As(III)-Oxidizing Bacteria Alleviate Arsenite Toxicity via Reducing As Accumulation, Elevating Antioxidative Activities and Modulating Ionome in Rice (Oryza sativa L.)

Paddy rice trends to accumulate more arsenic (As) from soils than other terrestrial crops. The toxicity and mobility of As mainly depend on its chemical species. Transformation of arsenite [As(III)] into arsenate [As(V)] would be a promising method to mitigate As toxicity. In the current study, As(III)-oxidizing strain SMS11 isolated from As-contaminated soils was employed for As remediation. Co-cultured with SMS11 alleviated As(III) stress to the rice plants by increasing the length and biomass of rice shoots up to 10% and 15%, respectively. Evaluation of oxidative stress indices showed that the activity of catalase in the rice shoots was weakened when exposed to As(III), increasing the risk of hydroxyl radical (·OH) formation. When co-cultivated with the bacteria, ·OH formation was significantly inhibited in the rice shoots. The ionomes of the rice plants were impacted by the external conditions. As(III) stress significantly disturbed ionome homeostasis in the rice plants. Uptake of As simultaneously elevated the levels of macro and nutrient elements such as Mg, P, K, Ca, and Zn in the rice shoots. The ionomic variation in the rice plants under As(III) stress was mitigated by inoculated with SMS11. The results represented that the As(III)-oxidizing bacteria alleviated external As(III) stress to the rice plants through elevating antioxidative activities and modulating ionome homeostasis.


Introduction
Different from other terrestrial crops, paddy rice trends to accumulate much more arsenic (As) from paddy field due to its flooded growing environment [1]. Once entering into the rice plant, As could not only influence the rice yield through inhibiting the plant growth, but also been transferred to rice grains, threatening human health through food chain [2,3]. Consumption of rice and its products is considered as one of the important ways for people expose to As [4][5][6]. The toxicity of As is dependent on its chemical species. Inorganic As forms including arsenite [As(III)] and arsenate [As(V)] are the dominant As species existed in the paddy field [7][8][9]. As(III) is the most toxic As species, followed by As(V). Moreover, inorganic As forms were recognized as non-threshold Class I carcinogens which can lead to serious health risk once uptake by human bodies [10,11]. In comparison, organic As forms are less toxic to human beings.
Translocation and transformation of As are also dependent on its chemical species [12,13]. Influx of As(V) into rice roots shares the same pathways as phosphate via phosphate transporters [14]. In rice roots, As(V) can be rapidly reduced to As(III) with the assistance of arsenate reductase, leading to dominant proportion of As(III) in the rice seedlings [13,15]. As(III) with greater toxicity and mobility than As(V) is predominant under anaerobic conditions. Influx of As(III) into rice plants was described through aquaporin channels due to its analogous properties with silicic acid [16]. Once entering into rice roots, As(III) could be sequestrated in the vacuoles for detoxification, transferred to the aerial parts or excluded to the environment via the transporter OsLsi2 [17]. Based on the specific transporters for different As species, silicon-and phosphorus-based additives have been widely applied in mitigation of As(III) and As(V) stresses to rice plants, respectively [18,19]. Moreover, iron-containing modifiers such as Fe oxide/hydroxide also exhibited positive effects on sequestering As in rice roots, reducing As solubility under flooded conditions and decreasing As content in rice grains [19]. Although these additives exhibited effective mitigation of As stress to the rice plants, relatively high cost restricted large scale application of the inorganic/organic amendments [20]. By contrast, microbiological remediation technologies are less costly and more eco-friendly. In the environment, microorganisms play a key role on As speciation and migration [21]. As(V) with high affinity to Fe oxide/ hydroxide is less toxic and less mobile than As(III). As(III)oxidizing bacteria isolated from As-contaminated soils possessed the capacity to convert As(III) to As(V) much more efficiently than natural oxidation process [22,23]. Therefore, oxidizing more mobile As(III) to less toxic As(V) was confirmed a feasible measure to control As contamination and restrict As translocation into rice roots [24,25].
Plant as an integral whole tends to keep the ionome homeostasis when exposed to external stress. Ionome defined as the total of elements is a dynamic network in plant [26]. It is noticed that elements in plant are interdependent and covary under external conditions. For instance, uptake or efflux of As in rice plants could not only change As content, but also alter other ionic balances [27]. In previous studies, As(III)oxidizing bacteria have been employed to alleviate As stress to rice plant [24,25]. However, few studies were concerning the variations of ionomes. Compared to individual element, ionome is more sensitive to environmental conditions [26]. Strong connection between ionomic profile and external stress has been observed in the rice plant tissues treated with different As species [28]. Therefore, investigation of ionomic variation is important to reveal the response of rice plant to external As stress. In this work, the As(III)-oxidizing strain SMS11 described in our previous study [22] was employed to mitigate As(III) stress to rice plants. The plant biomass, As translocation and As speciation were measured to evaluate the impacts on the rice plants with and without bacterium inoculation. Ionome and oxidative stress indices in the rice plant tissues were investigated to discuss the potential response mechanism to external stress.

