Introduction

The worldwide nanoparticles (NPs) utilization has been expected to reach 584,984 metric tons in 2019 from 225,060 metric tons (2014) with a 21.1% annual growth rate (McWilliams 2014). Among the various NPs utilized, titanium dioxide nanoparticles (TiO2 NPs) were the highly used metal oxide NPs which received enormous importance due to their specific properties like higher photocatalytic action, as a semiconductor, and super hydrophilicity nature under UV light (Bottero and Wiesner 2010; Fan et al. 2014; Huang et al. 2010). TiO2 NPs have been applied in various industrial sectors such as in agriculture for accelerating plant growth (Zahra et al. 2017), cosmetics and paints as a UV (ultraviolet) absorber and pigments (Mueller and Nowack 2008), the food sector as an additive and in packaging’s (Yang et al. 2014), the semiconductor industry (Bai et al. 2014), medicinal sectors (Ou et al. 2016), and a few other applications like in environmental remediation (Wang et al. 2016), construction and sports materials (Harifi and Montazer 2017), and sensing (Jiang and Zhang 2009), etc. They were also commonly utilized in consumer commodities from sunscreens to electronic products (Vance et al. 2015).

As a consequence of the enormous usage, TiO2 NPs gain access into the aquatic environment mainly through the consumer products, run-off, and wastes generated at both consumer and manufacturer levels (Gondikas et al. 2014; Gottschalk et al. 2015). Botta et al. (2011) observed a notable discharge of sub-micron-sized TiO2 NPs from sunscreens upon aging, which constitutes of about 30% of NPs initially present in the sunscreen. From the literature reports, it has been apparent that the excessive usage of TiO2 NPs, especially in the consumer products, was the foremost reason behind their remarkable exposure into the environment (Keller et al. 2013; Windler et al. 2012). Recently, Shandilya et al. (2015) inspected the impact of weathering (air and water) on TiO2 NPs coatings used in building materials. A slow release of TiO2 NPs into the air and water has been observed via abrasions and leaching which occurred in the coatings during the course of the period (7 months) analyzed in the study. As a result, the released NPs might interact with the organisms present in the aquatic system and may induce considerable toxic effects. Hence, it is indispensable to evaluate the impact of TiO2 NPs on various aquatic organisms.

Daphnia species are microscopic crustaceans which are usually present in the freshwater environment. They are non-selective filter feeders, which feeds on phytoplankton and tiny particles below the size range of 50 μm (Hund-Rinke and Simon 2006; Lampert 1987). Due to its ease of availability, handling, and high sensitivity to the pollutants in the ecosystem (Tatarazako and Oda 2007), Daphnia sp. has been widely used as a model organism for the ecotoxicological studies and risk assessment of nanoparticles. Das et al. (2013) assessed the effect of surface capping on the toxicity of TiO2 NPs on Daphnia magna. Uncapped TiO2 NPs induced higher mortality than the carboxy-functionalized TiO2 NPs. Several studies on TiO2 NPs have correlated the toxicity of TiO2 NPs with the ROS generation and indicated the importance of oxidative stress on the NPs toxicity (Clemente et al. 2014; Kim et al. 2010). Apart from the TiO2 NPs, food available in the test matrix acts as another stressor and impacts the toxicity of NPs on daphnia. Allen et al. (2010) noticed a decrement in the silver NPs toxicity in the presence of algae. It has been suggested that the nutrition availability alters the survival efficiency of daphnids by providing the sufficient energy required for their growth and reproduction (Conine and Frost 2017). Though the TiO2 NPs toxicity studies on aquatic organisms were carried out on different modes (Dalai et al. 2014; Wang et al. 2017), the impact of food on NPs toxicity remains ambiguous. Henceforth, the effect of fresh algae on NPs toxicity on daphnia has to be considered as it implies the environmental scenario.

As TiO2 NPs possess distinct crystalline forms (rutile, anatase, and brookite), the toxicity profile of TiO2 NPs concerning their crystalline forms has to be explored. Even though the crystallinity-based toxicity studies were available for TiO2 NPs especially the anatase and P25 forms (Clemente et al. 2014; Jacobasch et al. 2014), their detailed mechanism and the other crystalline form, rutile which also has a remarkable economic application (Yu et al. 2013), were not well explored in detail. It has been well known that the variation in irradiation has a significant impact on the toxic effect of TiO2 NPs owing to its photocatalytic differences upon irradiation (Li et al. 2016; Lu et al. 2017b). Other than these factors, aquatic components such as natural colloids and suspended matter also come into action in the aquatic system, which cannot be predicted with the present laboratory studies as they were tested mostly in the supplemented media (Hegde et al. 2016). In a study by Wormington et al. (2017), the presence of natural organic matter (NOM) inhibited the toxic effect of TiO2 NPs to a greater extent owing to their decreased ROS (reactive oxygen species) generation but not by UV attenuation. Henceforth, it is imperative to evaluate the toxic profile of TiO2 NPs in a similar environmental matrix with importance to their crystallinity and irradiation condition.

Owing to the wider release of distinct NPs into the aquatic environment, NPs interactions with other nanoparticles were inevitable. As a consequence, it may induce significant alterations in their toxic effects. Investigations on the toxic effects of nanoparticles as a mixture have been emerging gradually in the recent few years (Kathawala et al. 2015; Ko et al. 2017; Zou et al. 2014). Only a few studies have been evaluated on aquatic organisms regarding the toxicity of nanoparticles as a mixture (Costa 2015; Ko et al. 2018; Ye et al. 2017). Recently, Lu et al. (2017a) analyzed the mixture effects of copper (Cu) and chromium (Cr) NPs on D. magna in artificial freshwater and observed a distinct toxic effect when compared with their individual NPs. Co-exposure of Cu and Cr NPs resulted in similar toxicity as of Cu NPs and less toxic in comparison with Cr NPs alone. In another study, a binary mixture of zinc oxide (ZnO) and graphene oxide NPs exhibited discrete effects based on the type of organisms tested. Mixture exhibited an antagonistic effect on Danio rerio, and additive effect on both Scenedesmus obliquus and D. magna (Ye et al. 2018). Upon realizing the current scenario on nanoparticles buildup, it is mandatory to scrutinize the impact of NPs mixture in the freshwater system.

