Engineered nanomaterials (ENMs) have gained importance in the field of consumer products and nanodevices due to their unique structural, optical, physiochemical, magnetic, and surface characteristics (1). The exponential increase in production and consumption of ENMs would lead to their inadvertent release in ecosystems causing adverse impact. This has raised a global concern among the environmental scientists (2, 3). Since bacteria play an important role in the geochemical cycling, they are an appropriate model to understand the toxicity mechanism of ENMs.
To attribute the toxicity to ENMs, it is logical to assess their internalization in the cell. Transmission electron microscopy [TEM; (4)], scanning electron microscopy + backscattered electron + energy-dispersive X-ray spectroscopy [SEM + BSE + EDS (5)], confocal and fluorescence microscopy (6), reflection-based imaging (7–9), and dark field imaging (10, 11) are the commonly used methods to detect ENMs presence in the cell. Although these techniques have the advantage of tracking ENMs in the cell as well as cellular organelles, there are certain drawbacks, for example, in electron microscopy, the samples have to be fixed, hence, uptake in live cells cannot be monitored. It is also resource intensive and time consuming. Confocal and fluorescence microscopy, on the other hand, requires probe tagging or fluorescence doping of ENMs for their detection. Because the ENMs are now modified, it is likely to alter their behavior as well as bioavailability.
Earlier studies with mammalian cells have shown that the intensity of the side scatter (SSC) can be used as a parameter to detect the NPs internalization in the cells (12, 13). Therefore, in the present study, a flow cytometry method was developed and validated to detect the uptake of ENMs in live bacteria. The uptake of commercially important ENMs (ZnO and TiO2) was studied in a ubiquitous Gram-negative bacterium Escherichia coli (E. coli).
Materials and Methods
ZnO nanopowder (purity > 99%; CAS No. 1314-13-2) and titanium (IV) oxide nanopowder (TiO2; 99.7%, anatase; CAS No.1317-70-0) were purchased from Sigma Chemical Co. (St. Louis, MO). Luria Bertani (LB) broth was purchased from HiMedia Laboratories (Mumbai, India). All other chemicals were of analytical reagent grade and procured locally.
Nanoparticle stock suspension (80 μg/ml) was prepared in 0.22-μm filtered PBS. Size (hydrodynamic diameter) and zeta potential of the NPs were determined by dynamic light scattering and phase analysis light scattering, respectively, using a Zetasizer Nano-ZS, equipped with 4.0 mW, 633 nm laser (Model ZEN3600; Malvern Instrument, UK). The NPs were further characterized by TEM on an accelerating voltage of 120 kV in a JEOL model 1200EX (JEOL, Tokyo, Japan) instrument.
To prepare liver S9 fraction, adult male albino Wistar rats (200 g) were obtained from the Animal House facility of CSIR-Indian Institute of Toxicology Research, Lucknow. The animal experimentation was approved by the Institutional Animal Ethical Committee. Liver S9 fraction was prepared according to the protocol described by Maron et al. (14). S9 was used in the experiment to mimic the mutagenicity assays in bacteria where exogenous metabolic activation system from mammalian liver is added to ensure metabolism of chemicals, because bacteria do not possess a well-developed metabolic system.
E. coli strain K12 substrain DH10B was used in the study and procured from MTCC, Chandigarh, India. E. coli were cultured overnight and subcultured up to early log phase at an OD600 of 0.4–0.5 (1 × 109 cells/ml) in LB broth. The experimental design to assess the uptake of nanoparticles in E. coli was as follows:
E. coli (5 × 109 cells; concentrated 5 ml culture) were treated with different concentrations (10, 40, and 80 μg/ml) of ZnO and TiO2 nanoparticles with and without S9 in PBS for 60 min.
To assess whether the NPs are at the surface or have been taken up by the bacteria, studies were extended up to a further 90 min to account for the next few generations. To undertake this study, 100 μl of the treated cultures were reinoculated in 10-ml fresh LB media without S9 and incubated for 30, 60, and 90 min at 37°C in an environmental shaking incubator at 180 rpm.
Cells as well as nanoparticles (ZnO and TiO2 NPs) with and without S9 fraction were also run in parallel to account for the background signals.
From both the experiments, 50 μl of treated cultures was added to 950 μl PBS and then analyzed by flow cytometer (FACSCanto II; BD Biosciences, San Jose, CA) using FACSDiva 6.1.2 software (BD Biosciences). In the dot plots, X-axis reflects the FSC intensity in logarithmic scale, and Y-axis corresponds to the SSC intensity in linear scale. The gating of the data as P1 and P2 was based on SSC and FSC of the control cells and control cells + S9, respectively. This allowed us to differentiate the cells in which internalization of nanoparticles occurred (P1; with or without S9) from those where there was either no internalization or there was internalization along with adsorption (P2). The setting of regions (gate) in the dot plots was identical for both the NPs, because the basic optical and scattering property (characterized by the software Mie Tab based on Mie solution to the Maxwell equations) of NPs are almost identical. No compensation was done in the experiments.
Dead cell discrimination of the treated bacterial culture was carried out according to the protocol described by Jung et al. (15) using propidium iodide dye. Results were expressed as mean ± SEM of three experiments, and data were analyzed using rank sum test/nonparametric Mann–Whitney U test (two-tail testing) to determine significance relative to control and groups, respectively. In all cases, P < 0.05 was considered significant.
