healthmedicinet

Assessment of particle size and surface charge of latex nanoparticles

The interaction of nanosized-materials with body fluids is an early event; we and
others have shown that components of extracellular fluids adsorb to the particles
31]–33]. Importantly, we recently showed that polystyrene latex nanoparticles adsorb components
of the tissue culture medium 31], which is likely to alter the surface charge and format of the NPs presented to the
cells. Here the hydrodynamic diameter and surface charge density of each surface group
(Table 1) in distilled water (DW) and in tissue culture medium (DCCM1 and RPMI) were measured.
All NPs were monodisperse in DW but formed small agglomerates in DCCM1 and RPMI (Table 1), as indicated by the increase in average hydrodynamic diameters and polydispersity
index values (PDI), likely reflecting adsorption of proteins and other components
of the medium 31]. The surface charge densities (measured as zeta potential; Table 1) of the NPs also depended on the dispersing medium; the 50 nm ANP (+43.7?±?1 mV),
CNP (?46.7?±?1.27 mV) and UNP (?50.5?±?2.56 mV) show strong positive and negative
surface charge densities in DW relating to their surface functional group (?OH for
UNP,-COOH for CNP and –NH2 for ANP; Table 1), but their surface charge became very similarly moderately negative, regardless
of their original charge, in DCCM1, as follows: ANP (?11.4?±?0.90 mV), CNP (?13.1?±?0.89 mV)
and UNP (?15.5?±?0.81 mV) and in RPMI ANP (?35.2?±?0.56 mV), CNP (?33.0?±?0.97 mV)
and UNP (?14.8?±?2.10 mV).

Table 1. Surface chemistry, hydrodynamic diameter, and surface charge density of unmodified-
(UNP), carboxyl-modified (CNP) and amine-modified (ANP) latex nanoparticles in distilled
water (DW) and tissue culture medium (DCCM1 and RPMI)

Effect of size and surface chemistry on cell viability

Previously we reported that 50 nm ANP induced cell death and pore formation at the
cell membrane of TT1 cells, together with increased release of IL-6 and IL-8 and activation
of caspase 3/7 and 9 31]. Here, we further investigated the effect of NPs on their ability to induce intracellular
reactive oxygen species (ROS) as this might impact on cell viability and bioreactivity.
The 50 nm ANPs were reported to be more toxic than 100 nm ANPs; 50 nm ANPs were also
more toxic than 50 nm UNPs and CNPs after 24 h exposure 31]. In this study, we showed that following 4 h exposure, there was only a slight toxic
effect of ANP on TT1 at concentrations between 50-100 ?g/ml (Additional file 1: Figure S1). We also investigated if the same response profile would occur with AT2
and MACs using a similar NP concentration range and 24 hour exposure. The cytotoxicity
of 50 and 100 nm NPs against TT1, AT2 cells and MACs showed a similar pattern of response,
where ANPs, but not UNPs or CNPs, were very cytotoxic (Fig. 1a-f). The 50 nm ANPs were toxic even at the lower concentrations of 10 and 25 ?g/ml (t?=?24 h, ?=?6), inducing approximately 20 % cell death, compared to 10 % cell death for the
same concentrations of 50 nm UNP and CNP and 100 nm ANPs. ANPs, 50 ?g/ml, induced
approximately 50 % cell death in MACs, compared to 30?35 % cell death (t?=?24 h, ?=?6) in the epithelial cells, although the highest concentration, 100 ?g/ml ANP caused
a similar degree of cell death in all cell types (~60 %; Fig. 1b, d and f). At the highest concentration of 100 nm NPs there was approximately 45?50 % AT2
and MAC cell death (Fig. 1d and f), although TT1 cells exhibited less than 20 % cell death at this concentration (Fig. 1b). We suspected that oxidative stress might associate with ANP toxicity, and therefore,
the addition of antioxidant N-acetyl-cysteine (NAC, 10 mM) would reduce cell death
following ANP exposure. When TT1, AT2 cells and MACs were exposed to the 50 nm NPs
(t?=?24 h) in the presence of NAC (Fig. 1g-i, Additional file 1: Figure S2), there was little effect on the UNP- or CNP-treated cells, as expected
due to little effect of the NPs alone (Additional file 1: Figure S2). Regarding the marked cell death induced by 50 nm ANP, there was no improvement
in cell viability of TT1 cells (Fig. 1g). In MACs, NAC had only a small protective effect on cell death at 10 ?g/ml ANP (Fig. 1i). However, in AT2 cells (Fig. 1h), NAC prevented the effects of 10 and 25 ?g/ml 50 nm ANP, and caused a small improvement
in the effect of 100 ?g/ml ANP on AT2 cell viability.