The current study aims to elucidate the toxic effects of TiO2 NPs (rutile, anatase, and their mixture) on Ceriodaphnia dubia in the presence of algae under a similar freshwater environment, with an importance to their crystallinity. Here, the algal diet was provided to daphnids as the algal species were also available in the freshwater environment along with the NPs, which may impact the NPs toxicity. In addition, the sterile freshwater has been employed as a test medium to mimic the chemical matrix of the freshwater environment. The biochemical markers such as malondialdehyde (MDA), catalase, and reduced glutathione (GSH) were also analyzed to elucidate the oxidative stress induced by the TiO2 NPs on C. dubia. These biomarkers were selected based on the mechanism of antioxidant systems and the type of end product formed in response to ROS induced by the NPs. Moreover, the influence of irradiation on TiO2 NPs toxicity was also determined by analyzing the effect of TiO2 NPs under two different irradiation conditions such as visible and UV-A—the two common light sources which were available to the daphnids in the natural freshwater system.

Materials and methods

Chemicals

BG-11 broth; trichloroacetic acid (TCA); 2-thiobarbituric acid (TBA); TRIS HCl; 5,5-dithiobis-[2-nitrobenzoic acid] (DTNB); and malondialdehyde (MDA) were purchased from Hi-media Laboratories Pvt. Ltd., Mumbai, India. Reduced glutathione (GSH) was procured from the manufacturer, Sigma-Aldrich, MO, USA. Hydrogen peroxide (30% w/v H2O2) was obtained from Sd fine-chem limited, Mumbai, India. All the other reagents utilized in the study were of analytical grade.

Nanoparticles and their preparation

Distinct TiO2 NPs varying in their crystalline phases such as anatase (CAS No: 1317-70-0) and rutile (CAS No: 1317-80-2) NPs were obtained from the manufacturer, Sigma-Aldrich, USA. The particle sizes of the anatase and rutile NPs were advertized as < 25 nm and < 100 nm (∼ 10 nm diam. × 40 nm L), respectively. NPs suspension with the stock concentration of about 5 mM was prepared in the de-ionized water and utilized for the toxicity assays. Further, they were sonicated at 20 kHz frequency for about 20 min with the help of a 130-W ultrasonicator (Sonics, USA) to achieve a homogenous suspension of TiO2 nanoparticles. Both the TiO2 NPs were characterized by various techniques (electron microscopy, dynamic light scattering (DLS)) and reported in our previous studies (Iswarya et al. 2015). NPs suspension characterized by transmission electron microscopy (TEM, Fig. 1) disclosed that the primary sizes of the TiO2 NPs were noted to be 32.63 ± 2.73 nm in length and 5.97 ± 0.49 nm in breadth for rutile NPs, and 11 ± 0.54 nm for anatase NPs. It was also remarked that the rutile NPs were rod-shaped, and the anatase NPs were spherical. Henceforth, the size analysis through the TEM and DLS disclosed that the sizes of rutile and anatase NPs were almost similar. Thus, the abovementioned NPs were further utilized for the assessment of the crystalline effect of TiO2 NPs.

Fig. 1
figure 1

TEM images of the a rutile NPs, and b anatase NPs

Full size image

Experimental matrix and test species

Sterile freshwater prepared from the freshwater obtained from the VIT Lake was utilized as an experimental matrix for the entire experiment carried out in the present study. The collected freshwater was initially filtered with the series of filters such as blotting paper and Whatman no. 1 and followed by its sterilization to be devoid of debris and microbes present in it. The sterilization process was attained by autoclaving the filtered freshwater at 121 °C, 15 psi for about 15 min such that the chemical composition of the freshwater remains intact. This filter-sterilized freshwater was further expressed as sterile freshwater throughout the study. Thus, the freshwater system has been mimicked by the utilization of sterile freshwater as an experimental matrix in this study. The physicochemical parameters of the sterile freshwater have been characterized and detailed in our previous study (Iswarya et al. 2018).

Freshwater daphnia, C. dubia, was used as a test species for evaluating the toxicity assessment of TiO2 NPs. Chlorella sp., a freshwater alga, has been provided as a feed for C. dubia. Both the organisms were isolated from the VIT Lake located in the Vellore Institute of Technology, Vellore, India, and maintained in our laboratory under specified conditions (visible light, day/night rhythm of 16 h/8 h, 24 °C). Experimental procedures regarding the isolation of freshwater organisms were reported briefly in our previous studies (Iswarya et al. 2015; Iswarya et al. 2016a). White fluorescent lights (Philips Life Max, 18 W with an intensity of 1.8 W/m2) have been utilized to provide the visible light for the test organisms to maintain the day and night rhythm. Sterile freshwater was used as a sub-culture media for the maintenance of C. dubia, while the BG-11 medium was used for the subsistence of Chlorella sp.

Experimental design

Algal cells devoid of sub-culture media were harvested from the 28-day old stock cultures of Chlorella sp. by means of centrifugation at 7000 rpm at 4 °C for about 10 min. From the algal cells harvested, an optical density (OD) of 0.1 OD which consists 1 × 105 cells was prepared in the sterile freshwater. To the exposure medium, i.e., the sterile freshwater comprising 0.1 OD of algal cells, TiO2 NPs (rutile, anatase, and mixture) and 24-h-old neonates were added. The number of daphnids added to the exposure medium was varied based on the assay performed, viz., ten for mortality and 30 for oxidative stress assessments. As per the OECD guideline of 2 mL of the exposure medium per daphnid (OECD 2004), about 20 and 60 mL of the exposure medium was maintained for mortality and oxidative stress experiments, accordingly. Various concentrations of TiO2 NPs in an array of 75–1200 μM were selected for the toxicity assays. Whereas the rutile and anatase NPs were taken in the equivalent ratio for the binary mixture, i.e., the binary mixture comprises an equal concentration of rutile and anatase NPs. For instance, 300 μM of binary mixture consists of 150 μM of rutile NPs and 150 μM of anatase NPs such that the total concentration of the mixture is 300 μM. After the addition of neonates, they were kept under visible and UV-A irradiations for about 48 h under photoperiodic conditions, i.e., 16 h light and 8 h dark. A similar experiment has been followed for the negative control, i.e., in the absence of TiO2 NPs, which has been further portrayed as untreated daphnids (0 μM). Moreover, separate control groups were maintained for visible and UV-A irradiation. White fluorescent tubes (1.8 W/m2 intensity, Philips) and black tubes (2.3 W/m2 intensity, Philips) were utilized to afford visible and UV-A light for the daphnids, respectively. After the 48-h exposure to TiO2 NPs, the mortality assessment and the oxidative stress assays were performed.

