Introduction

Gold nanoparticles (Au NPs) are used in different applications such as commercial products (Eustis and El-Sayed 2006), diagnostics and therapeutics (Sokolov et al. 2003), chemical and biological sensing (Anker et al. 2008), drug delivery (Jain et al. 2007), cancer treatment (Wu et al. 2011), and immunoassays (Alkilany and Murphy 2010). It is estimated that global investment in gold has reached about US$12 billion in 2015 (Keel et al. 2010). Due to their increased usage, discharge of Au NPs (from the products and effluents) into the environment will impose a hazard to the aquatic ecosystem and also to the other organisms.

Microorganisms form the base of the food web, which helps in the maintenance of the ecosystem as a producer and also a decomposer (Schaechter et al. 2006; Osler and Sommerkorn 2007). The toxicity of Au NPs on Pseudomonas flourescens differed depending on the surface capping agents. Polyvinyl pyrrolidone (PVP)-capped Au NPs showed a measurable toxic effect than citrate-capped Au NPs (Nur 2013). As a primary producer, algae help in the maintenance of the aquatic food web (Aruoja et al. 2009). Hoecke et al. (2013) reported the eco-toxic effect of two types of polymer-coated gold NPs on algal species (Pseudokirchneriella subcapitata). A recent report on Chlamydomonas reinharditii highlighted the importance of surface capping on the toxicity of Au NPs (Behra et al. 2015).

Chuang et al. (2013) studied the cytotoxic effects of gold nanoparticles in five different types of mammalian cell lines such as AGS (human gastric adenocarcinoma cells), A549 (human lung adenocarcinoma epithelial), NIH3T3 (mouse embryonic fibroblast), PK-15 (porcine kidney), and Vero (African green monkey kidney). Biologically synthesized Au NPs exhibited lower toxicity on A549 cells than chemically synthesized Au NPs (Sathishkumar et al. 2014). Literature is replete with reports on the size- and shape-dependent toxic effects of gold nanoparticles in different cell lines (Chueh et al. 2014; Coradeghini et al. 2013) and murine models (Schulz et al. 2012). The results of Au NP toxicity are often conflicting in nature since different investigators use different cell lines, diverse sizes of nanoparticles, special surface groups, different doses, and different sampling points.

Reports also suggest that the in vivo toxic effect of nanoparticles is different from that of an in vitro study (Schipper et al. 2008; Sayes et al. 2007). Sayes et al. (2007) compared the in vitro and in vivo toxicity effects of nano-fullerenes (C60 and C60(OH)24) on the lung tissues of rats and found them to be distinct. Therefore, it is necessary to study the in vivo toxicity of Au NPs. Yang et al. (2014) examined the in vivo toxic effects of Au NPs based on the nanoparticle size and gestational age. Zhang et al. (2011) reported the size-dependent (5, 10, 30, and 60 nm) in vivo toxicity of PEG-coated gold nanoparticles in mice. A genotoxic effect of different sizes of gold nanoparticles was observed by Schulz et al. (2012) in the lungs of rats. Yah (2013) discussed various in vivo and in vitro toxicity effects of gold nanoparticles in relation to their physiochemical properties.

In the absence of comprehensive prior reports, a study was warranted for Au NP toxicity across trophic levels using various model organisms. The purpose of the current study was to assess the in vitro and in vivo toxicity of Au NPs, on different trophic organisms, affected by their size and surface functionalization. To investigate the size effect of Au NPs, two different sizes of citrate-capped Au NPs (CIT30 and CIT40) were selected and compared in this study. Since CIT30 was found to be the more potent toxicant between the two, it was modified with the polymer polyvinyl pyrrolidone (PVP) to make the Au NP less toxic as well as determine the toxicity of the PVP coating alone, if any. The toxicity of Au NPs at different trophic levels was assessed using freshwater bacteria, Bacillus aquimaris; freshwater green algae, Chlorella sp.; a human SiHa cell line, and a mice model.

Materials and methods

Chemicals

Tetra aurochloric acid (HAuCl4), trisodium citrate, and polyvinylpyrrolidone (PVP, K-30) were purchased from Sisco Research Laboratories Pvt Ltd (Mumbai, India). BG-11 broth; nicotinamide adenine dinucleotide, reduced (β-NADH); and Eagle’s minimal essential medium (Eagle’s MEM) were procured from Himedia Labs Pvt Ltd (Mumbai, India). 2′,7′-Dichlorofluorescin diacetate (DCFH-DA), MTT, and sodium pyruvate were obtained from Sigma-Aldrich (St. Louis, MO, USA). All the other chemicals used were of analytical grade.

Synthesis of Au NPs

Citrate-capped gold nanoparticles (Au NPs) of two different sizes (30 and 40 nm; herein referred to as CIT30 and CIT40, respectively) and one polyvinyl pyrrolidone (PVP)-capped Au NP (herein referred to as CIT30–PVP capped) were prepared and further used for the toxicity assessment of Au NPs in various organisms.

CIT30 Au NPs

CIT30 Au NPs were synthesized by following the standard Turkevich protocol with minor modifications (Turkevich et al. 1953; Rohiman et al. 2011). 23.75 mL of Milli-Q water was heated in a round-bottom flask until the temperature reached 90 °C. One milliliter of 136 mM trisodium citrate and 250 μL of 25.4 mM tetra aurochloric acid were added to the solution under continuous stirring condition. Then, the solution was further heated for about 10 min till a wine-red color was obtained which indicates the formation of Au NPs. The synthesized Au nanoparticle solution was allowed to cool at room temperature and stored at 4 °C.