Fig. 1. Viability of TT1, AT2 and MAC following 24 h exposure to 50 nm and 100 nm polystyrene
nanoparticles. 50 nm ANPs, at 50-100 ?g/ml, caused significant cell death in all cell
types (a, c, e). 100 nm ANP, 50 and 100 ?g/ml, caused significant cell death in AT2 and MAC but
not TT1 cells (b, d, f). Addition of antioxidant N-acetylcysteine (NAC), 10 mM, did not prevent TT1 cell
death (g), whereas NAC protected AT2 and MAC from the effects of low concentrations of ANPs
(h–i). UNP and CNP had no effect on cell viability. *p??0.05, **p??0.001 and ?=?3 TT1 replicates and ?=?6 subject samples for AT2 and MAC

We also observed toxicity of NPs in all cell types using the lactate dehydrogenase
(LDH) assay to examine membrane integrity (?=?6, t?=?4 and 24 h; Additional file 1: Figure S3–S4). We previously observed a marked release of LDH by TT1 cells exposed
to 50 nm ANP (but not UNP or CNP) which paralleled the formation of “holes” within
the cell membrane 31]. AT2 cells exposed to all types of 50 and 100 nm ANPs released LDH (Additional file
1: Figure S3); the most significant release of LDH was on exposure to 50 nm CNPs, ANPs
and 100 nm ANPs (p??0.001, ?=?6). The release of LDH was time and NP concentration dependent; the smaller NPs
induced the highest LDH. At the higher concentrations of 50 nm ANPs (25-100 ?g/ml),
there was up to 60 % LHD release, which mirrored MTT estimation of AT2 cell death
(Fig. 1c-d). In contrast, in the MAC studies, the 100 nm NPs had the greatest effect on LDH
release compared to 50 nm NPs following 24 h exposure (p??0.001, ?=?6; Additional file 1: Figure S4), which was also concentration and time dependent. ANPs exhibited higher
toxicity than UNPs and CNPs, and there was marked release of LDH at the higher concentrations
of 100 nm ANPs (25-100 ?g/ml) causing up to 75 % LDH release following 24 h exposure,
in parallel to the MTT measure of cell death (Fig. 1e and f). Interestingly, LDH did not exactly mirror MTT measure of cell death for MAC experiments
exposed to 100 nm UNP and CNP.

Effect of nanoparticles on the release of inflammatory mediators, IL-6 and IL-8

Following 24 h exposure, all types of NPs activated significant increased release
of IL-6 and IL-8 (p??0.001, ?=?6) by AT2 cells and MAC (Additional file 1: Figure S5). The increase in IL-6 release by AT2 was much the same regardless of
NP concentration or surface modification. This may in part reflect the similarity
in surface charge density (zeta potential) between the NPs, which was similar in DCCM1
medium, between ?11.4 and ?15.5 mV (Table 1), but does not address the lack of effect of increasing NP concentration. The marked
increase at even low NP concentrations may be due to induction of the maximal IL-6
response. Neither is it clear why there are no differences between the magnitude of
the effect of 50 and 100 nm NPs, considering the marked increase in NP numbers and
surface area/unit weight of the 50 nm NPs, as discussed previously 31]. In contrast, an effect of particle size was observed for release of IL-8 (Additional
file 1: Figure S5d-f); 50 nm NPs induced a significantly greater release of IL-8 than that
of the 100 nm NPs. However, as for IL-6, there was no effect of increasing NP concentration
on AT2 cell IL-8 release. In contrast, for MACs (Additional file 1: Figure S6) increased release of IL-6 and IL-8 correlated with NP concentration.
Surface chemistry was important (p??0.001, ?=?6); ANPs and UNPs induced the greatest, similar levels of IL-6 and IL-8 release,
compared to CNPs. Again, there was little effect of NP size on IL-6 release, whereas
100 nm UNP and CNP induced more IL-8 release than did the 50 nm NPs. We previously
reported that all three types of 50 nm NPs stimulated a concentration-dependent release
of IL-6 and IL-8 by TT1 cells 31]. Others have shown that polystyrene latex stimulated IL-8 release by A549 cells;
the smallest, 60 nm NPs, caused the highest release compared to 200 and 500 nm NPs
34]. Prietl et al. reported a similar pattern of IL-8 release by macrophages exposed
to 20, 500 and 1000 nm carboxyl modified latex particles 35].