As(III)-Oxidizing Bacterial Culture Preparation
The As(III)-oxidizing strain SMS11 (genus Pseudomonas, NCBI GenBank accession No. MW940738) obtained in our previous study [22] was introduced in this work. Prior to inoculation, the strain was cultured in 100 mL of chemically defined medium (CDM, the detailed information is provided in Text 1 of Supplementary Materials) at 150 rpm and 25 °C till logarithmic phase (OD 600 = 0.4-0.5). The cultures were centrifuged at 4000 rpm for 10 min and the supernatants were discharged. Then the collected strain cells were resuspended in normal saline (0.85% NaCl solution) for further inoculation.

Plant Cultivation and Harvest
A japonica cultivar rice (Oryza sativa L. var. Nipponbare) was obtained from Hunan Agricultural University. Before germination, the seeds were sterilized by 1% active NaClO solution for 15 min and then rinsed with ultrapure water several times. Seed germination was performed at 28 °C for 7 days in an incubator. A series of growing-well and uniform seedings were selected and then transferred to 1-L black plastic bottles filled with Yoshida nutrient solution [29]. The composition of Yoshida nutrient solution is listed in Table S1. The seedlings were transferred in a growth chamber and grown at 28/20 °C light/dark with a 16-h photoperiod for 15 days. The nutrient solution in the bottles was replaced every 3 days. After that, the rice seedlings were divided into three groups for subsequent exposure experiment in different cultivation solutions: (1) Control group, nutrient solution without any addition; (2) As(III) group, nutrient solution amended with 1 mg L −1 As(III); (3) As(III) + B group, nutrient solution supplemented with 1 mg L −1 As(III) and 2% (v:v) prepared cell suspensions. The culture solutions were refreshed every 3 days. In order to confirm the variation of As species, the culture solutions were collected each 3 days before renewal. The collected solutions were stored at − 20 °C for As speciation.
The rice seedlings were harvested after another 15 days of growth under the same condition. To remove apoplastic As, the rice roots were soaked in ice-cold desorption solution consisting of 1 mM K 2 HPO 4 , 0.5 mM Ca(NO 3 ) 2 , and 5 mM MES (pH 6.0) for 10 min. Then the rice seedlings were washed thoroughly, rinsed with ultrapure water, blotted dry and separated to roots and shoots. Length and biomass of the fresh rice roots and shoots were recorded. Ten randomly seedlings for each group were selected and kept at − 80 °C for reactive oxygen species (ROS) and antioxidant analysis. Three replicates from the remaining rice plant tissues for each group were freeze dried, homogenized and then kept at − 20 °C for As speciation and ionomic analysis.