Mortality assessment

Various concentrations of TiO2 NPs such as 0, 75, 150, 300, 600, and 1200 μM were selected for the mortality studies on C. dubia. About ten neonates were utilized to assess the mortality of C. dubia upon exposure to TiO2 NPs. After the 48-h exposure to TiO2 NPs, the number of neonates alive was counted and the data were expressed in terms of mortality percentage after its normalization with the untreated daphnids. From the mortality data, a mathematical model described by Abbott (1925) and Chesworth et al. (2004) has been executed to determine the type of interaction induced by the rutile and anatase NPs when they were co-exposed together. Expected mortality has been computed from the percentage mortality obtained for the individual NPs such as rutile (R) and anatase (A) NPs with the help of Eq. 1:

$$ \mathrm{Expected}\ \mathrm{mortality}=R+A-\left(R\times A/100\right) $$
(1)

where, R and A indicates the percentage mortality observed for rutile and anatase NPs individually, respectively. Further, the ratio of inhibition was computed using Eq. 2 by comparing the observed mortality with their expected mortality:

$$ \mathrm{Ratio}\ \mathrm{of}\ \mathrm{Inhibition}\ \left({R}_{\mathrm{I}}\right)=\mathrm{Observed}\ \mathrm{mortality}/\mathrm{Expected}\ \mathrm{mortality} $$
(2)

Based on the RI values obtained, the effect of the mixture has been categorized into synergistic (RI > 1), additive (RI = 1), and antagonistic (RI < 1). The obtained effects were considered as an additive if the RI values were not statistically varied from 1. In addition, the differences noted among the observed and expected mortality were statistically validated with the help of two-way ANOVA (Tukey multiple comparison tests) at the p value less than 0.05.

Aggregation profile of TiO2 NPs in the sterile freshwater

The aggregation profile of TiO2 NPs such as rutile, anatase, and binary mixture has been assessed in the sterile freshwater to determine the colloidal stability of TiO2 nanoparticles in the freshwater system. The aggregation profile of NPs was analyzed only at the specific concentrations of TiO2 NPs such as 75, 300, and 1200 μM under both visible and UV-A irradiations. The effective diameter of TiO2 NPs (rutile, anatase, and mixture) was determined using a Nanobrook particle size analyzer (90 Plus PALS, Brookhaven Instrument, USA) at distinct time intervals like 0 and 2 h. The abovementioned analysis was terminated after 2 h due to the micron size of NPs which cannot be measured with the particle size analyzer whose sensitivity was limited to 10 μm.

Additionally, the surface interactions between rutile and anatase NPs when they coexist as a binary mixture have been analyzed using transmission electron microscopy (TEM). Higher concentration of the binary mixture (1200 μM) was prepared in the sterile freshwater and subjected to irradiation under both visible and UV-A conditions for about 48 h. An aliquot of the suspension was placed on the copper grids and observed under high-resolution TEM (FEI Technai G2 T20 S-Twin, USA).

Oxidative stress markers

The oxidative stress induced by the TiO2 NPs has been further assessed on the daphnids treated with TiO2 NPs at the selected concentrations (0, 75, 300, and 1200 μM) with the help of certain biomarkers such as MDA, catalase, and GSH, etc. For oxidative stress analysis, about 30 daphnids (< 24 h) were exposed to TiO2 NPs for about 48 h under distinct irradiations—visible and UV-A. Control daphnids were also kept in parallel without any exposure to NPs under a similar experimental condition for both visible and UV-A irradiations. After 48-h exposure, the live daphnids (NPs-treated and untreated) were collected from the experimental medium and utilized for further assessments. Moreover, about 30 daphnids were maintained per sample from the replicates kept for the study such that the mass of the daphnids was quite enough to perform the assays. In case of insufficient daphnids, daphnids were collected from other replicates and maintained as individual samples such that their biomass was kept unique. The collected daphnids were washed in de-ionized water to eliminate the loosely bound TiO2 NPs if any. Then, the daphnids were homogenized in 1 mL of potassium phosphate buffer (0.1 M with pH 7.4) by means of an ultrasonicator (130 W, Sonics, USA) for 2 min and centrifuged at 10,000 rpm for 15 min at 4 °C. The whole homogenization process was performed on ice to avoid the generation of artifacts (ROS) during the course of ultrasonication. The pellet obtained after the process of centrifugation was utilized for MDA assay, while the supernatant was utilized for other assays. After the biomarker analysis, the obtained values were computed for individual daphnid by normalizing the biomarker data with the total number of daphnids utilized for the assay.

MDA assay

A renowned lipid peroxidation biomarker, malondialdehyde (MDA), was estimated to determine the oxidative stress induced by the nanoparticles on the daphnids. MDA assay has been performed as per the protocol described by Buege and Aust (1978). To the pellet obtained, 1 mL of 1 M TRIS HCl with the pH of 7.4 was added, well mixed and incubated at 37 °C for 30 min. Then, 2 mL of TCA-TBA reagent (0.375% TBA in 5%TCA) was added and vortexed for a while. The reaction mixture was then kept in a hot water bath at 90 °C for about 45 min. After a few minutes of cooling, the solution obtained was centrifuged at 4 °C for about 10 min at the speed of 3000 rpm. The absorbance of the supernatant obtained after the centrifugation step was measured at 535 nm using a UV-visible spectrophotometer (U2910, Hitachi, Japan). The amount of MDA produced in the sample was computed with the MDA standard curve and expressed in the unit of picomolar/daphnid after its normalization with the total number of daphnids utilized for the assay.