Effect of nanoparticles on the activation of intracellular reactive oxygen species
(ROS)

We used 2?, 7?-dichlorodihydrofluorescein diacetate (H2-DCFDA; to detect peroxide
and singlet oxygen) and dihydroethidium (DHE; to detect superoxide radicals) dyes
to monitor intracellular ROS. Both dyes indicated a similar pattern of ROS induction
within TT1 cells; however, only DHE effectively detected ROS in AT2 and MAC, indicating
differences in ROS production between the cells. ANPs (50 nm) significantly initiated
ROS production in TT1 cells in a concentration dependent manner at 4 h, when there
was no cell death (p??0.001, ?=?3; Additional file 1: Figure S1, S7a-d, s). The ROS detected by H2-DCFDA remained over 24 h, when there
was significant cell death, as can be seen by cell loss in Additional file 1: Figure S7f. UNPs did not initiate ROS detected by H2-DCFDA in TT1, whereas CNPs
took up to 24 h to induce ROS (Additional file 1: Figure S7). ANPs induced massive production of ROS (detected by DHE and H2-DCFDA)
in TT1 cells, p??0.001, ?=?3, at both 4 and 24 h (Fig. 2d, aa. Additional file 1: Figure S7a-d, f, s-t). Although UNPs and CNPs induced TT1 ROS, it was much lower
than that observed with ANP (Fig. 2b and c, aa; Additional file 1: Figure S7g-h, t). In contrast, all types of NPs initiated significant production
of ROS (detected by DHE) in AT2 (p??0.001, ?=?3, Fig. 2j, k, l and ab) and MACs (Fig. 2r, s, t and ac), which in AT2 cells was concentration-dependent (Additional file 1: Figure S7i-r, u). This effect was completely eliminated in AT2 cells by co-incubating
the antioxidant N-acetyl-cysteine (10 mM) with the NPs (Fig. 2n, o, p and ab), which also prevented ROS formation in TT1 cells exposed to UNP and CNP (Fig. 2f– g, aa), though only partially eliminated ROS in TT1 cells exposed to ANP (Fig. 2h, aa). In contrast, NAC had little effect when added to NP-exposed MACs, regardless of
surface modification (Fig. 2v-x, ac). Although the induction of oxidative stress was observed following NP exposure to
all types of NPs, differences were observed in the magnitude and profile of ROS activity
relating to both surface modification and cell type (Fig. 2aa, ab, ac). Although the cells were seeded at the same density, in the case of the primary
MACs not all the seeded cells adhered to the plate and the final number of MACs was
less than those for AT2 and TT1 cells. This resulted in an overall reduction in the
measured intensity of ROS in MAC (Fig. 2ac) compared with TT1 and AT2 cells (Fig. 2aa and ab).