CIT40 Au NPs

CIT40 Au nanoparticles were obtained by a Turkevich method (Turkevich et al. 1953) with slight modification as per the procedure followed by Kumar et al. (2006). 0.25 mM tetra aurochloric acid was added to 25 mL of Milli-Q water, which was previously heated at 90 °C on a hot plate in a stirring condition. Then, 250 μL of 30.4 mM trisodium citrate was added quickly to the reaction mixture. An immediate color change from yellow to wine red was obtained within 10 min. The solution was further allowed to continue the reaction at the same boiling point for about 15 min to ensure that the reaction is fully completed. Then, the solution was allowed to cool to room temperature and stored at 4 °C.

CIT30–PVP capped Au NPs

PVP-capped Au NPs (CIT30–PVP capped) were prepared by coating PVP (1.25 mM) on the surface of CIT30 Au NPs. After the synthesis of CIT30 Au NPs (red color formation), 1 mL of 1.25 mM PVP was added and the solution was stirred for 30 min, allowed to cool to room temperature, and stored at 4 °C.

Characterization of Au NPs

The primary size and shape of the synthesized Au NPs were examined using transmission electron microscopy (TEM, FEI Tecnai T20 S-TWIN TEM). The hydrodynamic size of the Au NPs was analyzed by dynamic light scattering analysis (DLS, 90 Plus Particle Size Analyzer, Brookhaven Instruments Corp., USA). Similarly, the hydrodynamic size distribution of Au NPs (1 mg/L of CIT30, CIT40, and CIT30–PVP capped) in sterile filtered lake water and cell culture media (Eagle’s MEM) was evaluated to determine the stability of Au NPs in their respective matrices at different time intervals (0, 6, 12, and 24 h) by DLS.

Toxicity of Au NPs towards lower organisms

Toxicity assessment of Au NPs on bacteria

Test organism

The bacterial strain B. aquimaris, a gram-variable bacterium isolated from VIT lake at VIT University, Vellore Campus, Tamil Nadu (as detailed in our previous report, Kumar et al. 2014), was used as a test organism. The toxicity experiments were carried out in sterile filtered lake water as a freshwater matrix. Lake water collected from VIT Lake was filtered through blotting paper and then with Whatman no 1, followed by sterilization to get rid of the biological disturbances such as larger colloids, suspended particles, etc. This sterile lake water was further used for the toxicity studies as a freshwater matrix.

Experimental setup

Bacterial cells with an initial cell count of 215 × 107 were interacted with three types of Au NPs (CIT30, CIT40, and CIT30–PVP capped) at various concentrations of 0.25, 0.5, and 1 mg/L in sterile filtered lake water at room temperature under visible light condition. After the period of exposure (2, 4, and 6 h), cell viability was assessed using the standard plate count assay. The percentage reduction in the cell viability of the treated samples was calculated with respect to the control.

Toxicity assessment of Au NPs on algae

Test organism

For algal toxicity experiments, the freshwater green alga Chlorella sp., a single-celled, green, and spherical-shaped microalga isolated from VIT lake, was used as the test organism (as detailed in our previous report, Iswarya et al. 2015). BG-11 broth, a specific growth medium, was used for growth and maintenance of the culture. A day/light rhythm of 16 h/8 h was maintained for the growth of algal cultures, and illumination was provided by white fluorescent lights (TL-D Super 80 linear fluorescent tubes, Phillips, India) with an intensity of 3000 lx. All the toxicity studies were carried out in a freshwater matrix, i.e., sterile lake water as per OECD guidelines (OECD, 2011).

Experimental setup

At the exponential phase, algal cells were harvested from the medium by centrifugation (7000×g, 10 min, 4 °C) and washed with sterile lake water. Then, the algal cells with an initial cell count of about 5 × 105 cells were prepared in sterile lake water and treated with Au NPs (CIT30, CIT40, and CIT30–PVP capped) at various concentrations of 0.25, 0.5, and 1 mg/L under visible light condition for 72 h.

Cell viability assessment

After the interaction period of 72 h, the number of cells in the treated and control samples was counted using a hemocytometer under a phase contrast microscope (Zeiss Axiostar Microscope, USA). The percentage decrease in cell viability of the treated cells was calculated with respect to the control.

Toxicity assessment of the capping solution used for Au NP synthesis

Toxicity of the capping solutions used in the Au NP synthesis was analyzed to check the toxic effect of capping materials alone. Both the bacterial and algal cells were interacted with different capping agents such as citrate and PVP, similarly as done for Au NPs. The cell viability assay was performed for the capping agents, viz. 136 mM citrate (CIT30 Au NP), 30.4 mM citrate (CIT40 Au NP), and 1.25 mM PVP (CIT30–PVP-capped Au NP), after interacting for 6 and 72 h for bacteria and algae, respectively. The percentage reduction in the cell viability was evaluated with respect to the control.

Biochemical assays

ROS assessment

The fluorescence probe, DCFH-DA, was used to quantify the generation of reactive oxygen species (ROS). Intracellular ROS generation was measured in both control and Au NP (1 mg/L)-treated cells after the interaction period (6 h for bacteria and 72 h for algae), according to the method of Wang and Joseph (1999) with minor modifications. Five milliliters of cell suspension was incubated with DCFH-DA with a final concentration of 100 μM at 37 °C for 30 min. The fluorescence intensity of the DCFH-DA dye was measured using fluorescence spectroscopy (Cary Eclipse Fluorescence Spectroscopy; Model No: G9800A) with excitation and emission wavelengths of 485 and 530 nm, respectively.

Lactate dehydrogenase assay

The extracellular lactate dehydrogenase (LDH) release was measured as an indicator of membrane permeability and cytotoxicity as per the method reported by Dalai et al. (2012). Bacterial and algal cells were treated with 1 mg/L of Au NPs (CIT30, CIT40, and CIT30–PVP capped) for about 6 and 72 h, respectively. After exposure, the cell suspension was centrifuged (7000×g, 10 min) and the LDH released in the supernatant was measured by following standard protocol. The reaction mixture contained 100 μL of the supernatant, 100 μL of 30 mM sodium pyruvate, and 100 μL of 6.6 mM NADH. LDH activity was determined as the rate of decrease in absorbance at 340 nm (UV-vis spectroscopy, Model U2910, Hitachi, Japan).