Catalase assay

Catalase assay has been performed according to the standard procedure prescribed by Aebi (1974). The reaction mixture comprising of 30 μL of the supernatant, 670 μL of potassium phosphate buffer (50 mM, pH 6.6), and 330 μl of 30 mM H2O2 prepared in potassium phosphate buffer was measured at 240 nm for about 3 min with the help of the UV-visible spectrophotometer. The rate of decrease in the absorbance was computed, and their enzyme activity was calculated in terms of milliunit/milliliter using its molar extinction coefficient, 0.0436 mM−1 cm−1. Then, the results were expressed in the unit of milliunit/milliliter/daphnid after its normalization with the total number of daphnids utilized for the assessment.

GSH assay

Reduced glutathione (GSH) assay has been performed as per the protocol described by Sedlak and Lindsay (1968) with slight modifications. Initially, the solution comprising 62.5 μL supernatant and 400 μL distilled water was precipitated with 5% TCA (12.5 μL). Then, the precipitated solution was centrifuged at 4 °C for 10 min at 6000 rpm. The supernatant obtained after the centrifugation step was utilized for GSH analysis. The mixture containing 50 μL of the supernatant, 200 μL of potassium phosphate buffer (0.2 M, pH 8.0), and 10 μL of DTNB (0.6 mM in 0.2 M potassium phosphate buffer, pH 8.0) were added to the 96-well plate, and their absorbance was read at 412 nm with the help of a microplate spectrophotometer (Powerwave XS2, Biotek). Further, the total amount of GSH produced by the daphnids was computed with the help of the curve obtained from the standard, reduced glutathione, and the results were expressed in the unit of picomolar/daphnid after its normalization with the total number of daphnids.

Statistical analysis

All the analyses were carried out in triplicates at least in three groups from the same batch of daphnids, and the results were represented in mean ± standard error. Analysis of variance (ANOVA) was executed with the help of the statistical software, GraphPad Prism, version 6.01. The statistical difference between the NPs-treated and NPs-untreated daphnids was evaluated with the help of two-way ANOVA (Tukey multiple comparison tests, p < 0.05). Furthermore, the variations among the different types of TiO2 NPs such as anatase, rutile, and mixture were also tested with two-way ANOVA at the level of significance, p < 0.05. In the same way, the significant variations between the visible and UV-A irradiations were also analyzed.

Results

Mortality assessment

Mortality induced by the distinct types of TiO2 NPs such as rutile, anatase, and mixture has been represented in Fig. 2. Under both visible and UV-A irradiations, a concentration-dependent incline in the mortality was noticed for the TiO2 NPs irrespective of its types such as rutile, anatase, and mixture. The daphnids exposed to 1200 μM of rutile, anatase, and mixture exhibited a mortality percentage of about 26 ± 2.45, 24 ± 2.45, and 30 ± 0% under visible irradiation. Mortality produced by the TiO2 NPs under visible irradiation was statistically different (p < 0.05) from their untreated daphnids at all the test concentrations employed in the study, except at 75 and 150 μM of individual NPs, while the daphnids exhibited 46 ± 2.45, 30 ± 4.47, and 28 ± 2% mortality under UV-A irradiation upon exposure to 1200 μM of rutile, anatase, and mixture, respectively. Similar to the visible irradiation, the mortality observed under UV-A irradiation was significantly different (p < 0.05) from the untreated daphnids at all the concentrations of TiO2 NPs analyzed in the study, except at 75 μM of all the types of TiO2 NPs and 150–300 μM of rutile NPs.

Fig. 2
figure 2

Mortality induced by the various types of TiO2 NPs (rutile, anatase, and mixture) under different irradiation conditions such as a visible and b UV-A irradiation on C. dubia. The asterisk (*) symbol represents that the mortality (%) observed for the TiO2 NPs was statistically significant (p < 0.05) from the untreated daphnids. Similarly, the alpha (α) and beta (β) symbol depicts that the mortality obtained for the rutile NPs was statistically different from the mortality observed for anatase NPs and mixture, respectively, while the “γ” symbol indicates that the mortality differences observed between the anatase NPs and the mixture were significant

Full size image

Individually, anatase and rutile NPs did not exhibit any significant difference in the mortality (p > 0.05) at all the concentrations tested under both the irradiation conditions, except 600 and 1200 μM at UV-A irradiation. Juxtaposing the individual NPs with the mixture, binary mixture showed the highest mortality than individual NPs under visible irradiation. The differences observed among the mortality of individual NPs and mixture were statistically significant (p < 0.05) at all the concentrations tested except at 1200 μM under visible irradiation. Like visible irradiation, binary mixture exhibited higher mortality than that of rutile NPs under UV-A irradiation at all the concentrations tested, excluding 600 and 1200 μM. The differences observed among the mortality of rutile NPs and mixture were statistically significant (p < 0.05) at all the test concentrations except at 75 μM under UV-A irradiation, while the anatase NPs did not show any significant difference (p > 0.05) in their mortality upon its comparison with the mixture under UV-A irradiation. Moreover, the differences observed between the mortality of visible and UV-A irradiations were statistically insignificant (p > 0.05) irrespective of the types of TiO2 NPs such as rutile, anatase, and mixture at all the concentrations tested, except at 600 and 1200 μM of rutile NPs.

Mathematical modeling of the mixture

A mathematical model has been utilized to characterize the mode of action that persists when the rutile and anatase NPs coexist as a mixture. RI values computed for the mixture has been represented in Table 1. Due to the discrepancy in the mortality values like zero mortality, RI values were not computed for certain concentrations of the mixture such as 75 and 150 μM under visible irradiation and 75 μM under UV-A irradiation. The highest RI value of about 2.29 ± 0.21 was observed at 300 μM of mixture, indicating its synergistic effect under visible irradiation. Whereas, the lowest RI value of about 0.61 ± 0.04 was noted at 1200 μM under UV-A irradiation representing the antagonistic action of mixture since its RI value was less than 1. On the contrary, an additive effect was noted at the concentration of 150–600 μM under UV-A irradiation and 600–1200 μM under visible irradiation with the RI value of 1. As the concentration of mixture increases, the RI value of the mixture was found to be decreased under both visible and UV-A irradiation. It signifies that the action of rutile and anatase NPs were decreased when they coexist as a mixture, particularly at their higher concentrations.