Fig. 2. Induction of reactive oxygen species by 50 nm UNP, CNP and ANP in the absence and
presence of antioxidant N-acetylcysteine. Cells were exposed to 25 ?g/ml of 50 nm
NPs for 4 h, alone (a–d, i–l, q–t) and in the presence (e–h, m–p, u–x) of 10 mM NAC. All types of NPs induced significant ROS production in AT2 (i-l) and MAC (q–t), whereas TT1 cells (a–d) were most susceptible to ANP. NAC treatment significantly prevented NP-induced ROS
in epithelial cells (aa, e–h, ab, m–p), but not macrophages (ac, u-x). ROS were measured in live cells and data are presented as mean fluorescence intensity
(MFI)?±?SD (?=?3 TT1 replicates and 6 subject samples for AT2 and MAC) in aa-ac *p??0.05; **p??0.001

Effect of nanoparticles on glutathione flux

As we have limited numbers of primary AT2 and MAC cells, we used TT1 cells as a model
to study the effect of oxidative stress on glutathione flux. Cellular glutathione
levels (GSH and GSSG) were measured in TT1 cells at 1 and 4 h after 50 nm NP exposure
(Fig. 3a, b). After 1 h, the GSSG/GSH ratio (indicating the ratio of oxidised GSSG to reduced
GSH) was increased, 2-3-fold above non-treated control cells, for all types of NPs
(at concentrations of 50 and 100 ?g/ml) reflecting oxidative stress (Fig. 3a-b). By 4 h (Fig. 3b), control (baseline) TT1 cell GSSG/GSH ratio had increased above that observed at
1 h. There was a further, significant increase in the GSSG/GSH ratio following ANP
exposure, even at the lowest concentration of 1 ?g/ml (1.5-fold control; p??0.05, ?=?3), which increased in a concentration dependent manner, reaching 6-fold that
of unexposed cells (p??0.001, ?=?3) at 50 ?g/ml ANP; this effect was not observed with UNP or CNP. Co-incubation
of NPs with NAC for 4 h prevented the reduction of total cellular glutathione (GSH
and GSSG combined; Fig. 3c). The highly significant fall (down to ~10 % of control, p??0.001, ?=?3) in glutathione following exposure to 50 nm ANPs could be markedly prevented
by NAC treatment (down to ~67 % of control). A similar trend was also observed in
TT1 cells exposed to 100 nm NPs, although the effect of all three NPs on increased
GSSG/GSH ratio at 1 h was more noticeable than that seen following exposure to the
50 nm NPs at the same time interval (Fig. 3d). Remarkably, this effect disappeared for UNP and CNP at 4 h, but remained at very
similar levels for the ANP-exposed cells (Fig. 3e). NAC prevented the reduction of cellular glutathione activated by 100 nm NPs (Fig. 3f) to a very similar extent to that seen with 50 nm NPs.

Fig. 3. Effect of polystyrene nanoparticles on total cellular glutathione and TT1 cellular
oxidised glutathione: reduced glutathione ratio (GSSG: GSH) at 1 and 4 h exposure.
GSSG:GSH ratio following exposure to 50 nm (a, b) and 100 nm (d, e) NPs for 1 (a, d) and 4 h (b, e) respectively. Total GSH following exposure to 50 nm (c) and 100 nm (f) for 4 h with and without antioxidant N-acetylcysteine (NAC, 10 mM). ANPs caused
the most significant increases in GSSG:GSH ratio, regardless of particle size. All
three NPs caused a reduction in total cellular GSH, though this was most marked following
ANP exposure (c, f). NAC treatment significantly prevented this effect (c, f). *p??0.05, p**??0.001; ?=?3 replicates

