The Effects of Fe2O3 Nanoparticles on Physiology and Insecticide Activity in Non-Transgenic and Bt-Transgenic Cotton

As the demands for nanotechnology and nanoparticle (NP) applications in agriculture increase, the ecological risk has drawn more attention because of the unpredictable results of interactions between NPs and transgenic crops. In this study, we investigated the effects of various concentrations of Fe2O3 NPs on Bt-transgenic cotton in comparison with conventional cotton for 10 days. Each treatment was conducted in triplicate, and each experiment was repeated three times. Results demonstrated that Fe2O3 NPs inhibited the plant height and root length of Bt-transgenic cotton and promoted root hairs and biomass of non-transgenic cotton. Nutrients such as Na and K in Bt-transgenic cotton roots increased, while Zn contents decreased with Fe2O3 NPs. Most hormones in the roots of Bt-transgenic cotton increased at low Fe2O3 NP exposure (100 mg⋅L-1) but decreased at high concentrations of Fe2O3 NPs (1000 mg⋅L-1). Fe2O3 NPs increased the Bt-toxin in leaves and roots of Bt-transgenic cotton. Fe2O3 NPs were absorbed into roots, then transported to the shoots of both Bt-transgenic and non-transgenic cottons. The bioaccumulation of Fe2O3 NPs in plants might be a potential risk for agricultural crops and affect the environment and human health.


INTRODUCTION
Iron oxide (Fe 2 O 3 ), the most common oxide of iron, has important magnetic properties. Iron (III) oxide is a convenient compound for the general study of polymorphism and the magnetic and structural phase transitions of NPs. Fe 2 O 3 NPs can be applied in the fields of photoelectrochemistry (such as solar energy conversion and water splitting) and photocatalysts for the removal of organic and inorganic species from aqueous or gas phases (Chirita and Grozescu, 2009). Ali et al. (2012) reported that the addition of Fe 2 O 3 NPs in cement could improve the strength and water permeability of the specimens. Introducing Fe 2 O 3 NPs into soil could significantly increase root elongation and photosynthesis rate in soybean as compared with bulk Fe 2 O 3 . Similar results have also been reported for rice seedlings (Oryza sativa L. var. Koshihikari) treated with 500, 1000, and 2000 mg·L −1 γ-Fe 2 O 3 NPs when compared with the control and bulk treatments (Alidoust and Isoda, 2014). Thus, nano-size effects might enhance plant growth relative to bulk metal. With the development of nanotechnology, a rapidly growing body of concerns has been raised regarding the potential risks and negative effects of NPs on the environment and human health.
The toxicity of metal NPs is not clearly known to living organisms, but previous studies have shown that nanotoxicity is generally affected by unique properties (particle size, shape, and surface properties) (Crane et al., 2008;Navarro et al., 2008). Nanotoxicity has been tested using various organisms, including bacteria, algae, protozoa, plants, and fish (U.S. EPA, 1993). The metabolic processes in living organisms have been investigated to assess its toxicity to the environment (Bitton, 1999). However, in the plant kingdom, investigation of phenotypic differences, seed germination, and plant biomass in response to NP exposure could be one of the most effective ways to assess NP toxicity (Wang and Liu, 2001;Boutin et al., 2004;Di Salvatore M, 2008;Parsons et al., 2010). For example, alumina (Al 2 O 3 ) NPs can cause phytotoxicity by inhibiting root elongation in corn, cucumber, soybean, cabbage, and carrot (Yang and Watts, 2005). This phytotoxicity is also evident for other plant species, including radish, rape canola, ryegrass, lettuce, corn, and cucumber, when treated with multiwall carbon nanotubes, aluminum (Al), alumina (Al 2 O 3 ), zinc (Zn), and zinc oxide (ZnO) NPs (Lin and Xing, 2007). In addition to NP-induced toxicity in plants, different plant species could respond differently in the same NP exposure. In a study by Li et al. (2014), CeO 2 NPs were revealed to have toxic effects on the root biomass of Bt-transgenic cotton under 100 and 500 mg·L −1 exposures, but to have no effects on conventional cotton. In addition, Le et al. (2014) reported that SiO 2 NPs negatively affected the activities of CAT in the roots of both Bt-transgenic and non-transgenic cottons.
Insect pests can significantly reduce crop yields and subsequently cause economic losses in agriculture all over the world. At present, chemical pesticides are still the major method to control pest damage; however, the disadvantages of chemical pesticide usage in agriculture include the fact that pests becomes resistant to the chemicals and overuse of pesticides can cause serious risks to the environment and human health . As such, transgenic plants have been widely applied in agriculture for the purpose of controlling pest damage. For example, over expression of Bacillus thuringiensis (Bt) insecticidal protein in cotton can significantly enhance the resistance of cotton to insects and control insect damage (Wang, 2007;Huang et al., 2010;Yu et al., 2014). Currently, Bt-transgenic cotton is widely used in agriculture, and China is the largest cotton producing country in the world (Du, 2001). However, as demand for nanotechnology and NP applications in agriculture increases, the ecological risks are drawing attention because of the unpredictable results in interactions between NPs and transgenic crops.
To our knowledge, this is the first study on the effects of Fe 2 O 3 NPs on Bt-transgenic cotton. In this study, the toxic effects of Fe 2 O 3 NPs on both conventional and Bt-transgenic cottons were investigated from the aspects of plant growth, nutrient levels, hormone levels, and the changes to Bt-toxic protein in Bt-transgenic cotton in the presence of Fe 2 O 3 NPs. Fe 2 O 3 NP accumulation and distribution were also assayed to reveal how NPs are distributed inside cotton. TEM images showed that Fe 2 O 3 NPs were found in the roots of both conventional and transgenic cotton.