Table 1 Mathematical model inferring the mode of action of TiO2 NPs when rutile and anatase NPs coexist together as a mixture
Full size table

Aggregation profile of TiO2 NPs in sterile freshwater

The aggregation profile of 75, 300, and 1200 μM TiO2 NPs (rutile, anatase, and mixture) in the sterile freshwater was analyzed at various time intervals such as 0 and 2 h using DLS and has been represented in Fig. 3. TiO2 NPs irrespective of their types showed an increase in their effective diameter as the time progressed. Initially at 0 h, the effective diameter of 1200 μM TiO2 NPs in the sterile freshwater was noted to be 988.75 ± 43.38, 732 ± 17.98, and 754.94 ± 7.19 nm for rutile, anatase, and binary mixture, respectively. Their initial sizes were gradually increased to 6945.01 ± 455.06 nm (rutile NPs), 3510.29 ± 31.87 nm, (anatase NPs), and 4617.25 ± 139.43 nm (mixture) after 2 h irradiation with visible light. Similarly, UV-A irradiation also enhanced their initial sizes into 7231.09 ± 270.83, 3481.44 ± 227.22, and 5338 ± 254.01 nm for rutile, anatase, and mixture, respectively. The difference in the sizes of TiO2 NPs (all the forms) noted between 0 and 2 h was statistically significant (p < 0.05) under both visible and UV-A irradiations tested in the study. Contrasting the visible and UV-A irradiation, all the types of TiO2 NPs such as anatase, rutile, and mixture showed a significantly higher aggregation (p < 0.05) under UV-A irradiation with respect to their concentrations studied, except 300 μM of anatase NPs and 1200 μM of TiO2 NPs irrespective of their forms. Individually, anatase NPs were noted to be quite stable than the rutile NPs. These differences were significant at all the concentrations for both visible and UV-A irradiations (p < 0.05). Comparing the rutile NPs with the binary mixture, the size of the rutile NPs was found to be statistically high (p < 0.05) than the size of the mixture under both irradiations at all the concentrations employed. In contrast, anatase NPs showed lesser aggregation in the sterile freshwater than the binary mixture at all the concentrations and irradiations tested in the study. The differences in their sizes were statistically different (p < 0.05).

Fig. 3
figure 3

Aggregation profile of different forms of TiO2 NPs such as rutile, anatase, and mixture under visible and UV-A irradiation at various exposure concentrations such as a 75 μM, b 300 μM, and c 1200 μM. The symbol * depicts that the size differences noted between 0 and 2 h were statistically significant at p < 0.05, while the alpha (α) and beta (β) symbols signify that the sizes observed for rutile NPs were statistically different from the sizes observed for anatase NPs and mixture, respectively. Similarly, the “γ” symbol represents that the size variations observed between the anatase NPs and mixture were significant

Full size image

The surface interaction between rutile and anatase NPs when coexisting as a mixture in the sterile freshwater has been illustrated with the help of transmission electron microscopy. TEM images of the binary mixture after visible and UV-A irradiations were represented in Figs. 4 and 5, respectively. Similar surface interactions were noticed under both visible and UV-A irradiations. The binary mixture showed a complex and rapid aggregation of NPs in the sterile freshwater (Figs. 4a and 5a) as similar to DLS data. It was also noticed that the anatase NPs were encompassed by the rutile NPs (Figs. 4c and 5c) owing to their interparticle interactions. Energy dispersive X-ray (EDAX) analysis confirmed the availability of dissolved carbon and inorganic ions such as Ca2+, SO42+, Cl−, Na+, and Mg2+ along with the binary mixture under both visible (Fig. 4d) and UV-A (Fig. 5d) irradiations. These ions were commonly referred as the natural colloids of the freshwater that would have played a prominent role in the surface interaction between the anatase and rutile NPs when co-exposed together as a mixture.

Fig. 4
figure 4

TEM images of the binary mixture depicting the rutile-anatase interactions under visible irradiation after 48 h. a Notable aggregation of NPs upon interaction between the rutile and anatase NPs when they coexist as a mixture in the sterile freshwater. b Anatase NPs (spherical shaped) were covered by the rutile (rod-shaped) NPs. c High-resolution image of binary mixture depicting the clear confinement of anatase NPs by the rutile NPs. Yellow arrows—anatase NPs; red arrows—rutile NPs. d EDAX image of the binary mixture confirming the presence of inorganic ions (blue-colored ovals) and their role in the surface interactions between anatase and rutile NPs

Full size image
Fig. 5
figure 5

TEM images of the binary mixture depicting the rutile-anatase interactions under UV-A irradiation after 48 h. a Substantial aggregation of NPs upon interaction between the rutile and anatase NPs when they coexist as a mixture in the sterile freshwater. b Anatase NPs (spherical-shaped) were enclosed by the rutile (rod-shaped) NPs. c High-resolution image depicting the clear confinement of rutile NPs over the anatase NPs. d EDAX data of the binary mixture confirmed the existence of inorganic ions (blue-colored ovals) and their role in the surface interactions between anatase and rutile NPs. Yellow arrows—anatase NPs; red arrows—rutile NPs

Full size image

Oxidative stress markers

Several oxidative stress biomarkers such as MDA, catalase, and GSH were estimated to determine the changes induced on the daphnids upon interaction with the TiO2 nanoparticles.

MDA assay

MDA assay has been performed to describe the level of lipid peroxidation on the cellular membranes of daphnids. MDA produced by the daphnids upon exposure to visible and UV-A irradiations was represented in Figs. 6a and 7a, respectively. TiO2 NPs-treated daphnids showed the higher amount of MDA than that of untreated daphnids under both the irradiation conditions. Daphnids treated with 1200 μM of rutile, anatase, and mixture depicted MDA levels of about 3.89 ± 0.41, 3.08 ± 0.08, and 0.97 ± 0.24 pM/daphnid under visible irradiation, and 3.40 ± 0.08, 2.19 ± 0.16, and 4.21 ± 0.08 pM/daphnid under UV-A irradiation, respectively. Under visible irradiation, the amount of MDA produced by the daphnids was noted to be significantly increased (p < 0.05) till 300 μM, which experienced a decline at 1200 μM in case of individual NPs. In contrast, a decline in the MDA level was observed for the binary mixture with an incline in the concentration of mixture. The changes noted in the MDA of NPs-treated daphnids were statistically different (p < 0.05) from the MDA of untreated daphnids at all the concentrations, except at 300 and 1200 μM of the mixture. In contrast to visible irradiation, the MDA level of TiO2 NPs-treated daphnids was increased with respect to its concentration under UV-A irradiation. In case of anatase-treated daphnids under UV-A irradiation, the amount of MDA produced by the daphnids was lesser than the MDA of untreated daphnids. This effect was not observed for the concentration of 1200 μM.

