Genotoxicity of titanium dioxide nanoparticles and triggering of defense mechanisms in Allium cepa

Abstract Titanium dioxide nanoparticles (TiO2NPs) are widely used and may impact the environment. Thus, this study used a high concentration of TiO2NP (1000 mg/L) to verify the defense mechanisms triggered by a plant system – an indicator of toxicity. Furthermore, this study aimed at completely characterizing TiO2NP suspensions to elucidate their toxic behavior. TiO2NPs were taken up by meristematic cells of Allium cepa, leading to slight inhibition of seed germination and root growth. However, severe cellular and DNA damages were observed in a concentration-dependent manner (10, 100, and 1000 mg/L). For this reason, we used the highest tested concentration (1000 mg/L) to verify if the plant cells developed defense mechanisms against the TiO2NPs and evaluated other evidences of TiO2NP genotoxicity. Nucleolar alterations and plant defense responses (i.e., increased lytic vacuoles, oil bodies and NP phase change) were observed in meristematic cells exposed to TiO2NP at 1000 mg/L. In summary, TiO2NPs can damage the genetic material of plants; however, plants displayed defense mechanisms against the deleterious effects of these NPs. In addition, A. cepa was found to be a suitable test system to evaluate the cyto- and genotoxicity of NPs.


Introduction
Titanium dioxide nanoparticles (TiO 2 NPs) are among the most used manufactured nanomaterials, being produced in thousands of tons per year around the world (Robichaud et al., 2009). This NP occurs in three forms in nature: anatase, brookite, and rutile (Clement et al., 2013). The frequent use of products containing TiO 2 NPs has been associated with inappropriate disposal of domestic and industrial effluents, which may result in the release of this NP into the environment. Thus, there is an emerging concern about the environmental impacts caused by NPs (Monica and Cremonini, 2009), in particular TiO 2 NPs.
Considering that higher plants are suitable models for assessing environmental toxicants and plants are fundamental organisms of all ecosystems, phytotoxicity studies become relevant because: (1) plants are considered the basis of the food chain; (2) plants provide the oxygen neces-sary for life; and (3) plants are distributed in different environments (terrestrial and aquatic). Thus, damages to plants can generate an imbalance in the ecosystems, emphasizing the importance of phytotoxicity studies (Ma et al., 2010). Within this context, there is a concern about the toxicological impacts of NPs on living organisms, such as plants -target organisms of NPs through different environment compartments (air, water, and soil).
The toxic effects of TiO 2 NPs on plants have been reported in the literature (Ghosh et al., 2010;Klancnik et al., 2010;Pakrashi et al., 2014). However, these studies considered only the nanopowder form, without an adequate characterization of TiO 2 NPs suspensions, fundamental to understanding the internalization, bio-uptake, and the behavior of the TiO 2 NPs inside the cells.
Allium cepa is an efficient plant system that has some advantages over other higher plants, in particular regarding its convenient chromosomal features that allow the evaluation of genetic damages. Genotoxic effects of different contaminants have already been reported in the literature using A. cepa (Grisolia et al., 2005;Rodrigues et al., 2010;Dourado et al., 2017), including the identification of the mechanisms of action (Fiskesjo, 1985;Leme and Marin-Morales, 2009;Klancnik et al., 2010;Pakrashi et al., 2014).
Moreover, plant systems allow the performance of complementary analysis. Ag-NOR banding technique is a new method to detect a genotoxicity biomarker used in A. cepa that has shown to be an efficient tool for studying these effects caused by contaminants (Mazzeo and Marin-Morales, 2015). These authors reported changes in Ag-NOR banding that may suggest aneugenic effects or increased genomic activity. Furthermore, Zhao et al. (2016) observed a cellular defense mechanism and morphological damages (e.g., ruptures of the plasma membrane, the appearance of oil bodies, and changes in the vacuoles).
Thereby, this study aimed to evaluate the toxic potential of TiO 2 NPs in higher plants to provide knowledge about the cellular responses after TiO 2 NPs exposure. For this purpose, several biomarkers of toxicity in A. cepa test system (e.g., germination rate and root development, morphological and DNA damages, nucleolar alterations) were used. In addition, selected area electron diffraction (SAED) analysis was carried out to verify NP phase-change as a plant defense mechanism.