Toxicity of Au NPs towards higher organisms

Toxicity assessment of Au NPs on human cell line

Cancer cell maintenance

The SiHa cell line used in this study was a kind gift from Dr. Chinmay Panda, Chittaranjan National Cancer Institute, Kolkata, India. It was an HPV16 containing cervical cancer cell line first isolated from a Japanese patient. The cells were maintained in Eagle’s MEM containing sodium pyruvate, glutamine, and non-essential amino acids supplemented with 10 % FBS (Gibco) and 1× antibiotic-antimycotic solution (HiMedia) in a humidified incubator at 37 °C with 5 % CO2.

Cytotoxicity assay

Cytotoxicity of the gold nanoparticles in SiHa cells was assessed by standard MTT reduction assay (Mossman 1983). MTT is a yellow water-soluble tetrazolium dye that is reduced to purple-colored water-insoluble formazan crystals by mitochondrial dehydrogenases of viable cells. 103 cells were seeded in a 96-well plate and incubated overnight. Cells were then treated with Au NPs in a concentration range of 0.1 to 0.5 mg/L with regular increments of 0.1 mg/L for 24 h. Then, 100 μL of MTT (5 mg MTT/mL PBS) was added to each well, mixed, and incubated in a CO2 incubator for another 3 h. Media were removed from the wells, and the formazan crystals were dissolved with 100 μL of DMSO (Merck). Absorbance was recorded in a microplate reader (BioRad) at 595 nm. From the absorbance data, survival percentages in treated and control sets were calculated.

DNA fragmentation

DNA laddering is a hallmark of apoptotic cell death. For DNA fragmentation assay, DNA was isolated according to the method of Herrmann et al. (1994) with modification. SiHa cells were seeded in 60-mm petri dishes and incubated overnight. The cells were interacted with 1 mg/L of Au NPs (CIT30, CIT40, and CIT30–PVP capped) for 24 h. After the interaction, the cells were harvested, washed with phosphate-buffered saline, and pelleted by centrifugation. An equal amount of 2× lysis buffer (200 mM NaCl, 20 mM Tris, 50 mM EDTA, 1 % SDS, and 2 % NP-40) was added to the pellet, mixed well, and incubated on ice for 30 min. This was followed by extraction with phenol-chloroform-isoamyl alcohol (25:24:1). Nucleic acids were collected in the aqueous phase and precipitated with isopropanol and sodium acetate. The pellet was dissolved in TE buffer and incubated with RNase A. Finally, DNA was precipitated with isopropanol and sodium acetate. Collected pellets were washed in 70 % ethanol and air-dried. DNA was dissolved in TE buffer and separated by electrophoresis on 1.5 % agarose gels.

Toxicity assessment of Au NPs on mice

Animals

Healthy male Swiss albino mice (4–6 weeks old) weighing 25–30 g were purchased and acclimatized for a week in the institutional animal house. The animals belonging to the same litter (for each of the three nanoparticles and capping agents) were randomly housed in cages (seven animals per cage for each group of concentrations), maintained under standard laboratory conditions (temperature of 22 ± 2 °C; humidity of 50–70 %, and 14 h:10 h light and dark cycle). They were kept in cages with autoclaved paddy husk for bedding and raised on standard rodent pellets (consisting of crude protein and fiber) and drinking water ad libitum throughout the study. The ethical clearance for the use of animals in the study was obtained from the institutional animal ethics committee of Calcutta University.

Treatment schedule

The mice were divided into seven groups, randomly consisting of seven animals each. They were gavaged for seven consecutive days with solutions of CIT30, CIT40, and CIT30–PVP capped Au NPs at doses of 0.01, 0.05, and 0.1 mg/kg body weight and capping agents (citrate 30.4 and 136 mM and PVP 1.25 mM) at 0.1 mL/10 g body weight (b.w.). The stock solutions Au NPs were diluted with Milli-Q water to get the following group of doses:

Group I:

Mice was administered with Milli-Q water as vehicle control.

Group II:

Mice were administered with Au NP solution (0.01 mg/kg b.w.).

Group III:

Mice were administered with Au NP solution (0.05 mg/kg b.w.).

Group IV:

Mice were administered with Au NP solution (0.1 mg/kg b.w.).

Groups V–VII:

Mice were administered with solutions of capping agent (30.4 or 136 mM citrate or 1.25 mM PVP).

The animals were sacrificed by cervical dislocation on the 8th day, and the target organ, liver, was processed to evaluate oxidative stress and DNA damage.

Evaluation of DNA damage in liver cells by single-cell gel electrophoresis (comet assay)

The alkaline comet assay was assessed following the method described in Tice et al. (2000) with minor alterations (Manivannan et al. 2013). Briefly, the liver tissue was minced in 0.075 M NaCl solution containing 0.024 M Na2EDTA (pH 7.5). The cell suspension was interacted with 1 % low-melting-point agarose. Then, it was layered on the slide base that was coated with 1 % normal-melting-point agarose. The slides thus prepared were placed in the chilled lysing solution (2.5 M NaCl, 100 mM Na2EDTA, 10 mM Trizma, 10 % DMSO, and 1 % Triton X-100, pH 10.0) for 1 h at 4 °C. Then, the slides were treated with chilled alkaline solution (300 mM NaOH and 1 mM Na2EDTA, pH > 13) for 20 min to DNA unwinding. After that, electrophoresis was performed at 0.7 V/cm and 300 mA using a freshly prepared electrophoresis buffer (1 mM EDTA disodium salt and 300 mM NaOH) at 4 °C for 25 min. After electrophoresis, the slides were neutralized with Tris buffer (400 mM, pH 7.4). The slides were stained with ethidium bromide (EtBr, 20 μg/mL) and stored at 4 °C in a humidified slide box until scoring. The slides were scored at a final magnification of ×400 using an image analysis system (Komet 5.5, Kinetic Imaging, Andor Technology, Nottingham, UK) attached to a fluorescence microscope (Leica, Germany), equipped with an attached CCD camera. Images from 150 random cells (per animal) were analyzed. The comet parameters used to measure DNA damage in the cells were tail DNA (%).