Characterization of Fe 2 O 3 Nanoparticles
Fe 2 O 3 NPs were purchased from Shanghai Hufeng Bioscience Technology Company (Shanghai City, China). Scanning electron microscopy (JEOL JSM 5600, Japan) was used to determine the morphology of Fe 2 O 3 NPs. TEM images were obtained (JEM 200CX, Japan) at 200 kV. Following Amrut et al. (2010), the samples were prepared by dispersing drops of the colloid on a copper grid, which was then covered with a carbon film, and the solvent was evaporated. Fe 2 O 3 NP suspensions were prepared at a concentration of 2 mg·L −1 for measurement of hydrodynamic size and zeta potential (Nicomp 380 DLS Zeta potential/Particle system, Santa Barbara, CA, USA).

Experimental Exposure
Bt-transgenic cotton (Bt-29317) and conventional cotton (Jihe 321) were purchased from the Chinese Academy of Agricultural Sciences, China Agricultural University. Four cotton plants were allowed to acclimatize in a pot containing 2.0 L of nutrient solution for 4 days and then exposed to 0, 100, or 1000 mg·L −1 Fe 2 O 3 NPs for 10 days. The nutrient solution was made following previous reports by Li et al. (2014). Each treatment was performed in triplicate, and each experiment was repeated three times. Fe 2 O 3 NPs were dispersed in DI H 2 O for 30 min using an ultra-sonicator (KQ3200DE) (Le et al., 2014;Li et al., 2014). The experiments were conducted in a greenhouse with natural light, humidity, and temperature at China Agricultural University.

Measurement of Biomass, Plant Height, Root Length, and Root Hairs
The plant height (mm) and root length (mm) were measured from the growth point in shoot to the cotyledon node and from the growth point to the root point, respectively, after 10 days of exposure to Fe 2 O 3 NPs (Le et al., 2014). The root hairs (root) were counted one by one in each cotton plant.
The samples were collected at day 10. The shoots and roots were rinsed with both tap water and DI H 2 O for 5 min three times and then dried separately at 80 • C for 24-36 h until a constant dry weight was obtained. The dry weight was used to determine the effects of the Fe 2 O 3 NPs on plant growth.

Measurement of Nutrient Contents
Nutrient contents were determined using inductively coupled plasma (ICP) mass spectrometry (DRC-II) and ICP atomic emission spectroscopy (iCap 6000) .
Fine powders (20-30 mg) of oven-dried shoots and roots were digested in 5 mL of HNO 3 at room temperature for 24 h; then, 3 mL of H 2 O 2 was added to each sample to further digest samples at 180 • C for 4-5 h.