Material and Methods
Chemical characterization of the titanium dioxide nanoparticle suspensions The TiO 2 NP (powder) was purchased from Sigma-Aldrich ® (CAS nº 1317-70-0), with physical characteristics being a particle size 21 nm (TEM), ³99.5% trace metal basis, and 100% anatase. Morphological characteristics of the TiO 2 NP powder were also determined to verify the crystalline structure by X-ray diffraction, the specific surface area by Brunauer-Emmett-Teller Theory (BET), and the surface chemistry using the X-ray photoelectron spectroscopy (XPS).
The TiO 2 NP was evaluated regarding its toxic potential using the Allium test system at 10, 100, and 1000 mg/L. The choice of the highest concentration used in this study was defined based on other studies (Klancnik et al., 2010;Zhu et al., 2010). TiO 2 NP suspensions were prepared in ultrapure water and dispersed in an ultrasonic water bath (42.000 Hz, 160 W) for 30 min immediately before the start of the bioassays with A. cepa. The highest concentration (1000 mg/L) was also chosen to observe if the cells developed defense mechanisms, what would be an indication of toxicity, and to observe the internalization, bio-uptake, and behavior of the TiO 2 NPs inside the cells.
The TiO 2 NPs suspensions were also analyzed. The characterization of these suspensions was performed by the Zetasizer ® Nano Series ZS90 (Malvern Instruments, Worcestershire, UK) to determine the average particle size (dynamic light scattering -DLS), polydispersity index and zeta potential (Laser Doppler Velocimetry and electrophoresis -LDV). We considered an angle of 90º and a wavelength of 633 nm (Malvern, 2015).
A complementary analysis of the NPs structure was performed by the selected area electron diffraction (SAED) technique, as described below).

Test system and exposure condition
One hundred A. cepa seeds (2n = 16 chromosomes) of the same batch and variety (the "baia periform" onion) were submitted to germination with the TiO 2 NP suspensions (10, 100, and 1000 mg/L), ultrapure water as negative control (NC), 10 mg/L of methyl methane sulfonate (MMS; Sigma-Aldrich, CAS 66-27-3) as positive control (PC) for cyto-and genotoxicity testing, and zinc sulfate heptahydrate (Sigma-Aldrich ® , CAS 7446-20-0) at 6 mg/mL (PC for seed germination and root elongation toxicity test) in polystyrene Petri dishes (diameter 85 mm) covered with a nylon net (100 seeds/plate) . TiO 2 NP suspensions were replaced by fresh ones every 24 h to assure its bioavailability for the test system.
All experiments were carried out at 25°C in the dark. The seed germination and root elongation toxicity tests were performed using triplicate plates per treatment (100 seeds/plate), while the other assays were carried out using a single plate per treatment (Lin and Xing, 2007;.

Seed germination and root elongation toxicity test
The seed germination and root elongation toxicity tests were carried out according to the protocol described by Rank (2003). After 5 days of exposure, seed germination (number of seedlings) and root length were measured. Toxicity was expressed as the difference of seed germination and root elongation when compared to the NC. The results were statistically analyzed using a Shapiro-Wilk test followed by Student's t-test (p < 0.05).

Cyto-and genotoxicity assessments
Allium cepa roots of 2 cm in length (~5 days) were fixed in a mixture of ethanol and acetic acid (3:1-v/v, Merck). The fixed roots were stained with Schiff's reagent, as described by Feulgen and Rossenbeck (Mello and Vidal, 1978), and the slides were prepared using the meristematic region according to the protocol described by .
Cytotoxicity was assessed by recording the changes in the mitotic index (MI) of the meristematic cells. Genotoxicity was determined by scoring different types of chromosomal aberrations (CAs) and nuclear abnormalities (NAs). Micronucleated cells were also scored to determine the mutagenicity . Additionally, the mode of action of TiO 2 NP was defined based on the analysis of different types of CAs, which were grouped as clastogenic (chromosome bridges and breaks) or aneugenic (chromosomal losses, chromosomal delay) according to .
These parameters were evaluated under a light microscope (Olympus BX-40-magnification -400 x) and 10 slides per treatment were analyzed (500 cells/slide). The treatments were statistically compared using the non-parametric Kruskal-Wallis test followed by the Student-Newman-Keuls test (p < 0.05).

Complementary tests
Complementary tests to the genotoxicity test were performed to confirm the toxic effects of TiO 2 NP and to visualize its behavior inside the cells.
In order to perform these tests, only the highest concentration (1000 mg/L) was chosen, because this concentration had already been shown to be genotoxicologically toxic, and the current objective was to confirm the toxicity and observe the effects caused to the plant system by the TiO 2 NPs.