Fig. 6
figure 6

Oxidative stress biomarkers assessed on the daphnids after its treatment with various types of TiO2 NPs such as rutile, anatase, and mixture under visible irradiation. a MDA, b catalase, and c GSH. The symbol “*” indicates that the levels of markers generated by the TiO2-treated daphnids were statistically significant with respect to their respective markers induced on the untreated daphnids. The symbol “α” depicts that the differences noted between the rutile and anatase NPs were statistically significant. Similarly, the symbol “β” represents the significant differences noted among the rutile NPs and mixture, and the “γ” symbol indicates the significant differences between the anatase NPs and mixture

Full size image
Fig. 7
figure 7

Oxidative stress biomarkers assessed on the daphnids treated with TiO2 NPs (rutile, anatase, and mixture) under UV-A irradiation. a MDA, b catalase, and c GSH. The asterisk symbol (*) represents that the markers produced by the TiO2-treated daphnids were significantly different from the markers of untreated daphnids. The alpha symbol (α) indicates that the variations observed between the rutile and anatase NPs were statistically different, while the significant differences noted between the rutile NPs and the mixture were represented by the “β” symbol and the differences between the anatase NPs and the mixture were indicated by the “γ” symbol

Full size image

Comparing the MDA levels of visible and UV-A irradiations, MDA levels were high under visible irradiation irrespective of its phases, except the mixture which produced MDA in higher amounts at the concentrations of 300 and 1200 μM under UV-A. Among the MDA levels of individual NPs, rutile-treated daphnids showed higher MDA production than anatase-treated daphnids under both visible (except 300 μM) and UV-A irradiations. Contradicting the individual NPs with the mixture, binary mixture induced higher amounts of MDA production under UV-A irradiation. In contrast, the amount of MDA induced by the mixture was lower under visible irradiation.

Catalase assay

Catalase activity of the daphnids treated with TiO2 NPs was evaluated and represented in Figs. 6b and 7b for visible and UV-A irradiations, respectively. Under both the irradiation conditions, catalase activity was noted to be significantly increased (p < 0.05) for NPs-treated daphnids as compared to their catalase activity of untreated daphnids. Upon visible irradiation, highest catalase activities of about 64.31 ± 1.31, 68.25 ± 0, and 65.62 ± 0 mU/mL/daphnid were observed for rutile (300 μM)-, anatase (75 μM)-, and mixture (300 μM)-treated daphnids, respectively. A concentration-dependent incline in the catalase activity was observed for rutile- and mixture-treated daphnids till 300 μM under visible irradiation. In contrast, the anatase-treated daphnids showed a significant decline in the catalase activity upon an incline in the concentration of NPs under visible irradiation. In the case of UV-A irradiation, catalase activity was higher at the concentration of 1200 μM which was noted to be 34.12 ± 0, 44.62 ± 0, and 56.44 ± 1.31 mU/mL/daphnid for rutile-, anatase-, and mixture-treated daphnids. As the concentration of NPs increased, the enzyme activity of catalase was also increased significantly (p < 0.05) under UV-A irradiation.

Within the individual TiO2 NPs, anatase NPs induced higher catalase activity on C. dubia than that of rutile NPs under UV-A irradiation. A similar effect has been noticed under visible irradiation only at the concentration of 75 μM. The differences noted in the catalase activity of individual NPs were significant at the p value < 0.05. Juxtaposing the enzyme activity of the mixture with the individual NPs, mixture-treated daphnids depicted higher catalase activity under UV-A irradiation and were significant (p < 0.05) at the concentrations of 75 and 1200 μM. In case of visible irradiation, the catalase action did not vary significantly (p > 0.05) with respect to their individual NPs. Only at the lower concentration (75 μM), the catalase action of mixture-treated daphnids was noted to be significantly (p < 0.05) high when compared with the rutile NPs-treated daphnids and lower with the anatase NPs-treated daphnids. Among the two irradiations tested in the study, catalase activity was significantly high (p < 0.05) under visible irradiation at all the concentrations employed, except 1200 μM.

GSH assay

Reduced glutathione was considered as a vital biomarker which has been commonly utilized to determine the defense capacity of the organism against the oxidative stress. The level of reduced glutathione on NPs-treated daphnids was analyzed, and the data were represented in Fig. 6c for visible irradiation and Fig. 7c for UV-A irradiation. An incline in the GSH levels was noted for NPs-treated daphnids than that of untreated daphnids under both visible and UV-A irradiations. Though an incline in the GSH levels was noted, a concentration-dependent effect has not been observed for NPs-treated daphnids under both visible and UV-A irradiations. The level of GSH was noted to be high at 300 μM under both visible and UV irradiations employed. It has been noted to be about 142.42 ± 0, 172.73 ± 0, and 187.88 ± 15.15 pM/daphnid under visible irradiation and 66.67 ± 15.15, 81.81 ± 0, and 112.12 ± 0 pM/daphnid under UV-A irradiation for rutile-, anatase-, and mixture-treated daphnids, respectively.

Among the individual NPs, GSH levels were higher for anatase-treated daphnids upon visible irradiation. These differences were significant (p < 0.05) at all concentrations, excluding 1200 μM. In contrast, UV-A irradiation did not induce any significant difference (p > 0.05) in the GSH levels upon treatment with the individual NPs. Comparing the individual NPs with the mixture, mixture-treated daphnids showed higher GSH levels than the individual NPs-treated daphnids under both the irradiation conditions, except at 75 μM of anatase NPs under visible irradiation. Contradicting the visible irradiation with UV-A, the reduced glutathione was significantly produced in higher quantity under visible irradiation irrespective of their types at all the test concentrations, except at 75 μM of the mixture.