Further to confirm the presence of ENMs in bacteria, TEM was performed according to Jung et al. (15). Ultrathin sections (60 nm) were prepared using Reichert–Jung ultra microtome and stained with uranyl acetate and Reynold's lead citrate. The grids were examined under a TEM (JEOL model JEM-2100) operated at an accelerating voltage of 100 kV using a 20-μ aperture.
RESULTS AND DISCUSSION
In the present study, a rapid method for the detection of uptake of nanoparticles in bacteria, using flow cytometry, has been established and validated. The novelty lies in the fact that the uptake can be evaluated in live cells for several generations. Flow cytometry provides rapid, multiparametric, single-cell analysis with robust statistics (reduction of false negative or type II errors) due to large number of events measured per treatment in three dimensions when compared with TEM, which is time consuming and can only be done on a few number of fixed cells in two dimensions.
The mean hydrodynamic diameter of the ZnO and TiO2 NPs through DLS was 165 and 124 nm, whereas, through TEM, the diameter was 30 and 50 nm, respectively. These differences in sizes of NPs in different methods (e.g., TEM and DLS) are due to the different principles used to measure the size (16, 17). This is due to (a) DLS measures Brownian motion and subsequent size distribution of an ensemble collection of particles in solution and gives mean hydrodynamic diameter, which is usually larger than BET or TEM diameter as it includes a few solvent layers; (b) during DLS measurement, there is a tendency of particles to aggregate in aqueous state, thereby giving the average size of clustered particles and individual particles; (c) it reports an intensity weighted average hydrodynamic diameter of a collection of particles so any polydispersity of the sample will skew the average diameter toward larger particle sizes (16, 17). The percent intensity distributions of the nanoparticles measured by DLS are depicted in Figure 1.
The correlation of the FSC with the cellular size (12, 18), and SSC intensity with granularity of cells and the cellular mass are well established (12, 18, 19). The hypothesis behind the experiment is the positioning of blood cells in FSC–SSC dot-plot due to their differential forward and side-scattering property. Earlier reports showed that granulocytes scattered more light at 90° direction, due to the presence of the granules in the cytoplasm (12, 18). It is likely that NPs in the host cells serve as granules and scatter the light in a dose-dependent manner.
Our data demonstrated that E. coli cells after 60-min ZnO NP treatment exhibited an increase in SSC intensity (indicating granularity), which was more pronounced with S9 (76, 94, and 181% increase) rather than without S9 (10.5, 24.5, and 125.9% increase) when compared with the control (Figs. 2A and 2B). However, a statistically significant higher SSC intensity (139 and 203% with S9 and 128 and 198% without S9) was observed at lower TiO2 NPs exposure (10 and 40μg/ml; Figs. 2D and 2E). The dose-dependent increase in SSC intensity of treated cells can be attributed to increased internalization of ZnO and TiO2 NPs, which was confirmed by TEM (Figs. 3 and 4). There was no significant cytotoxicity observed in the cells (Table 1).
Table 1. Dead cell discrimination in E. coli exposed to ZnO and TiO2 nanoparticles (NP) using propidium iodide staining
Dead cells (%)
Dead cells (%)
Data represents mean ± SE of three experiments. The statistical analysis was done by rank-sum test/nonparametric Mann–Whitney U test (two tail testing).
3.5 ± 0.8
3.5 ± 0.8
ZnO NP (10μg/ml)
4.8 ± 1.2
TiO2 NP (10 μg/ml)
4.3 ± 1.1
ZnO NP (40μg/ml)
9.6 ± 2.1
TiO2 NP (40 μg/ml)
6.7 ± 1.2
ZnO NP (80μg/ml)
13.3 ± 2.8
TiO2 NP (80 μg/ml)
10 ± 2.1
Our observations suggest that ENMs influence the scatter parameters of bacteria either when bound to them or after getting internalized. A marked difference in the positioning of treated bacteria in dot-plots was also observed despite of their similar optical properties. The enhanced internalization of ZnO NPs with S9 is due to the protein coating of NPs by S9 (20). The comparative EDS spectra of NPs + S9 showed a significant increase in the concentration of nitrogen (N) and carbon (C) with respect to NPs alone, demonstrates the protein coating of NPs. The SSC intensity of TiO2 NPs treated cells is higher than ZnO NPs treatment in all experimental conditions. The plausible reason may be that TiO2 NPs particles were of a smaller size when compared with ZnO that have their enhanced internalization in the bacterial cells irrespective to the experimental conditions.
In the multigeneration study, the extended culture showed a time-dependent decrease in the granularity of the E. coli cells. This can be attributed to the fact that the granularity of cells on account of ENPs will get reduced on cell division. This hypothesis was substantiated by our data that showed a time-dependent decrease in the intensity of the SSC (Figs. 2C and 2F) in all the experimental conditions except the ZnO NPs treatment without S9. It is possible that due to the size, ZnO NPs may not have entered into the cells in large numbers without S9 and washed off from the cell surface after 30 min incubation in fresh media.
In summary, we present a rapid method to detect the uptake of NPs in bacteria, which is cost effective, easy to perform, and more effective than the conventional methods of NP uptake. It will open up new avenues in nanoparticle research to the biological perturbations observed with the internalization of NPs. This also has wide range implications in assessing the safety/toxicity of NPs used in consumer and therapeutic products, which may ultimately come in contact with resident bacteria of humans or environment.