Silver-stained nucleoli and nucleolar organizer region (Ag-NOR)
Ag-NOR staining and nucleoli analysis were carried out using fixed roots of 1000 mg/L TiO 2 NP treatment and an NC group, according to the previous protocol described by Mazzeo and Marin-Morales (2015). Thus, this analysis was used to verify whether the nucleolar pattern was altered after TiO 2 NP exposure.
Images were obtained using a motorized Axio Imager Z2 epifluorescence microscope (Carl Zeiss, Jena, Germany), equipped with an automated scanning V Slide (Metasystems, Altlussheim, Germany). Five thousand cells of both treatments were analyzed. This analysis was performed only on A. cepa interphase cells. The number of nucleoli per cell was determined by the counting tool implemented in Anati-Quanti software (Aguiar et al., 2007), and their size was determined by Image J using a specific tool that measures the area of each nucleolus (Passoni et al., 2014). The results were statistically analyzed using a Shapiro-Wilk test followed by a Student's t-test (p < 0.05).

Morpho-anatomical analysis
Onion roots of 2 cm in length (~5 days) were fixed in Karnovsky solution (Karnovsky, 1965) for light microscopy (LM) and transmission electron microscopy (TEM) analyses. For both methodologies, visual analysis was performed.

Light microscopy (LM)
Fixed roots were dehydrated with ethanol, embedded in acrylic resin (methacrylate) and cut into 5 mm longitudinal sections (manual rotary microtome, Reichert). Sections were stained with toluidine blue stain at pH 4.0 (O'Brien and McCully, 1981) and the slides were sealed using Entellan ® . Photographs of the stained sections were taken using a LM Zeiss Axioskop 2 equipped with a digital camera (MRC3). Five slides per treatment were analyzed.

Transmission electron microscopy (TEM)
Onion roots of 2 cm in length (~5 days) fixed in Karnovsky solution were rinsed in 0.05 M cacodylate buffer (3 x for 10 min) and post-fixed in 1% osmium tetroxide for 1 h. The samples were contrasted with uranyl acetate at 0.5%-v/v (overnight), dehydrated in acetone and embedded in resin (SPURR). After polymerization, ultrathin sections (70 nm) were placed on copper grids (300 mesh), counterstained with uranyl acetate and lead citrate (Reynolds, 1963), and observed under TEM (Jeol, JEM 1200EX-II) operating at 80 kV to avoid sample damage. The images were digitally captured by a CCD camera Gatan model Orius SC1000B.

Selected area electron diffraction (SAED)
Analyses of the TiO 2 NP crystal structure were made using the SAED technique with TEM (Jeol, JEM 1200EX II), and the electron diffraction figures were captured by a CCD Orius SC1000B camera. To determine the interplanar spacing of the crystal structure, the following expression was used: d (nm) = lL (nmpx)/D (px). In this formula, when measuring the distance (D) from the diffracted point to the center of the transmitted beam and knowing the camera constant (lL) and the TEM operating conditions, it is possible to measure the interplanar spacing (d), which defines certain characteristics of the crystalline material present in the sample. To determine the camera constant, which was 51.9 ± 0.2 nm pixel, a gold film was used. This technique was applied to A. cepa roots exposed to 1000 mg/L TiO 2 NP suspension to observe NP internalization. Additionally, this analysis was performed with the nanopowder form, obtained from 1000 mg/L TiO 2 NP suspension, before exposure to A. cepa roots.

Physicochemical characterization of TiO 2 NPs
The nanopowder form was analyzed by transmission electron microscopy, as well as by X-ray diffraction. The analysis showed that the crystal structures of TiO 2 NPs comprise 100% anatase phase, consisting of 28.42% titanium and 71.58% oxygen. The DLS method revealed an average size of 45 nm and 107 nm for the aggregated particles, while the BET method showed a specific area of 83.47 m 2 /g.
The analysis of the TiO 2 NP suspensions showed that their polydispersity indices were 91.1% (10 mg/L -pH 5.68), 76.5% (100 mg/L -pH 5.72), and 57.3% (1000 mg/L -pH 4.90). The size distribution of the TiO 2 NP suspensions is shown in Figure 1. The zeta potential values ranged from 21.2 to 2.99 mV, which confirms their instability.
Genotoxicity TiO 2 NP in Allium cepa 427 The SAED pattern analysis of 1000 mg/L suspension showed that all TiO 2 NPs displayed a tetragonal format, which is a feature of the anatase phase, with an average size of 25 nm. On the other hand, the SAED analysis also revealed that the internalized TiO 2 NP in the vacuole compartment of meristematic cells shows an orthorhombic format, a feature of the brookite phase, with several sizes up to 450 nm.