Discussion

Toxicity of TiO2 NPs on C. dubia

In the present study, the toxic effect of TiO2 NPs (rutile, anatase, and mixture) on C. dubia was explored in the presence of algae under distinct irradiation conditions such as visible and UV-A irradiations in the milieu of a freshwater environment. As the concentration of TiO2 NPs increased, toxicity was also noted to be increased (Fig. 2). Variation in the crystallinity and irradiation did not induce any effect on the mortality of TiO2 NPs, except at the higher concentration of rutile NPs under UV-A. These results were in contrary with the earlier reports (Lu et al. 2017b) which state that the differences in the crystallinity and irradiation impact the TiO2 NPs toxicity. Since they were mostly performed in the absence of algae, it has been apparent that the differences noted in the toxicity were chiefly due to the algal diet. Upon comparison with our previous toxicity data on TiO2 NPs, it has been revealed that the TiO2 NPs toxicity obtained in the presence of algae was lower than the toxicity obtained through both water-borne (Iswarya et al. 2016a) as well as a dietary exposure (Iswarya et al. 2018). A similar effect has been observed by Allen et al. (2010) and Conine and Frost (2017) while examining the toxic effects of silver nanoparticles on D. magna. They also discussed that algal availability mitigates the toxic effects of nanoparticles due to the alteration in the nutrition available to the daphnids. In other words, it can be presumed that the toxic effect of individual NPs was diminished when afforded with the nutrition-rich sources like algae. Besides, the EPS (exo-polymeric substances) secretion by the algal cells and the surface interaction between TiO2 NPs and algae might have enhanced the NPs aggregation and reduced the bioavailability of TiO2 NPs in the test matrix, as stated by Sendra et al. (2017) and Dalai et al. (2013). Consequently, the TiO2 NPs available for the ingestion by the daphnids becomes less and in turn, diminished the NPs toxicity.

From the mortality data, it was observed that the action of the binary mixture was dependent on both concentration as well as the type of irradiation exposed. The effect of the mixture was decreased, as the concentration of TiO2 NPs in the mixture was increased. Upon visible irradiation, the toxicity of individual NPs was increased when they coexist as a mixture. It indicates the synergistic action of the mixture, especially when the rutile and anatase NPs were co-exposed in lower quantities. From the synergistic action, the effect of the mixture was slowly transformed into additive action upon an incline in the mixture concentration. In the case of UV-A irradiation, the binary mixture exhibited almost similar toxicity when compared with anatase NPs and higher toxicity with rutile NPs except at their higher concentrations. It implies the additive effect of the mixture under UV-A irradiation except at 1200 μM which depicted an antagonistic effect. These effects have been confirmed with the RI values computed for the mixture. Thus, the daphnids were highly sensitive to the binary mixture than their individual NPs under both the irradiation conditions. In our previous study, anatase-rutile mixture induced an antagonistic effect on C. dubia upon dietary exposure under both irradiation conditions (Iswarya et al. 2018). Similarly, the daphnids exposed to the mixture in the absence of algae (i.e., water-borne) experienced an antagonistic and additive effect upon visible and UV-A irradiation, respectively (Iswarya et al. 2016a). The differences noted in the mortality of daphnids were due to the differential feeding behavior of daphnids which prefers algal cells than the nanoparticles (Dalai et al. 2014). Another rationale behind the changes observed in the mixture toxicity was due to the surface interactions between rutile and anatase NPs in a mixture, which has been evident in the form of NPs aggregation (DLS and TEM data). As a result, the NPs agglomerated algal cells were also available for the ingestion, which in turn induces a significant change in the toxic effect of TiO2 NPs.

Aggregation profile of TiO2 NPs in the sterile freshwater

The changes occurring in the size of TiO2 NPs in the test matrix have been determined over a period of 2 h. TiO2 NPs either individual or mixture were not stable in the sterile freshwater at the concentrations tested in the study. They have attained micron size within 2 h of irradiation time (Fig. 3). The aggregation profile of different forms of TiO2 NPs follows a sequential order of rutile > mixture > anatase under both the irradiation conditions analyzed. This result implies that the rutile NPs were highly susceptible to aggregation in the sterile freshwater matrix studied under both visible and UV-A irradiations. TiO2 NPs undergoes aggregation owing to the interaction with the natural colloids such as dissolved carbon and inorganic ions present in the sterile freshwater (Sillanpää et al. 2011). Furthermore, UV-A exposure induced higher aggregation than the visible irradiation, irrespective of their forms. This effect has been prominent only at the lower concentrations tested. Sun et al. (2014) reported a similar observation, i.e., higher aggregation of P25 NPs upon UV-A irradiation in comparison with non-irradiated NPs suspension. They also elucidated that an incline in the hydroxyl radical’s generation, especially the bridging hydroxyls by TiO2 NPs, was the major factor behind their enhanced aggregation under UV-A irradiation.

Aggregation observed for the binary mixture was also influenced by a typical interfacial interaction between the anatase and rutile NPs in a mixture. It has been evident from the TEM images of the binary mixture that the anatase NPs were encompassed by the rutile NPs (Figs. 4c and 5c) which confirmed their surface interactions when they exist as a binary mixture. EDAX results conveyed the clear information regarding the role of dissolved inorganic ions and carbon in exacerbating the aggregation of individual NPs while they exist as a mixture. Dissolved inorganic ions such as Ca2+, SO42+, Cl−, Na+, and Mg2+, etc., and organic matter (carbon) present in the freshwater induces a surface complexation with the nanoparticles, which ultimately results in the NPs aggregation (Domingos et al. 2010).

