). Fish can sometimes also make metal storage granules in the liver.169 However, these granules should not be misinterpreted as stored internalised ENMs. In the case where the ENMs dissolve to ions (e.g., in the acidic stomach pH) and the metal is subsequently taken up in the ionic form, it is expected that these ions might be mineralised and stored, especially by the invertebrates.170 Interestingly, in the case of at least one ENM where there was no dissolution (and therefore no re-mineralisation), the particles were not retained in the tissues in vivo in terrestrial isopods (CoFe2O4 NPs171). A review of the uptake and excretion kinetics of ENMs in invertebrates is reported elsewhere (see van den Brink et al.172).
The USL has the potential to concentrate ENMs, but it is also a protective layer. In animals that produce mucus, the sloughing of mucous secretions protects the underlying epithelium, and together with particle settling in the gut lumen, the overall effect is to reduce the bioavailability of most EMNs to the epithelium to a fraction of the ingested dose. Apart from a few studies on fish or rats, more measurements of the gut surface-associated fraction of ENMs is needed in order to understand how much of the potentially labile pool of surface adsorbed ENMs is available for true uptake. The effects of ENMs on mucus secretion rates and intestinal sloughing is also poorly understood. The various extracellular matrices, such as the cuticle of invertebrates, also likely serve to limit the access of ENMs to the underlying epithelial cells. However, more measurements of particle interactions with cuticle, and peritrophic membranes are needed, especially with respect to the integrity and apparent permeability of these matrices in the presence of ENMs.
The mechanisms of ENM uptake by gut epithelial cells is most well-known for mammalian models and these include active uptake by receptor-mediated endocytosis pathways and/or macropinocytosis. While these mechanisms also exist in the gut of invertebrates, it is also crucial to understand that many invertebrates have another potential pathway – the uptake of ENMs via the ‘intracellular digestion’ process including phagocytosis and pinocytosis by cells in the gut lumen, and subsequent delivery to, and across, the epithelium. The importance of this pathway for invertebrates requires further research. Rarely, for a few pristine very small lipophilic materials such as C60, diffusional uptake across the gut epithelium may occur, but this phenomenon, which has mostly been demonstrated with artificial lipid bilayers, needs to be explored in vivo in the complex chemical matrix of the gut lumen. Care also needs to be taken with electrophysiology and its interpretation. Paracellular uptake of ENMs is not likely in a healthy epithelium, and TEER measurements alone that relate to electrolyte/water flux do not inform on the integrity of the paracellular route for ENMs. Transcellular uptake via vesicular-dependent pathways remains the most likely route to cross the gut epithelium of animals. The basal lamina acts as a barrier to the serosal compartment in many animals, and in invertebrates where the gut epithelium is in direct contact with underlying loose connective tissue or muscle fibres, there are immune phagocytic cells that will remove ENMs from the extracellular fluid. Such housekeeping roles of the immune system and the propensity of particle settling in the high ionic strength of extracellular fluid will limit the distribution of ENMs in animals. The role of all the blood cells for the incidental transport and distribution of ENMs to the internal organs requires more investigation. However, at least some measurements such as single particle inductively coupled mass spectrometry (spICP-MS) are detecting ENMs inside some internal organs or tissues. Invertebrates especially, have the ability for biogenic mineral formation and any nano scale granules present in the tissue may not be the original ENMs.
Footnote |
† Electronic supplementary information (ESI) available. See DOI: 10.1039/d0en00174k |
This journal is © The Royal Society of Chemistry 2020 |