Determination of Hormone Concentration
Extraction and purification of ABA, IAA, ZR, and GA were performed with enzyme-linked immune absorbent assay (ELISA) kits using monoclonal antibodies (Phytodetek, Agdia, Elkhart, IN, USA) as described in He et al. (2005) and Dong et al. (2008). All samples were measured at 405 nm after the purification step (Wang et al., 2012).

Determination of Bt Toxin
Approximately 0.5-g samples of roots and leaves of Bt-transgenic cotton were harvested and stored at −40 • C until analysis. The samples were placed in 9.5 mL of extraction solution (1.33 g of Na 2 CO 3 , 1.46 g of NaCl, 0.5 g of vitamin C, and 0.25 g of dithiothreitol in 250 mL of DI H 2 O). The mixture was gently shaken for 30 min before being centrifuged at 4,000 rpm for 15 min. The supernatants were used to analyze the levels of the Bt toxin protein using an ELISA procedure (Rui et al., 2005).

Transmission Electron Microscopy Observation
Fe 2 O 3 NP-treated shoot and root samples of both conventional and Bt-transgenic cotton were harvested and washed with DI H 2 O at day 10. Each sample was prefixed in 2.5% glutaraldehyde, then washed in 0.1 mol·L −1 pH 7.0 phosphate buffer mixed with 1% osmium tetroxide for 2 h before being dehydrated in a graded ethanol series and finally embedded in epoxy resin. The samples for TEM observation were sectioned using the procedure described in Zhang et al. (2012).

Data Analysis
The results are presented as the mean ± SD. One-way analysis of variance was used to calculate statistical analysis (SPSS 22.0 software). A confidence interval of 95% (p < 0.05) was considered significant in all cases.

Characterizations of Fe 2 O 3 NPs
Fe 2 O 3 NPs were purchased from Sigma Inc. The advertised diameter was smaller than 50 nm, the specific surface area was from 50 from 245 m 2 g −1 , and the density was 5.25 g·cm 3 ; the crystal phase was cubic. Fe 2 O 3 NPs were characterized by scanning electron microscopy (JEOL JSM 5600, Japan) and dynamic light scattering (DLS). In addition, Fe 2 O 3 NPs were prepared at concentration of 2 mg·L −1 for measurement of hydrodynamic size (154.3 nm) and zeta potential (−9.27 mV; Nicomp 380 DLS Zeta potential/Particle system, Santa Barbara, CA, USA). Figure 1 gives the SEM image; the diameter was larger than the advertised data.

Effects of Fe 2 O 3 NPs on the Growth of Cotton Plants
In control groups, no significant difference in plant height was observed between conventional and Bt-transgenic cotton. However, a decrease in plant height for Bt-transgenic cotton was shown with both 100 and 1000 mg·L exposure doses of Fe 2 O 3 NPs, while similar phenotypic difference was only observed in 1000 mg·L Fe 2 O 3 NP-treated conventional cotton (Figure 2A). Although root length in conventional cotton is approximately 12.5% longer than in Bt-transgenic cotton without the Fe 2 O 3 NP exposure, a significant decrease in the root length in 100 mg·L Fe 2 O 3 NP-treated transgenic cotton was evident. The opposite phenomenon in Fe 2 O 3 NP-treated conventional cotton was attained ( Figure 2B). Similarly, a low exposure dose of Fe 2 O 3 NPs could stimulate development of root hairs in conventional cotton as compared with its respective control group, while no difference was found in transgenic cotton among the treatments in Figure 2C. Regarding fresh biomass of separated shoots and roots in both types of cottons, shoot biomass in conventional cotton was similar among treatments; however, root biomass in Fe 2 O 3 NP-treated conventional cotton was 30.8-41.2% higher relative to its control group (Figures 2D,E). Analysis of shoots and roots biomass in Bt-transgenic cotton showed that Fe 2 O 3 NPs seemed to have no impact on biomass. These results suggest that Bt-transgenic cotton is more sensitive to NP exposure than conventional cotton. A similar result was reported by Li et al. (2014), in which 100 and 500 mg·L −1 CeO 2 NPs could cause more toxicity to Bt-transgenic cotton by reducing root biomass, while no impact was found in non-transgenic cotton. However, phytotoxicity might be dependent on NP species. For example, as SiO 2 NP concentration increased up to 2000 mg·L −1 , the plant height and biomasses of both conventional and Bt-transgenic cottons significantly decreased (Le et al., 2014). Other NPs could also impact plant biomass, such as ZnO and CuO NPs. Ryegrass (Loliumperenne) biomass was significantly reduced by