Evaluation of oxidative stress in liver

Oxidative stress was indirectly measured by evaluating the amount of lipid peroxidation (LPO), reduced glutathione (GSH), glutathione S-transferase (GST), and catalase (CAT) in the metabolically active liver cells. The respective absorbances were read with a UV 2300 II Spectrophotometer (Techcomp Ltd., Hong Kong).

Lipid peroxidation: The lipid peroxidation level in the liver homogenates was measured using the levels of malonaldehyde (MDA), which is the end point of lipid peroxidation (Buege and Aust 1978).

GSH: The GSH level was estimated by the method of Sedlak and Lindsay (1968).

GST: The activity of GST was measured according to the method of Habig et al. (1974).

CAT: Catalase activity was monitored as per the procedure given by Abei (1984).

Protein estimation: Protein quantification in liver tissue homogenate was carried out by the method of Bradford (1976) with bovine serum albumin as standard.

More details of the above methods can be found in the Supplementary Information.

Statistical analysis

All the cytotoxicity and biochemical assays were repeated at least three times, and the data are presented as mean ± standard error (SE). For the bacterial and algal cell viability assessments and also for the stability studies, the statistical analysis was performed using GraphPad Prism, Version 5.0, using two-way ANOVA (p < 0.01). One-way ANOVA (Tukey multiple comparison test) was followed for the other assays (p < 0.05). The data for the in vivo assays were analyzed using the Statistical Programme Sigma Stat 3.0 (SPSS Inc., Chicago, IL, USA).

Results

Characterization of Au NPs

Transmission electron micrographs of citrate- (CIT30 and CIT40) and PVP-capped (CIT30–PVP capped) Au NPs revealed that all the three types of NPs had a spherical shape (Fig. 1). The sizes were observed to be 24, 40, and 17 nm for CIT30, CIT40, and CIT30–PVP capped Au NPs, respectively. The hydrodynamic sizes of the Au NPs (CIT30, CIT40, and CIT30–PVP capped) were analyzed using dynamic light scattering and represented by the effective diameter. The effective diameters of the citrate-capped Au NPs (CIT30 and CIT40) in Milli-Q water were found to be 29.01 ± 2.42 and 43.11 ± 0.48 nm, respectively. PVP-capped Au NPs (CIT30–PVP capped) showed an effective diameter of about 65.23 ± 3.64 nm.

Fig. 1
figure 1

Transmission electron microscopy (TEM). TEM micrograph of Au NPs: a CIT30 Au NPs, b CIT40 Au NPs, and c CIT30–PVP capped Au NPs

Full size image

The hydrodynamic sizes of the Au NPs in sterile filtered lake water and cell culture media were studied at different time intervals (0, 6, 12, and 24 h) (Fig. 2a, b). Significant (p < 0.01) change in the effective diameter of Au NPs was observed in the sterile lake water for CIT30 and CIT40 Au NPs. The effective diameter of CIT30 Au NPs increased from 170.63 ± 5.71 nm at 0 h to 2017.26 ± 23.48 nm after 24 h. Likewise, CIT40 Au NPs aggregated from 53.54 ± 2.05 nm at 0 h to 997.39 ± 6.39 nm after 24 h at a slower rate than CIT30 Au NPs. Among the citrate-capped Au NPs, CIT30 Au NPs aggregated more rapidly than CIT40 Au NPs in sterile filtered lake water. In contrast, CIT30–PVP capped Au NPs were found to be highly stable in the sterile lake water, till 24 h (97.35 ± 0.40 nm). It was also observed that CIT30–PVP capped Au NPs were more stable than CIT30 Au NPs. The effective diameter of Au NPs in the cell culture media remained unchanged for all the three Au NPs (CIT30, CIT40, and CIT30–PVP capped), during the experimental period of 24 h. In comparison with CIT30 and CIT40 Au NPs, the effective diameter of CIT40 Au NPs in the cell culture media was found to be smaller than the size of CIT30 Au NPs. This effect could be due to the higher reactivity of lower sized nanoparticles to interact more readily with the components present in the cell culture media.

Fig. 2
figure 2

Stability study of Au NPs in the different experimental matrix used in the study. Aggregation profile of different Au NPs (CIT30, CIT40, and CIT30–PVP capped Au NPs) in the a sterile and filtered lake water and b cell culture media (Eagle’s MEM). Asterisks (*) indicate that the size of the NPs at 6 h was statistically significant with respect to particle size at 0 h (p < 0.01). The symbol # indicates that the size of the NPs at 12 h was statistically significant with respect to particle size at 6 h (p < 0.01). The symbol δ indicates that the size of the NPs at 24 h was statistically significant with respect to particle size at 12 h (p < 0.01)

Full size image

Toxicity of Au NPs towards lower organisms

Toxicity assessment of Au NPs on freshwater bacteria

The toxic effect of the three types of Au NPs (CIT30, CIT40, and CIT30–PVP capped) and capping agents was analyzed on the freshwater bacterial isolate, B. aquimaris. Particle size-, exposure period-, and concentration-dependent decrease in the cell viability were observed for the citrate-capped Au NPs (CIT30 and CIT40) (Fig. 3a). The viability of B. aquimaris cells decreased significantly (p < 0.01) with increasing period of exposure. With a decrease in particle size, the toxicity of citrate-capped Au NPs increased (CIT30 > CIT40). The cell viability decreased significantly with respect to the exposure period and concentration of PVP-capped Au NPs (CIT30–PVP capped). Between CIT30 and CIT30–PVP capped Au NPs, CIT30 Au NPs showed higher toxicity than CIT30–PVP capped Au NPs. It clearly shows that surface capping (PVP) reduced the cytotoxicity of gold nanoparticles.