Seed germination and root elongation toxicity test
The results of the A. cepa toxicity test showed a significant reduction in both the germination rate (16-25%) and root development (11-18%) for all treatments with TiO 2 NP suspensions (concentration-dependent manner) ( Table 1).

Cyto-and genotoxicity assessments
A. cepa meristematic cells exposed to the TiO 2 NPs suspensions showed a significant reduction in the MI at 100 mg/L and 1000 mg/L, and a tendency for the MI to decrease at the 10 mg/L concentration in a concentration-dependent manner. Significantly higher levels of both CA and MN were also observed after exposure of the A. cepa roots to 428 Santos Filho et al. TiO 2 NP at 1000 mg/L, showing a tendency to increase CA and MN at 10 mg/L and 100 mg/L ( Table 2). The analysis of the different types of CA is shown in Table 3. The main types of CAs were the chromosomal bridge, which was significantly higher than the NC at 100 and 1000 mg/L of TiO 2 NPs, and chromosomal breaks (significant frequencies for all tested concentrations). Delayed chromosomes, chromosomal adherence, and nuclear buds were also observed in the meristematic cells of A. cepa exposed to the TiO 2 NPs, and significantly higher frequencies were observed only at the highest tested concentration (1000 mg/L).
These results indicated the genotoxicity, mutagenicity, and cytotoxicity of TiO 2 NPs and were the starting point for complementary tests that confirmed the toxicity of these NPs and allowed to observe what happens inside the cells. All these tests were performed only at the highest concentration (1000 mg/L), since the objective was the observation of intracellular mechanisms. The results are reported below.

Nucleolar organizer region (NOR) analysis
Significant increases in the NORs were observed in the A. cepa meristematic cells exposed to TiO 2 NP at 1000 mg/L ( Figure 2). Moreover, the Ag-NOR staining data show a significant increase in the nucleolar score (1.50 ± 0.09 in NC to 1.82 ± 0.27 in TiO 2 NP) and average size of the nucleoli (2221 ± 159 in NC to 2516 ± 173 in TiO 2 NP) for the A. cepa cells exposed to TiO 2 NP compared with the NC (Student's t-test, p < 0.05).

Morpho-anatomical analysis
The LM analysis showed that the NC meristematic cells are thin-walled and relatively small. These cells contain numerous small vacuoles and large nuclei with one or two nucleoli ( Figure 3A,B). With the TiO 2 NP treatment, the nuclei appear more condensed ( Figure 3C,D) with up to three nucleoli ( Figure 4D) and more and larger vacuoles.
The TEM analysis revealed that the meristematic cells from NC have organelles with peripheral disposition and a smooth cell wall. Additionally, two types of vacuoles can be characterized: with hyaline content, which are small, rounded, and have lytic vacuoles that have a higher amount of electron-dense material and lenticular shape. Oil bodies, isolated or associated with lytic vacuoles, were also observed in the cytoplasm of NC cells.
Rupture of the plasma membrane and a large number of oil bodies with peripheral disposition were observed in cells exposed to TiO 2 NP (1000 mg/L) ( Figure 4B-E). In addition, A. cepa meristematic cells exposed to TiO 2 NP also exhibited a greater number of lytic vacuoles, which were larger ( Figure 4E) compared to NC.

Selected area electron diffraction (SAED)
The electron diffraction analysis performed in A. cepa roots exposed to 1000 mg/L of TiO 2 NPs revealed the presence of TiO 2 NPs inside lytic vacuoles in A. cepa cells. These particles had an approximate size of 450 nm and, ac-Genotoxicity TiO 2 NP in Allium cepa 429 Table 1 -Seed germination and root growth inhibition of Allium cepa test system exposed to titanium dioxide nanoparticles (TiO 2 NP).      cording to the analysis of interplanar spacing and orthorhombic format, they were in the brookite phase. On the other hand, the SAED nanopowder analysis confirmed the manufacturer information saying that the TiO 2 NP in the powder form was 100% anatase phase and with an average size of 25 nm. Table 4 shows the interplanar distances (d) obtained through the analysis performed on the A. cepa roots and in the nanopowder form. Figure 4F-I shows the images of this analysis, and Figure 5 shows the two forms of TiO 2 NP (anatase and brookite phases).