Effects of Fe 2 O 3 NPs on Nutrient Contents in Cotton Plants
Nutrient contents (macronutrients and micronutrients) in both types of cotton treated by Fe 2 O 3 NPs were investigated. K is one of the most important micro nutrients in plants. K contents in both types of cottons were increased in the 100 mg·L −1 Fe 2 O 3 NP treatment, although this increase was not statistically significant in Bt-transgenic cotton compared with its corresponding control ( Figure 3A). Similar trends were also observed for Ca contents in shoots (Figure 3C). Fe 2 O 3 NPs inhibited Na uptake in shoots of conventional cotton at both 100 mg·L −1 and 1000 mg·L −1 , whereas an increase in Na in the shoots was found for 100 mg·L −1 Fe 2 O 3 NP-treated Bt-transgenic cotton ( Figure 3B). Cotton type could also determine the nutrient content. For example, in the absence of Fe 2 O 3 NPs, Mg content in conventional cotton was approximately half that of transgenic cotton. When treated with various concentrations of Fe 2 O 3 NPs, increases in Mg were observed in conventional cotton shoots treated with Fe 2 O 3 NPs, whereas a decrease in Mg in 100 mg·L Fe 2 O 3 NPtreated Bt-transgenic shoots was found ( Figure 3D). Similar results were also found in B content in both conventional and transgenic cottons (Figure 3I). Fe 2 O 3 NPs significantly decreased Mo contents in shoots of both conventional and Bt-transgenic cottons ( Figure 3G). Other elemental contents, including Mn, Zn, P, and Cu (Figures 3E,F,H,J), showed no change in either type of plant in the presence of Fe 2 O 3 NPs. These results suggest that Fe 2 O 3 NPs could more seriously interrupt nutrient uptake in Bt-transgenic cotton as compared with conventional cotton. Figure 4 shows the trend of nutrient uptake in roots of both types of cotton. Fe 2 O 3 NPs significantly enhanced K uptake in the roots of both Bt-transgenic and non-transgenic cotton ( Figure 4A). A significant increase in Na content was only observed in the roots of Bt-transgenic cotton in Figure 4B. Fe 2 O 3 NPs only caused a decrease in Ca uptake in conventional cotton, whereas 100 mg·L −1 Fe 2 O 3 NPs enhanced Ca content in Bt-transgenic root ( Figure 4D). The opposite results were found for Mg content in both types of cotton in Figure 4D. In conventional cotton, Fe 2 O 3 NPs significantly stimulated Mg uptake at both concentrations, while a decrease in Mg content was evident for Bt-transgenic cotton relative to its corresponding control. No significant effects of Fe 2 O 3 NPs were seen for Mn content in roots, regardless of type of cotton, although conventional cotton could take up more Mn than Bt-transgenic cotton in the absence of Fe 2 O 3 NPs ( Figure 4E). Fe 2 O 3 NPs decreased the Zn content in the roots of Bt-transgenic cotton, while no significant difference was reported in Zn content for conventional cotton roots ( Figure 4F). The presence of 100 mg·L −1 of Fe 2 O 3 NPs could enhance Cu uptake in the roots of both types of cottons, and Cu contents in both plants slightly decreased at 1000 mg·L −1 (Figure 4J). Fe 2 O 3 NPs decreased Mo  contents in the roots of both conventional and Bt-transgenic cottons at 1000 mg·L −1 (Figure 4G). P contents were not affected by Fe 2 O 3 NPs (Figure 4H), except for 1000 mg·L −1 Fe 2 O 3 NP-treated conventional cotton, whose P content was slightly lower relative to its corresponding control group. Fe 2 O 3 NPs had no impact on B content in roots, regard less of plant types ( Figure 4I).