Effect of nanoparticles on mitochondrial function and structure

Mitochondrial membrane potential and mitochondrial structure of TT1 cells following
NP exposure (Fig. 4–5) were observed to compare with changes in ROS production, using MitoTracker® fluorescent
probe, confocal microscopy and transmission electron microscopy (TEM). The MitoTracker®
probe reflected mitochondrial membrane potential of the intact mitochondria; a significant
decrease in mean fluorescence intensity (MFI), indicating reduction of mitochondrial
membrane potential in all cell types exposed to 50 nm ANPs (t?=?4 h, Fig. 4a-c; p??0.001, ?=?3 replicates TT1 and 6 subject samples AT2 and MAC). This was accompanied by mitochondrial
swelling and disruption of the mitochondrial network in ANP-exposed cells (Fig. 4–5), as shown by TEM (Fig. 4, ?=?60 observed cells) and confocal microscopy (Fig. 5, ?=?45 observed cells). Mitochondrial swelling is a pathology of mitochondria indicated
by an increase in volume of mitochondria due to the fluid influx as a result of altered
mitochondrial membrane potential. The enlarged size of the mitochondria can be seen
at the same magnification (same scale bar) as we show here in Fig. 4k, n and o, in comparison to the control cells in Fig. 4d, h and l. The structure of cristae collapse during the swelling process cannot be detected
by the osmium contrast agent when using TEM. We also investigated changes in mitochondrial
structure at the lower NP concentration range (1-25 ?g/ml) using transmission electron
microscopy (TEM) and did not see a difference compared to non-treated cells (data
not shown). In the normal healthy cells the mitochondria form a network where each
mitochondrion is linked to another as seen in control cells in Fig. 5 (the connected green fluorescent feature, control panel). Disconnection of the green
fluorescent feature indicated the disconnected mitochondria within the network (change
from green connect line to green dot), as seen following exposure to ANP (ANP, green
fluorescence panel). Cytochrome C was stained with a red fluorescent signal and when
co-localised with the mitochondrial fluorescent green signal, showed yellow. However,
in ANP-exposed cells, cytochrome C is released from the mitochondria and this can
be observed in a clear pure red fluorescent signal, indicating loss of mitochondrial
integrity and apoptosis (arrows on the right ANP column in Fig. 5). Unlike TT1 cells, all NPs induced ROS production in AT2 cells, however only ANPs
induced mitochondrial swelling and loss of mitochondrial membrane integrity, as seen
by TEM (Fig. 4k), and breakdown of the mitochondrial network (Fig. 5). This was associated with the release of cytochrome C (Cyt C indicated with arrow)
within the cells, though this was not as noticeable as that observed in TT1 cells
(Fig. 5). Again, all types of NPs induced ROS in MACs; interestingly, unlike the epithelial
cells, in MACs, both CNPs and ANPs initiated mitochondrial swelling (Fig. 4–5); however, breakdown of the mitochondrial network and release of Cyt C could only
be observed in MACs exposed to ANPs (see arrows in Fig. 5).

Fig. 4. Effect of 50 nm polystyrene nanoparticles on mitochondrial membrane potential (a–c) and structure (d–o) following 4 h exposure. ANPs caused a significant reduction in the mitochondrial membrane potential (**p??0.001, ?=?3 TT1 replicates and 6 subject samples for AT2 and MAC) of all cell types (a–c) and altered mitochondrial structures by causing mitochondrial swelling (arrows in
g, k, o) compared with the control (d, h, l). There was only slight mitochondrial swelling in MACs following CNP exposure (arrows
) not seen following UNP exposure (m). The number of total observed cells analysed/sample
was 60 (?=?60); scale bar 500nm

Fig. 5. Effect of polystyrene nanoparticles on cytochrome C (Cyt C) release and the mitochondrial
network (Mito) in TT1, AT2 and MAC. Exposure to 50 ?g/ml 50 nm UNP and CNP had no
effect on the release of Cyt C or the mitochondrial network. ANP caused disruption
of the mitochondrial network (arrows indicate breakdown of Mitochondria in green)
and initiated the release of Cyt C (arrows indicate the red area of Cyt C release)
in all cell types). Cell nuclei, mitochondrial networks and cytochrome C are stained
blue, green and red, respectively; ?=?45 cells analysed/sample