Based on the information obtained from DLS (Fig. 3) and TEM images (Figs. 4 and 5), the probable interaction of TiO2 NPs (individual as well as a mixture) with the abiotic components of the freshwater has been hypothesized. It has been speculated that the inorganic ions of the freshwater form a thin layer of ions around the NPs to which the natural organic matter (NOM) interacts and binds with them. For individual NPs, this inorganic layer acts as a bridge and provokes the interaction between the NPs and NOM (Chowdhury et al. 2012; Domingos et al. 2010; Metreveli et al. 2016). While in case of the mixture, this inorganic layer and NOM acts as intermediates and induces an electrostatic interaction between anatase and rutile NPs. This NOM binding varies based on the irradiation exposed. It has been presumed that UV-A irradiation infringes the NOM due to the higher ROS generated by the TiO2 NPs upon photolysis. Thereby, these charged, smaller particles promote the aggregation of NPs in the sterile freshwater upon TiO2 NPs interaction. In turn, it may lead to sedimentation and affects the bioavailability of NPs to the aquatic organisms present in the system (Aiken et al. 2011; Nur et al. 2015). A schematic representation illustrating the interaction between the anatase and rutile NPs in a mixture has been depicted in Fig. 8.

Fig. 8
figure 8

Schematic diagram representing the transformations and surface interactions between the rutile and anatase NPs in the sterile freshwater when they coexist as a mixture

Full size image

Oxidative stress markers

Analysis of oxidative stress markers was of prime importance to elucidate the stress impregnated by the nanoparticles on daphnids. It also helps us to determine the defense mechanisms portrayed by the daphnids to cope up with the stress caused by the NPs. Kim et al. (2010) elucidated that the mortality induced by the TiO2 NPs was ascribed to the oxidative stress generated by the nanoparticles which have been indirectly linked with their antioxidant activities. Hence, some of the important oxidative stress markers such as MDA, catalase, and GSH were evaluated in the current study. MDA is a reactive substance which is highly produced upon the peroxidation of lipids as a consequence of reactive oxygen species (ROS) (Barata et al. 2005). A concentration-dependent incline in the MDA was observed for all the types of NPs tested under UV-A irradiation (Fig. 7a). Even though an increase in MDA was noticed, the MDA levels of anatase NPs were almost similar to the MDA levels of untreated daphnids. In contrast to UV-A, visible irradiation did not induce a concentration-dependent effect on the MDA levels, except the mixture which showed a decrement in the MDA production upon an increase in the concentration (Fig. 6a). Among the different types of TiO2 tested, rutile NPs depicted the higher amount of MDA production under visible irradiation except at 300 μM, whereas the mixture induced higher quantities of MDA under UV-A irradiation. It depicts the crystalline effect of TiO2 NPs on the induction of lipid peroxidation. Previous reports illustrated that the excess ROS produced by the nanoparticles interact with the organism’s polyunsaturated fatty acids and releases a variety of lipid peroxides including MDA, which eventually activates the antioxidant enzymes like catalase, GST, etc. for detoxification (Bagnyukova et al. 2006; Borgeraas and Hessen 2000; Gao et al. 2017). In addition to ROS, the direct interaction of nanoparticles with the organism also attributes the lipid peroxidation (Federici et al. 2007).

Catalase is an enzymatic antioxidant which catalyzes the H2O2 accumulated in the cell and helps in the detoxification mechanism (Halliwell and Gutteridge 2015). Upon exposure to TiO2 NPs, catalase activity of daphnids was increased significantly than the untreated daphnids (Figs. 6b and 7b). Similar to MDA, an increase observed in the catalase activity was noted to be concentration-dependent under UV-A irradiation and independent upon visible irradiation except the mixture, which showed a decrease in the catalase activity under visible irradiation. Among the different types of TiO2 tested, mixture depicted higher catalase activity under UV-A. Individually, anatase NPs showed higher catalase activity under UV-A irradiation. However, the catalase activity observed between the particles was not comparable under visible irradiation. The changes observed in the catalase activity were consistent with MDA data, except the anatase NPs under UV-A. Beyond these differences, catalase activity of daphnids was higher under visible irradiation when compared to UV-A irradiation. Reeves et al. (2008) reported that the by-products of H2O2 especially the hydroxyl radicals induced the cytotoxicity on fish cells upon exposure to UV-treated TiO2 NPs and damaged the major cellular components such as the DNA. Vlahogianni et al. (2007) mentioned that the catalase activity was enhanced to counteract the indirect damages caused by the H2O2 accumulation.

Reduced glutathione (GSH) is a non-enzymatic antioxidant molecule which functions as a cofactor for the enzyme glutathione-S-transferase (GST) and protects the integrity of both enzymes and proteins (Mwaanga et al. 2014). In the present study, the amount of reduced glutathione was highly increased in NPs-treated daphnids under both UV-A and visible irradiations (Figs. 6c and 7c). Also, glutathione levels were noted to be independent of concentration under both the irradiations for all the types of TiO2 NPs tested. A similar concentration-independent effect on GSH levels was observed in mice upon treatment with 30 nm-sized gold NPs (Iswarya et al. 2016b). Contradicting with individual NPs, mixture depicted higher GSH levels under both the irradiation conditions. Among the individual NPs, anatase NPs showed the GSH quantity in higher amounts than the rutile NPs under visible irradiation. In contrast, the GSH levels observed under UV-A irradiation did not show any significant difference among the individual NPs tested in the study. An increase noted in the GSH levels represents a measure taken up by the antioxidant mechanisms to quench the oxidative stress induced. Owing to the large nucleophilic sulfhydryl moieties, GSH was actively involved in the detoxification of H2O2 and various other free radicals generated (Knapen et al. 1999). GSH depletion observed at 1200 μM indirectly indicates that the GSH levels were highly utilized by the free radicals generated by the TiO2 NPs for its neutralization.

Conclusion

The present study explored the toxic effects of various TiO2 NPs towards C. dubia in the presence of algae under distinct irradiation conditions such as visible and UV-A, in context with the freshwater scenario. The outcomes of the study confirm that the availability of algae in the test matrix mitigated the toxic behavior of TiO2 NPs when exposed individually. The decline in toxicity could have been due to surface modulation of the NPs by interactions with the algal cells or the EPS released by the algae. The binary mixture of rutile and anatase NPs showed enhanced toxicity compared to the individual NPs in the presence of an algal diet, especially under visible irradiation. The surface interaction between the anatase and rutile NPs possibly influenced the toxicity of the mixture. The results from antioxidant enzyme assays supported the mortality data.