Fig. 3
figure 3

Cytotoxicity assessment of Au NPs on Bacillus aquimaris . Cell viability of Bacillus aquimaris on exposure to a different concentrations of different Au nanoparticles at various exposure periods (2, 4, and 6 h), * significant with respect to control, p < 0.01, and b capping agents after 6 h interaction, * significant with respect to control, p < 0.05

Full size image

Capping agents used in the synthesis such as citrate (136 and 30.4 mM) and PVP (1.25 mM) were considered as a vehicle control, and their toxic effect has been evaluated similarly. Among the three capping controls, 1.25 mM PVP solution caused a significant (p < 0.05) reduction in the viability of cells than citrate solution (Fig. 3b).

The generation of reactive oxygen species in bacteria after an interaction (6 h) with 1 mg/L of Au NPs (CIT30, CIT40, and CIT30–PVP capped) was evaluated by dichlorofluorescein diacetate (DCFH-DA) (Fig. 4a). The increase in fluorescence intensity was statistically significant (p ≤ 0.05) and related to the particle size of citrate-capped Au NPs. CIT30–PVP capped Au NPs showed less ROS production than CIT30 Au NPs.

Fig. 4
figure 4

Biochemical assays revealing the toxic mechanism by the Au NPs on Bacillus aquimaris. a Oxidative stress (ROS) assessment of bacterial cells after 6 h interaction with various Au NPs (1 mg/L), b LDH release after 6 h interaction of Bacillus aquimaris with various Au NPs (1 mg/L). * Significant with respect to control, p < 0.05

Full size image

The cell membrane integrity was compromised by the Au NPs as revealed in the LDH assay (Fig. 4b). The release in the LDH content of 1 mg/L of Au NPs in bacteria increased significantly (p < 0.05) than that of the control. ROS production and cell membrane damages were higher in CIT30 Au NPs than those in other Au NPs.

Toxicity assessment of Au NPs on freshwater green algae

The effect of Au NPs (CIT30, CIT40, and CIT30–PVP capped) and capping agents was studied on the freshwater algal isolate Chlorella sp. On exposure to CIT30 and CIT40 Au NPs (1 mg/L), the cell viability was found to be 46.84 ± 0.69 and 52.46 ± 0.77 %, respectively. Toxicity was found to be particle size and concentration dependent (Fig. 5a). Similar to bacteria, the highest toxicity was observed for CIT30 Au NPs than other sizes. Upon exposure to CIT30–PVP capped Au NPs, the cell viability was found to be about 64.56 ± 3.29 % at the concentration of 1 mg/L. As in bacteria, CIT30 Au NPs showed higher toxicity than CIT30–PVP capped Au NPs in Chlorella sp.

Fig. 5
figure 5

Cytotoxicity evaluation of Au NPs on Chlorella sp. Cell viability of Chlorella sp. on exposure to a different concentrations of different Au nanoparticles, i.e., CIT30, CIT40, and CIT30–PVP capped Au NPs after 72 h, * significant with respect to control, p < 0.01, and b capping agents after 72 h interaction, * significant with respect to control, p < 0.05

Full size image

Cell viability in the treatment with the capping agents was 92.41 ± 1.52, 88.78 ± 3.66, and 72.51 ± 2.43 % for 30.4 mM citrate, 136 mM citrate, and 1.25 mM PVP, respectively. The PVP solution (1.25 mM) induced a significant reduction in viability (p < 0.05) than the citrate solutions (Fig. 5b).

Reactive oxygen species (ROS) generation for treated and control algal cells is represented in Fig. 6a. Significant ROS production (p < 0.05) was observed in CIT30 and CIT40 Au NPs. Analogous to the cell viability results, the ROS production increased with the decrease in particle size. Compared to the control, ROS generation induced by CIT30–PVP capped Au NPs was found to be insignificant.

Fig. 6
figure 6

Biochemical assays revealing the toxic mechanism by the Au NPs on Chlorella sp. a Oxidative stress assessment after 72 h interaction of algal cells with various Au NPs (1 mg/L). b LDH release after 72 h interaction of algal cells with various Au NPs (1 mg/L). * Significant with respect to control, p < 0.05

Full size image

The release of the cytoplasmic enzyme, LDH, in the treated and control algal cells is represented in Fig. 6b. LDH release was found to be significantly high (p < 0.05) in CIT30–PVP capped-interacted algal cells than the control. In CIT30 and CIT40 Au NP-interacted algal cells, LDH release was found to be insignificant.

Toxicity of Au NPs towards higher organisms

Assessment of toxicity of Au NPs on SiHa cell line

The cytotoxicity induced by the capping agents and Au NPs was determined by MTT assay in the SiHa cell line. The capping agents used in the present study did not induce any cytotoxicity in SiHa cells (Fig. 7a). At an exposure period of 24 h, the survival rate of the SiHa cells exposed to CIT30 Au NPs was significantly less than that of the control at the concentration 0.5 mg/L (Fig. 7b). With increasing concentration (0.1 to 0.5 mg/L) and period of exposure (48 and 72 h), the survival rate of cells decreased gradually and was statistically significant in CIT30, CIT40, and CIT30–PVP capped Au NP sets (Fig. 7c, d). In general, the survival rate decreased with increasing concentration of the Au NPs and the period of exposure used. A higher impairment of cell viability was noted for CIT30 Au NPs. The suppressive effect of Au NPs was indirectly associated with their sizes in SiHa cells.