Effects of Fe 2 O 3 NPs on Hormone Concentration in Cotton Plants
The effects of Fe 2 O 3 NPs on plant hormones in conventional and Bt-transgenic cottons are shown in Figures 5 and 6. IAA contents in leaves of both types of cottons significantly increased (p < 0.05) at an exposure dose of 100 mg·L −1 Fe 2 O 3 NPs, but no change was found at 1000 mg·L −1 Fe 2 O 3 NP treatment (Figure 5A).
A decrease in ABA concentration was only found in the Fe 2 O 3 NP-treated conventional cotton, while Fe 2 O 3 NPs had no impact on ABA concentrations in the leaves of Bt-transgenic cotton. In addition, no significant difference in ABA concentration was seen between Bt-transgenic and non-transgenic cotton leaves under Fe 2 O 3 NP exposure ( Figure 5B). Similar results were found for ZR concentration in the leaves of both cotton plants in Figure 5D. The GA content in the leaves of conventional cotton declined for both Fe 2 O 3 NPs treatments, whereas it significantly increased in the leaves of Bt-transgenic cotton ( Figure 5C).
Fe 2 O 3 NPs enhanced the concentrations of all four plant hormones (including IAA, ABA, GA, and ZR) in the roots of Bt-transgenic cotton, especially at 100 mg·L −1 Fe 2 O 3 NP exposure. However, IAA, ABA, and ZR concentrations in the roots of Fe 2 O 3 NP-treated conventional cotton significantly decreased. In addition, Fe 2 O 3 NPs had no impact on GA concentration in the roots of conventional cotton (Figure 6). This suggests that the root hormone concentrations in conventional and Bt-transgenic cotton had different responses to Fe 2 O 3 NP exposure. Le et al. (2014Le et al. ( , 2015 reported that IAA and ABA concentrations in the roots of conventional and Bt-transgenic cottons were altered by SiO 2 and CeO 2 NPs. Gui et al. (2015) reported that phytohormones such as IAA and ABA intransgenic and non-transgenic rice (O. sativa) were affected by 2, 20, and 200 mg·L −1 Fe 2 O 3 NP exposures. IAA stimulates growth processes such as cell elongation and division, whereas ABA controls plant senescence and responses to stress (Davies, 1995;Mok and Mok, 2001). Thus, our results further demonstrate that NPs could have negative effects on the plant growth process, development, and senescence by manipulating the phytohormone concentration in plants. Figure 7 shows the total Fe contents in shoots and roots of both conventional and Bt-transgenic cotton treated with two concentrations of Fe 2 O 3 NPs. No difference was seen in Fe contents in both cotton shoots for treatment with 100 mg·L −1 Fe 2 O 3 NPs when compared with the control group. However, as the concentration of Fe 2 O 3 NPs increased to 1000 mg·L −1 , Fe contents in both cotton varieties were significantly higher than the control groups, and Fe content in the Bt-transgenic cotton was significantly higher than that in conventional cotton ( Figure 7A). Fe contents in roots of both Fe 2 O 3 NP-treated cottons are shown in Figure 7B. Fe content in the roots of Bttransgenic cotton treated with 1000 mg·L −1 Fe 2 O 3 NPs was 5.30 times higher than in its corresponding control, while this value was 2.8 times for 1000 mg·L −1 Fe 2 O 3 NP-treated conventional cotton as compared with its corresponding control. In addition, Fe content in Bt-transgenic cotton roots was 1.27 times higher than that of conventional cotton at 1000 mg·L −1 Fe 2 O 3 NP treatment. This indicates that the Bt-transgenic cotton was capable of taking up more Fe relative to the conventional cotton, which agrees with a previous study  in which CeO 2 NPs aggregates more easily penetrated into the roots of Bt-transgenic cotton than conventional cotton. A small amount of Fe 2 O 3 NPs was transported to the shoots. Our results are in agreement with previous studies that showed that CuO NPs were taken up by maize through the root system (Wang et al., 2012) and CeO 2 and SiO 2 NPs were transported to the shoots from the cotton root system (Le et al., 2014;Li et al., 2014).