Uptake of nanoparticles by TT1, AT2 and MAC

TEM and scanning electron microscopy (SEM) were employed to observe nanoparticle-cell
interactions and particle uptake (Fig. 6, 7 and 8). We recently showed that the uptake and transport of the same set of latex nanoparticles
(t?=?4 h) through TT1 cells involved both passive and active transport depending on
their size and surface chemistry 16]. The 50 nm NPs largely entered TT1 cells via passive transport, while the 100 nm NPs entered mainly via clathrin- and caveolin-mediated endocytosis; 50 nm ANPs were internalised more rapidly
than the UNPs and CNPs 16]. 3?8 % of 50 nm UNP and CNP translocated across the TT1 monolayer, without interfering
with TT1 monolayer integrity 16]. It was demonstrated that the NPs can traverse between TT1 cells, until they reach
a tight junction 16], also shown here in Fig. 7h and which suggests that the integrity of the tight junction (white tj, arrow), and
its location, controls the translocation of these NPs between epithelial cells. The
aim of the study did not include the effect of NPs on cell monolayer integrity and
their translocation; however, work on the TT1 cell in this respect is described elsewhere
16]. In the current study, the cells were exposed to 50 ?g/ml NPs, as this concentration
exhibited very low toxicity (viability was 92?95 %, Additional file 1: Figure S1, 31]) at 4 h exposure and was a critical concentration at which a change in mitochondrial
structure was observed. 40?60 % of TT1 cells and 50?70 % of MACs internalised NPs,
whereas only 7?22 % of AT2 cells contained NPs (Fig. 6m-o). Interestingly, there was little difference between NP-functionalisation and the
number of NPs taken up by each cell type, despite marked differences in cell viability,
where ANP were most cytotoxic. Neither was there any difference between the particle
sizes. This indicates that surface charge is an important component of the cytotoxic
effect of the ANP, with the exception of TT1 cells, where 100 nm ANP-functionalised
NPs caused relatively little cytotoxicity. The uptake of NPs by AT2 cells was much
lower than that of the TT1 cells and the number of cytosolic NPs in AT2 cells was
also much lower than that of the TT1 cells (data not shown). TT1 cells internalised
all types of NPs following 4 h exposure (Fig. 6b-d, m, Fig. 7). The 50 nm and 100 nm NPs were observed within TT1 cell vesicles, suggesting active
uptake (Fig. 6c-d, Fig. 7). Particles in endo/lysosomal compartments of TT1 were in the form of agglomerates,
possibly aggregates (Fig. 6b-d). Cytoplasmic NPs were present as individual particles (Fig. 6b, c, l and Fig. 7c-e and k), suggesting passive uptake of NPs or that NPs might escape from endo/lysosomes.
Use of the LysoTracker® fluorescent probe indicated an effect of NPs on lysosomal
membrane integrity. The decrease in mean fluorescence intensity (MFI) of the probe
indicated a decrease in the number of intact lysosomes within the cells following
NP exposure. ANPs, but neither UNPs nor CNPs, caused a significant reduction in the
MFI of LysoTracker® (p??0.001, ?=?3, Fig. 6p) suggesting that the amine-surface modified NPs precipitated lysosomal membrane damage
and, subsequently, escaped to the cytoplasm possibly via a ‘proton sponge’ mechanism 36]. We previously reported that ANPs caused pore formation in the cell membrane which
may be one mechanism of passive uptake of 50 nm ANPs 31]. In addition, NPs (Fig. 7c and k) appeared to adhere to the TT1 cell membrane and penetrate into the cell cytoplasm.
The uptake of individual particles was also observed to occur at the lateral, paracellular
and cell-cell interface, where NPs had tracked between the cells, up to the tight
junction, before translocation as individual NPs across the cell membrane (Fig. 7d-e and h-i) as we previously reported 16]. It is difficult to assess whether particles that appear to be within the cytosol
are membrane-bound. We used a sample preparation and staining technique, with osmium,
uranyl acetate and lead citrate post-stain, to specifically identify membranes and
believe that any membranes, including vesicular membranes should have been apparent.
Thus, we believe that some particles appear to be free within the cytosol. And, importantly,
this latter, paracellular process was less obvious with ANP suggesting different uptake
mechanisms (Fig. 7j-k). In contrast to TT1 cells, most of which internalised all types of 50 nm NPs, only
a small proportion of AT2 cells (6?20 % from Fig. 6, 16], 29]) were found to contain NPs (Fig. 6f-g, n). Intracellular ANPs were only found in the endosomal compartment (Fig. 6h), indicating active uptake.