Fig. 7
figure 7

Cytotoxicity evaluations of Au NPs on SiHa cell line. a Cell viability assessment with different capping agents, b SiHa cells exposed to Au NPs for 24 h, c SiHa cells exposed to Au NPs for 48 h, and d SiHa cells exposed to Au NPs for 72 h. * Significant with respect to control, p ≤ 0.05

Full size image

Distinct genomic DNA bands without any laddering were observed (Fig. 8). The absence of DNA laddering supports that there was no apoptotic cell death of the SiHa cells.

Fig. 8
figure 8

DNA fragmentation assay. Agarose gel showing the intact genomic DNA bands. Lane 1: 100-bp marker, lane 2: control, lane 3: CIT30 Au NPs, lane 4: CIT40 Au NPs, lane 5: CIT30–PVP capped Au NPs

Full size image

Toxicity assessment of Au NPs on mice model

DNA strand breaks in liver cells of mice after oral treatment with Au NPs and the different capping agents used for the synthesis are presented in panels a and b of Fig. 9, respectively. Of the two sizes of citrate-capped Au NPs, the percent tail DNA was significantly higher in CIT30 Au NPs than that in CIT40 Au NPs. Functionalizing CIT30 Au NPs with PVP also caused an increase in DNA strand breaks leading to greater DNA migration out of the nucleus into the tail of the comet. In general, exposure to various concentrations of Au NPs (0.01, 0.05, and 0.10 mg/kg body weight) enhanced DNA fragmentation. The induction of DNA damage was directly dependent on the size of the Au NPs and their concentration in solution. With respect to the vehicle control, citrate capping (136 mM) induced significantly more DNA strand breaks.

Fig. 9
figure 9

DNA strand break assay on mice model. Assessment of DNA strand breaks in liver cells of mice exposed to treatment with a different sizes and concentrations of gold nanoparticles (Au NPs). b Capping agents. * Significant with respect to control, p ≤ 0.05

Full size image

Effects of exposure to different types of Au NP caused an alteration of the oxidative stress-related end points in mice hepatocytes. The results of lipid peroxidation are represented in Fig. 10a, b. As compared to the control, MDA level was markedly increased by citrate-capped Au NPs (CIT30 and CIT40). The lipid peroxide level was significantly increased in the treatment with citrate-capped Au NPs, CIT30, at the concentration of 0.05 mg/kg b.w. and with CIT40 at concentrations of 0.01 and 0.05 mg/kg b.w. Citrate as a capping agent has caused a statistically significant rise in the level of MDA content in the tissue. Administration of PVP (1.25 mM) and functionalizing CIT30 Au NPs with PVP did not generate much reactive oxygen species. Hence, there was no significant rise in cellular MDA content with respect to the control.

Fig. 10
figure 10

Lipid peroxidation assay. Assessment of lipid peroxidation in liver cells of mice exposed to treatment with a different sizes and concentrations of gold nanoparticles (Au NPs). b Capping agents. * Significant at p ≤ 0.05 with respect to control

Full size image

A decrease in GSH corresponds to increased oxidative stress due to ROS generation. In the present study, GSH levels decreased in cells exposed to the lowest concentration (0.01 mg/kg b.w.) of CIT40 Au NPs and at all concentrations for CIT30–PVP capped Au NPs (Fig. 11a). The consumption of glutathione for scavenging of ROS did not follow a dose-dependent trend for the various concentrations of CIT30 Au NPs. Citrate (30.4 mM) and PVP (1.25 mM) did not induce a positive response (Fig. 11b).

Fig. 11
figure 11

GSH activity. Assessment of GSH activity in liver cells of mice exposed to treatment with a different sizes and concentrations of gold nanoparticles (Au NPs). * Significantly higher with respect to control at p ≤ 0.05; # significantly lower with respect to control at p ≤ 0.05. b Capping agents. * Significant with respect to control at p ≤ 0.05

Full size image

The GST activity following exposure to different sizes and concentrations of gold nanoparticles (CIT30 and CIT40) did not show a dose-dependent change except for CIT30–PVP capped, where the level of GST has decreased significantly at all the concentrations with respect to the control (Fig. 12a). PVP (1.25 mM) as capping agent has caused a statistically significant decrease in GST activity (Fig. 12b). The cellular responses (in terms of GSH) to the various concentrations of citrate capping were more or less similar to that of the control ones.

Fig. 12
figure 12

GST activity. Assessment of GST activity in liver cells of mice exposed to treatment with a different sizes and concentrations of gold nanoparticles (Au NPs). b Capping agents. # Significantly lower with respect to control at p ≤ 0.05

Full size image

The catalase activities of Au NPs (CIT30, CIT40, and CIT30–PVP capped) are represented in Fig. 13a. There was a significant rise in catalase activity in cells exposed to Au NPs (CIT30–PVP capped) at the concentration 0.05 mg/kg b.w. Catalase activity remained unaffected in CIT30 Au NPs and decreased significantly (when compared to the control) in liver cells exposed to CIT40 Au NPs (at concentrations of 0.05 and 0.1 mg/kg b.w.). The catalase activity of the capping agent PVP was significantly high (Fig. 13b).

Fig. 13
figure 13

Catalase activity. Assessment of catalase activity in liver cells of mice exposed to treatment with a different sizes and concentrations of gold nanoparticles (Au NPs). b Capping agents. * Significantly higher with respect to control at p ≤ 0.05; # significantly lower with respect to control at p ≤ 0.05

Full size image

Discussion

There are a significant number of studies on the genotoxicity and cytotoxicity of Au NPs on mammalian cell lines (Alkilany and Murphy 2010; Chueh et al. 2014; Coradeghini et al. 2013). The majority of nanotoxicity research has focussed on cell culture. We evaluated the overall toxicological impact of Au NPs in organisms of diverse trophic levels. They included bacteria (B. aquimaris), an alga (Chlorella sp.), a cervical cancer cell line (SiHa), and Swiss albino mice. Since an appropriate coating of nanoparticles ensures stability as well as partially prevents their harmful effects, we have coated the Au NPs with citrate and PVP for the possibility of producing safer Au NPs. Before studying the toxic potential of Au NPs, we characterized their size by dynamic light scattering as well as TEM.