Fe contents in the Shoots and Roots of Cotton Plants
Transmission electron microscopy images of the root sections of Bt-transgenic and non-transgenic cotton show the presence of dark dots (particles) primarily localized in the endodermis and vascular cylinder with 1000 mg·L −1 Fe 2 O 3 NP exposure (Figure 8). The absorption of Fe 2 O 3 NPs and their aggregation in the roots of both conventional and Bt-transgenic cottons are evident. Most Fe 2 O 3 NPs were found in the root outer epidermis, and only a few NPs were localized in the intercellular spaces. Thus, it can be concluded that the Fe 2 O 3 NPs could enter the endodermis and vascular cylinder of both Bt-transgenic and nontransgenic cotton. Previous studies also demonstrated that SiO 2 and CeO 2 NPs were localized in the intercellular spaces of both types of cotton (Le et al., 2014;Li et al., 2014). ZnO NPs were also observed in the endodermis and vascular cylinder of ryegrass roots (Lin and Xing, 2008). Adamczyk and Sumerford (2001) reported that the performance of Bt genes for controlling target insect pests varies according to cotton variety, age of plant (Wan et al., 2005), part of plant (Abel and Adamczyk, 2004), type of gene, and insertion site of the gene into the DNA of target plants (Gore et al., 2001;Gore and Adamczyk, 2004;Jackson et al., 2004). Bt toxin is the product of an exogenous Bt gene, whose concentration in shoots is the most important index for evaluating the insect resistant ability of Bt cotton; however, if the root expresses more Bt toxin, it can negatively affect the soil ecological system (Saxena et al., 1999;Obrycki et al., 2001;Fu et al., 2012). Figure 9 shows the expression of Bt toxin in leaves and roots of Bt-transgenic cotton in the presence of Fe 2 O 3 NPs. A significant difference in the contents of Bt toxin was seen in both leaves and roots of Bttransgenic cotton treated with various concentrations of Fe 2 O 3 NPs as compared with the control group. Bt toxin concentrations in leaves and roots were 845.89 and 886.94 ng·g −1 , respectively, at the exposure dose of 100 mg·L −1 Fe 2 O 3 NPs, which were 1.61 and 1.36 times higher than their respective control group. Upon exposure to 1000 mg·L −1 Fe 2 O 3 NPs, the Bt toxin in Bttransgenic cotton leaves and roots were still greater (p < 0.05) than that of the control group, although they were lower than in the lower exposure dose of Fe 2 O 3 NP-treated Bt-transgenic cotton. The expression of Bt toxin in Bt-transgenic cotton leaves increased, especially at 100 mg·L −1 Fe 2 O 3 NPs, which could play an important role in resisting insect damage. To reduce the risk of resistance development in target insect pests against Bt cotton and agricultural crops, there is a need to understand variations in the efficiency of Bt genes and the application of NPs in agriculture.

CONCLUSION
The present study demonstrated that Fe 2 O 3 NPs could inhibit the plant height and root length of Bt-transgenic cotton, as well as promote the root hairs and biomass of conventional cotton. The effects of Fe 2 O 3 NPs on nutrient contents in the shoots and roots of both types of cotton were investigated. Fe 2 O 3 NPs enhanced Na content in the roots of Bt-transgenic plants, and similar results were observed for K contents in the roots of both Bt-transgenic and non-transgenic cotton. Zn contents in the roots of Bt-transgenic cotton decreased upon Fe 2 O 3 NP exposure, while Cu contents in the roots of both types of cotton increased upon exposure to 100 mg·L −1 Fe 2 O 3 NPs and decreased at 1000 mg·L −1 Fe 2 O 3 NPs. Responses of hormone concentrations in the presence of Fe 2 O 3 NPs differed between the leaves and roots of both types of cottons. Most hormones in the roots of Bt-transgenic cotton increased at low Fe 2 O 3 NP exposure (100 mg·L −1 ), but decreased at a high concentration of Fe 2 O 3 NPs (1000 mg·L −1 ). In addition, Bt-toxin in the leaves and roots of both Bt-transgenic and non-transgenic cotton increased upon NP exposure. TEM images shows that Fe 2 O 3 NPs were evident in the root sections of both Bt-transgenic and non-transgenic cotton, and Fe contents in the shoots and roots increased with increasing exposure doses of Fe 2 O 3 NPs. The present study illustrates the bioaccumulation of Fe 2 O 3 NPs in plants, which might have potential risks for agricultural crops and affect the environment and human health.

AUTHOR CONTRIBUTIONS
LN and YR conceived and designed the experiment. CM provided scientific expertise. LN, CM, YR, WC, YD, LL, and BX performed the experiments and analyzed the data. LN, CM, and YR wrote and revised the paper. All authors have read and approved the final manuscript.