Discussion
Most of the phytotoxicity studies of TiO 2 NPs do not comprise a characterization of their suspensions (e.g., TiO 2 NPs suspensions), and only characterize the nano-powder (Ghosh et al., 2010;Klancnik et al., 2010;Kumari et al., 2011;Pakrashi et al., 2014). There is an imminent demand to elucidate the mechanisms of toxicity of NPs in plants in order to protect these key organisms of terrestrial and aquatic ecosystems, as they are the base of the food chain and support several ecosystems services, such as pollination. Within this context, the knowledge of NPs toxicity mechanisms can be better achieved when nanotoxicological studies include information about the features of NPs in suspension (Schwab et al., 2016).
When in aqueous media, the pH of samples is considered one of the most important factors that may alter the values of the zeta potential (Jiang and Oberdorster, 2009). Zeta potential (z) is the measurement of the particle potential at the surface of the hydrodynamic shear. In this study, when analyzing the zeta potential values, all the suspensions presented instability in their colloidal systems, facilitating the formation of particle aggregates, even after sonication of these suspensions. Besides that, regarding the zeta potential, the existence of a pH value in which the values of the negative and positive charges are present in the same amount around the particles is known as point of zero charge (pH pzc ). The pH pzc of TiO 2 NPs in anatase phase is pH 6.3 (Finnegan et al., 2007). In this case, TiO 2 NP is considered an acidic metal oxide, which means that its highly hydroxylated surface tends to donate protons by dissociating water, binding the OHions and releasing H + ions, leaving these NPs positively charged. Our TiO 2 NPs suspensions, at the highest concentration (1000 mg/L), whose pH was measured as 4.90, probably showed many particles with positive charges, which make the entrance of these NPs into cells possible by passing cell membranes. Additionally, the high polydispersity index of TiO 2 NPs in the three concentrations tested in this study, allowed the observation of particles of different sizes. This information is important because it indicates that particle size is not homogeneous, which may explain the damage caused by TiO 2 NPs, considering the existence of particles at the nanoscale.
Phytotoxicity is usually estimated by the seed germination and root elongation toxicity test. Our findings 432 Santos Filho et al.  showed that TiO 2 NPs slightly inhibited seed germination and root growth; however, cellular and genetic damages were observed to meristematic cells of A. cepa after TiO 2 NP exposure. Studies have pointed out that determination of phytotoxicity by macroscopic parameters is not always accurate (Lin and Xing, 2007;Klancnik et al., 2010;Castiglione et al., 2016;Cox et al., 2017), and genotoxicological analysis is required to predict the hazards of chemicals. DNA damages to plant cells after NP exposure have been reported (Klancnik et al., 2010;Kumari et al., 2011;Pakrashi et al., 2014;Castiglione et al., 2016). In this study, increased frequencies of CA and MN in concentrationdependent manner were observed in A.cepa meristematic cells exposed to TiO 2 NPs, indicating their internalization. In addition, different types of CA were found. Significant values of clastogenic CAs (e.g., chromosome breaks and bridges) were observed in all tested concentrations, while aneugenic CAs (e.g., chromosome delay, chromosome adhesion, and nuclear buds) could be detected only at the highest concentration tested (1000 mg/L).
Chromosomal bridges result from structural changes between sister chromatids or between different chromosomes due to breaks or terminal deletions. Bridges that persist at the end of anaphase can originated chromosomal fragments (i.e., breaks) during chromatin segregation (Humphrey and Brinkley, 1969). Chromosome breaks can also be caused by external agents, affecting the dynamics of the chromatin, and may damage the repair process (Terzoudi et al., 2011). Impairment of the mitotic spindle apparatus may lead to chromosomal adherences (Ventura-Camargo et al., 2011). Adherence is an irreversible abnormality that involves the proteinaceous matrix of chromatin rather that DNA itself, usually leading to cell death (Fiskesjo, 1988). The interruption in mitotic spindle polymerization may promote a unilateral binding of the fuse to chromosomes, making their movement to the poles unfeasible and leading to chromosomal losses (Shamina et al., 2003).
Chromosomal losses and breaks can originate in micronuclei and be from an aneugenic or clastogenic origin, while nuclear buds may originate from nuclear envelope formation prior to complete chromosome migration to the poles and their incorporation into the nuclei, as well as by cellular activities that promote the elimination of the amplified genetic material (Mazzeo et al., 2011).