Physicochemical characterization of Au NPs showed a typical TEM image of round-shaped Au NPs. The size of the CIT30–PVP capped Au NPs was slightly reduced from the size of CIT30 Au NPs that might be due to the addition of PVP solution. Pillai and Kamat (2004) cited that the covalent linkage or surface capping of nanoparticles with an organic molecule usually yields smaller sized NPs. The size of the citrate-capped Au NPs (CIT30 and CIT40 Au NPs) obtained from the TEM was found to be in agreement with the DLS results, while in CIT30–PVP capped Au NPs, a large variation between the DLS and TEM was noted. Nur (2013) illustrated that the difference between DLS and TEM size is always higher for the PVP-stabilized samples than the citrate-stabilized samples. This size difference was due to the PVP, a long-chain polymer whose size is larger than that of citrate, a much smaller molecule (Tejamaya et al. 2012).

DLS is a highly sensitive, analytical tool recommended for monitoring nanoparticle aggregation in solution (Diegoli et al. 2008; Jans et al. 2009; Powers et al. 2006). In the present study, DLS has been used as the primary size determination tool to analyze the aggregation rate of Au NPs in the respective experimental matrix such as sterile lake water and cell culture media. The effective diameter of CIT30 Au NPs in sterile lake water was increased at 0 h to more than 100 nm. This effect was not found in CIT40 and CIT30–PVP capped Au NPs, which could be due to the higher reactivity of lower sized nanoparticles (CIT30). According to Zeng et al. (2012), the surface energy of Au NPs increases, as the diameter decreases; as a result, smaller Au NPs interact more strongly with the compounds present in the solution and leads to size-dependent aggregation of Au NPs. In sterile lake water, CIT30 Au NPs agglomerated rapidly than CIT40 Au NPs. The DLS results depicted that PVP-capped Au NPs (CIT30–PVP capped) were more stable than the citrate-capped Au NPs (CIT30 and CIT40). Similar results were observed by Nur (2013) in environmentally relevant ionic strength solutions and bacterial growth media. The effect of natural organic matter (NOM, Suwannee River humic acid) on the aggregation rate of Au NPs stabilized with various capping agents was demonstrated by Stankus et al. (2011). They reported that due to the variations in ionic strength and pH, the rate of aggregation was high for citrate-capped Au NPs. In general, the citrate-capped Au NPs rely on the electrostatic interactions of citrate and are weakly bound to Au NPs (Tolaymat et al. 2010). But, polyvinyl pyrrolidone (PVP) relies on steric stabilization and is strongly bound to the core of NPs (Römer et al. 2011; Cumberland and Lead 2009). Thus, the charge-stabilized nanoparticles (citrate capped) were more unstable in the lake water than the sterically stabilized nanoparticles (PVP capped).

In contrast to the sterile lake water matrix, the hydrodynamic size of the Au NPs in the culture media was found to be unchanged for all the three types of Au NPs, during the experimental period of 24 h. The biological components of media such as proteins, etc. might have helped in the stabilization of Au NPs and thereby prevented the aggregation of nanoparticles due to the formation of protein corona or biomolecular coating on NPs (Cedervall et al. 2007).

In both bacteria and algae, the toxic effect of citrate-capped Au NPs was dependent on the particle size and concentration. These results are in agreement with the size-dependent toxicity reported by Coradeghini et al. (2013) and Pan et al. (2007) in several cell lines. Previous information on the toxicity of Au NPs in various organisms suggested that surface chemistry and charged surface functional groups of the Au NPs play a crucial role in determining their cellular toxic effects (Bozich et al. 2014; Kim et al. 2013). In our study, PVP-capped gold nanoparticles showed lesser toxicity than citrate-capped nanoparticles in the case of both bacteria and algae. Sathishkumar et al. (2014) mentioned that the protective coating of nanoparticles reduced the cytotoxicity and helped in the stabilization of nanoparticles. These findings suggest that the size and surface capping of the nanoparticles may play an important role in the cytotoxicity of Au NPs.

Mechanisms of toxicity implicated by the Au NPs in bacteria and algae were confirmed by the ROS and LDH assay. In bacteria, ROS generation and LDH release were dependent on the size and surface capping of Au NPs. Higher ROS generation and cell membrane damage were observed in CIT30 Au NPs compared to CIT40 and CIT30–PVP capped Au NPs. In contrast, a different mechanism might have been followed for Au NP toxicity on the algal cells. For algal cells, ROS generation seemed to be the major contributor in the toxicity of citrate-capped Au NPs (CIT30 and CIT40). In the case of CIT30–PVP capped Au NPs, enhanced membrane damage could be significantly imparting its toxicity.