Fig. 6. Transmission electron micrographs of TT1, AT2 and MAC following exposure to 50 nm
UNP, CNP and ANP for 4 h. The data are presented as TEM images (a–l) and quantitatively as bar graphs (m–o). Following exposure to 50 ?g/ml 50 nm NPs, TT1 and MAC internalised all three types
of NPs, (m, o, arrows in b, c, d, j, k, l), whereas much lower numbers of AT2 cells internalised NPs (). UNP and CNP were observed adhering to AT2 cell membranes (arrows in f–g). Once internalised, all types of NPs were observed mainly in endosomal structures.
Percentage cell uptake is presented in m-o (*p??0.05, **p??0.001, ?=?3, total 90 cells analysed). The reduction in mean fluorescence intensity (MFI)
of Lysotracker® probe (p) indicates the lysosomal disruption following 4 h exposure of TT1 cells to ANPs (**p??0.001, ?=?3)

Fig. 7. Interaction and uptake of 100 nm polystyrene UNP. CNP and ANP by TT1 cells 4 h after
exposure. Following exposure to 50 ?g/ml NPs, compared to non-exposed TT1 cells (a), UNPs were taken up via endocytosis as an agglomerate (arrows in b) and as individual particles (arrows in c). UNPs were also observed paracellularly and were taken up individually (arrows in
d–e). Similar observations were made following TT1 cell exposure to CNPs (f; arrows in g indicate endocytosis and macropinocytosis). The CNPs also travelled paracellularly
(left arrow in h, right arrow indicates tight junction-tj). The ANPs were taken up via endocytosis as agglomerates (arrows in j) and individually (arrows in k), but few were observed paracellularly. The percent cell uptake of all NPs by all
cell types is shown in Fig. 6m-o in comparison with the 50 nm NPs. Scale bars in a-k are 200 nm; a total of 90 cells
were examined

Fig. 8. Interaction of UNP, CNP and ANP with the cell surface membrane of AT2 cells and MAC.
AT2 cells exposed for 4 h to 50 ?g/ml of 50 nm polystyrene UNP (c–d), CNP (e–f) and ANP (g–h) compared to the non-treated control AT2 (a–b). Both individual and aggregated forms of all types of NPs were observed on the cell
surface membrane of AT2 cells, situated amongst microvilli (arrows in d, f, h; ?=?60 cell observations). Scanning electron micrographs of control MAC (i) and MAC exposed to 50 ?g/ml of 50 nm polystyrene UNP (j–k), CNP (l–m) and ANP (–o) for 4 h. MACs were activated following exposure to UNP (j–k), CNP (l–m) and ANP (–o). Both individual and aggregated forms of NPs were observed on the cell membrane
of MACs (arrows in k, m, o). UNP triggered filopodia formation, while CNP and ANP initiated membrane blebbing.
All type of NPs were observed in association with cell surface microvilli

To observe how NPs interact at the cell surface, we used scanning electron microscopy
(SEM, Fig. 8) to observe the cell-nanoparticle interface. We previously reported that latex NP
could initiate the protruding of microvilli in live cell experiments of TT1 interactions
31], 37] using scanning ion conductance microscopy 31]. A similar effect was observed here with AT2 (Fig. 8); the presence of all types of 50 nm NPs, 50 ?g/ml, induced microvilli formation,
possibly modifying NP interaction with AT2 cells, as the microvilli protruded and
surrounded the NPs (Fig. 8a-d; 31]). Both agglomerated and individual NPs adhered to the cell membrane, amongst the
cell surface microvilli, which were more dense and co-localised with the NPs (Fig. 8d, f and h). All types of 50 nm NPs were detected within the vesicle compartments of MACs, as
agglomerates, suggesting phagocytic uptake (Fig. 6k, o). SEM indicated altered MAC morphology after exposure to NPs (Fig. 8i-o). UNP-induced changes in MAC morphology were of the classic activated macrophage,
exhibiting extensive filopodia (Fig. 8j-k); in contrast, MACs exposed to CNP and ANP were devoid of filopodia and showed blebbing
of the cell membrane (Fig. 8l-o), possibly prior to apoptotic cell death. All three NPs were found adhered to the
cell membrane of MACs (Fig. 8k, m, o).