These genotoxicity results stimulated the accomplishment of complementary tests that allowed the observation of TiO 2 NPs internalization, which corroborated the toxicity results of this NP. These complementary tests were performed only with the highest concentration, because we aimed to understand TiO 2 NPs internalization and observe their effects. The significant increase in nucleolar score, as well as their sizes, in A. cepa interphase cells exposed to 1000 mg/L TiO 2 NP suggests that these results are due to an increase in genome activity (Mazzeo and Marin-Morales, 2015). According to Boulon et al. (2010), nucleoli are apparently major structures involved in the activation of cellular stress. Increased nucleoli volume suggests gene amplification and may be another indicative of genotoxicity (Mazzeo and Marin-Morales, 2015).
The cellular damage observed in A. cepa cells exposed to TiO 2 NPs indicates that this NP was taken up by the meristematic cells, causing deleterious effects. This study demonstrated that NPs were internalized by the meristematic cells of A. cepa. This internalization probably occurred due to the excess of positively charged particles that were easily attracted and internalized by the plasma membrane (negatively charged), allowing the observation of negative effects in these cells. These data agree with other studies of NPs (zinc oxide NPs and TiO 2 NPs) (Kumari et al., 2011;Larue et al., 2012), in which the authors also reported that NP uptake may result in damages from cellular defense mechanisms. The present results indicate that roots exposed to TiO 2 NPs show damage to both the plasma membrane and cell wall, suggesting that these barriers are not effective against TiO 2 NP uptake by meristematic cells.
Once the structure and function of both the cell wall and plasma membrane have been compromised, a physical barrier was formed by oil bodies located beneath them to reduce the uptake of TiO 2 NPs. According to Zhao et al. (2016), the increased number of oil bodies may indicate a cellular defense mechanism against toxicants. Moreover, a rise in the number and size of lytic vacuoles in the cytoplasm may also be related to this cellular defense mechanism, since this cellular compartment acts as a primary deposit site of toxic compounds . Defense mechanisms are common when the plant encounters an adverse situation, like an exposure to a contaminant (Medina et al., 2016;Car et al., 2019). The SAED analysis showed the presence of TiO 2 NPs in lytic vacuoles of A. cepa meristematic cells that were exposed to this agent, as well as in Triticum aestivum spp (TiO 2 NPs at 100 mg/L), as observed by Larue et al. (2012).
The TEM analysis carried out on A. cepa cells exposed to 1000 mg/L TiO 2 NP showed that the deposit of this NP in lytic vacuoles occurs in aggregates of ca. 450 nm. The SAED results showed an orthorhombic structure compatible with the brookite phase, differently from a tetragonal structure expected for the anatase phase. This finding suggests that TiO 2 NPs were taken up by the cells and that a phase change had occurred (anatase to brookite).
According to the literature, the anatase and brookite phases are metastable and can switch their shape (Alotaibi et al., 2018). The phase change showed in this work can be related to cellular defense mechanisms, as an attempt to minimize the damages caused by the internalized TiO 2 NPs. Therefore, the anatase to brookite phase conversion would be a way to "mitigate" the deleterious effects of TiO 2 NPs probably by reducing the damage mediated by reactive ox-ygen species (Larue et al., 2012). It can be hypothesized that the interaction with biological molecules (e.g., enzymes) produced by the plant system may be the responsible factor for the phase change (anatase to brookite) caused by TiO 2 NPs in A. cepa meristematic cells. However, further studies are needed to elucidate the crystalline phase change of TiO 2 NPs inside plant cells.
Finally, A. cepa was shown to be sensitive to the genotoxic and cytotoxic effects of TiO 2 NPs, thus being a suitable test system for predicting the hazard potential of NPs to plants. To minimize the toxic effects of TiO 2 NPs, plant cells exhibit cellular defense mechanisms that include increasing the number of oil bodies and lytic vacuoles. Furthermore, phase transformation of the crystal structure from anatase to brookite may also be an attempt to mitigate the toxic potential of TiO 2 NPs. In spite of the defense mechanisms, these NPs are still able to induce severe damages at nuclear (genotoxicity; changes in the nucleolar pattern) and cellular levels (cytotoxicity) in a concentrationdependent manner, which is indicative of their internalization. Probably, NPs were taken up by the meristematic cells through disruption of physical barriers (plasma membrane and cell wall). The phase conversion of TiO 2 NPs from anatase to brookite inside a plant cell was reported here for the first time, but the mechanisms associated with this change need to be elucidated by further studies.