In the bacteria model, the LDH release and ROS generation were well correlated for both citrate- and PVP-capped Au NPs. But in algae, enhanced LDH release by the CIT30–PVP capped Au NPs was found to be inconsistent with the decreased ROS generation. There can be numerous possible reasons for this discrepancy. Wang et al. (2011) reported that the oxidation of Au3+ ions and decarboxylation of citrate on the surface of Au NPs induced the generation of free radicals in the presence of light. Due to the unavailability of reactive citrate ions on the surface of PVP-capped Au NPs, CIT30–PVP capped Au NPs induced less ROS generation than the citrate-capped Au NPs (CIT30 and CIT40) in both bacteria and algae. The cell wall of algae is primarily made up of cellulose; while in the case of bacteria, the cell wall majorly consists of peptidoglycan (Cooper 2000; Remade 1990). Due to the variation in the cell wall structure, interaction between the CIT30–PVP capped Au NPs and the algal cell wall may be enhanced in comparison to that in bacteria. Renault et al. (2008) observed a specific surface interaction/adsorption of amine-coated Au NPs on the cell walls of Scenedesmus subcapitus. The effect of capping agents alone (citrate and PVP) on ROS generation and LDH release in algae was also studied. Following the LDH release and ROS generation data for CIT30–PVP capped Au NPs (Fig. 6), 1.25 mM PVP also showed enhanced membrane damage (0.49 ± 0.26 a.u.) than citrate (0.10 ± 0.05 and 0.09 ± 0.03 a.u. for 136 and 30.4 mM, respectively) with respect to the control. However, PVP induced less ROS generation (27.37 ± 5.13 a.u.) than the citrate (54.62 ± 3.27 a.u. for 136 mM and 33.06 a.u. for 30.4 mM) in comparison to the control.

Previous in vitro studies have shown that the presence of sodium citrate residues on the particle surface resulted in a decline in cell viability of epithelial and endothelial cells (Freese et al. 2012). The cytotoxicity induced by the capping agents and Au NPs was determined by MTT assay in vitro in the SiHa cell line. This cell model was specifically selected based on prior reports regarding extensive application of Au NPs for cancer cell imaging and therapeutics in SiHa cells (Gao et al. 2012; Kumar et al. 2011; Leonard et al. 2011). The concentrations selected were based on our in vivo studies and previous studies (Kumar et al. 2011; Leonard et al. 2011). In general, the survival rate decreased with increasing concentration of the Au NPs and the period of exposure. The suppressive effect of Au NPs was indirectly associated with their sizes in SiHa cells, and distinct genomic DNA bands without any DNA laddering support that there was no apoptotic cell death of the SiHa cells. There is only a handful of previous reports with regard to apoptosis signaling and onset after cell exposure to Au NPs.

In the present study, genotoxic potentials of the Au NPs (CIT30, CIT40, and CIT30–PVP capped) were investigated using the alkaline comet assay in vivo on mice hepatocytes. The liver was the preferred organ of study as it is the target site where Au NPs are generally deposited (Chen et al. 2013; Khan et al. 2013; Simpson et al. 2013). Selection of dose for in vivo nanotoxicity experiments is interlaced with many challenges that include stability and dispersion of the nanoparticle used (Dhawan and Sharma 2010). There are a number of previous in vivo toxicity studies involving low dose levels of Au NPs (Lasagna-Reeves et al. 2010; Cho et al. 2009; Zhang et al. 2009). Alkaline comet assay experiments detect many types of DNA damage, i.e., strand breaks, alkali-labile sites, and incomplete excision repair sites (Kumaravel and Jha 2006). Au NPs possibly induced some or all of these types of DNA damage. The induced DNA damage was directly dependent on the size of the Au NPs and their concentration in solution. Irrespective of the capping agent, CIT30 and CIT30–PVP capped Au NPs induced significant DNA damage. Citrate at higher concentration (136 mM) induced significant DNA damage. However, the intratracheal instillation of uncoated gold nanomaterials of 2-, 20-, and 200-nm sizes investigated for DNA damage revealed a weak size-related increase of the mean tail intensity (Schulz et al. 2012). Hwang et al. (2012) reported significant differences in toxicity of Au NPs in the livers of mice, depending on the diet given to the experimental animals (methionine- and choline-deficient diet, viz. normal chow diet). Therefore, specific effects of size and surface functionalization of nanoparticles cannot be simply distinguished, and further in-depth studies were required to obtain more clarity on this issue.

Au NPs have been shown to induce a variety of toxic effects, including generation of reactive oxygen species and oxidative stress (Li et al. 2010; Tedesco et al. 2010; Gao et al. 2011). In the current study, Au NPs exerted oxidative stress causing substantial damage to different cellular components, DNA breaks, and lipid membrane peroxidation. Further studies on Au NPs revealed that Au NPs induced the formation of reactive oxygen species (ROS) and increased the level of lipid peroxidation. The activities of glutathione peroxidase (GSH), catalase (CAT), and the content of glutathione (GST) did not follow a dose-dependent trend for the various concentrations and sizes of the Au NPs used. Several recent reports have shown ROS generation in the cells exposed to Au NPs. Li et al. (2010) showed an increased oxidative stress and lipid peroxidation in MRC-5 (human lung fibroblasts) cells exposed to Au NPs. Tedesco et al. (2010) also reported an increase in oxidative stress upon in vivo exposure of Au NPs. Gao et al. (2011) demonstrated GSH depletion induced hydrogen peroxide generation in HL7702 (human liver cell line) cells upon exposure to Au NPs.

Summary and conclusions

This study aimed to address the toxicity of gold nanoparticles relevant to their size and surface functionalization on diverse organisms. In general, the physicochemical properties of Au NPs (i.e., size and surface modifications) influenced the interaction with the different trophic levels studied in vitro (bacteria, algae, SiHa cell line) and in vivo (mice). Beyond the variability of experimental conditions related to the organisms selected, in general, as-synthesized Au NPs were significantly toxic both in vitro and in vivo. CIT30 Au NPs were most toxic to bacterial, algal, and SiHa cells. The addition of PVP over CIT30 Au NPs reduced the toxic effects across trophic levels. However, in the mice model, genotoxicity was found to be increased upon exposure to CIT30–PVP capped Au NPs as compared to CIT30 Au NPs. The capping agent, PVP, which is a long-chain polymer was individually inert in the majority of the biological targets tested. In summary, results from this study provide a comprehensive understanding of size- and surface functionalization-dependent toxic effects of Au NPs across the trophic levels.