Element Analysis
The rice tissues were digested according to a microwaveassisted method modified in our previous report [32]. Briefly, approximately 0.200 g (for shoot tissue) or 0.100 g (for root tissue) of the homogenized sample was mixed with digestion reagent containing 9 mL of concentrated HNO 3 and 2 mL of 30% H 2 O 2 in a polytetrafluoroethylene (PTFE) sealed vessel. The digestion program was as follow: 15 min to 120 °C, 15 min to 190 °C and 30 min at 190 °C. After cooling down, the digested solution was passed through a 0.45 μm membrane and diluted in a 25-mL volumetric flask with ultrapure water. The digestion procedure was performed in triplicate for each rice tissue in each group. Total concentrations of 20 elements including B, Mg, Al, P, K, Ca, V, Cr, Mn, Fe, Co, Ni, Cu, Zn, As, Sr, Mo, Cd, Sb, and Pb were determined using ICP-MS (Agilent 7700x, Tokyo, Japan). The detailed operating parameter for ICP-MS is listed in Table S2.

As Speciation Analysis
Extraction and determination of As species in the rice plant tissues were performed according to our previous method [32]. Approximately 0.200 g (for shoot tissue) or 0.100 g (for root tissue) of the homogenized sample was mixed with extraction reagent containing 20 mL 1% HNO 3 solution in a PTFE sealed vessel. The extraction program was as follows: 10 min to 90 °C and 90 min at 90 °C. After cooling down, the extracted solution was filtered and diluted in a 50-mL volumetric flask with ultrapure water. Three replicates were performed for each tissue in each group. For culture solution, 10 mL of each solution sample was filtered and ready for injection into the instrument. The As species in the final solutions were separated and quantified using HPLC (Agilent 1260, Tokyo, Japan) and ICP-MS [33]. The detailed operating parameters for HPLC and ICP-MS are listed in Table S3.

Multivariate Statistical Analysis
The significant differences between treatment groups were determined using one-way analysis of variance (ANOVA) followed by Tukey's test. The connections between two variables in the same rice tissue were measured using Pearson's correlation analysis. The ionomic variations between the rice tissues in different treatment groups were evaluated using principal component analysis (PCA). ANOVA and Pearson's correlation analysis were performed by SPSS v.13.0 and PCA was performed by R (3.4.4).

Growth of Rice Plant
The fresh length and biomass of rice roots and shoots in three treatment groups are demonstrated in Fig. 1. Compared with control group, length and biomass of rice shoots were significantly decreased by 21% and 35%, respectively when treated with As(III). However, no significant effect was observed on the rice roots between control and As(III) groups. Co-treatment with As(III) and strain affected the rice seedlings significantly compared with exposure to As(III) alone. The length and biomass of rice shoots in the As(III) + B group were 10% and 15% higher than those in the As(III) group, respectively. The values for rice roots were 8% and 37%, respectively.

As Speciation in Culture Solution and Plant Tissues
Variation of As species in the culture solutions collected after 3-day growth in As(III) and As(III) + B groups is shown in Fig. 2a. In all the collected culture solutions, only inorganic As forms were detected, while As(V) was predominant no matter with or without As(III)-oxidizing strain inoculation. Total As concentrations were decreased along with the growth of rice plants. However, the As(III) contents were maintained at a certain level (around 0.2 mg/ kg) after 6-days exposure. Distribution of As species in the rice shoots and roots from As(III) and As(III) + B treatment groups is illustrated in Fig. 2b. Similar to the culture solutions, only two inorganic As forms were found in the rice tissues. However, As(III) was predominant in the rice tissues different from the solutions. Total As concentrations in the rice shoots from As(III) + B group were significantly lower than those from As(III) group. In the rice roots, there was no significant difference of total As between two groups.

Oxidative Stress Indices in the Rice Tissues
The oxidative stress indices including inhibition ability of O 2 ·− and · OH, H 2 O 2 content, activity of SOD and CAT, and MDA content in the rice tissues from three treatment groups are demonstrated in Fig. 3. Effects of As(III) and strain on the oxidative stress indices in the roots were at variance with the shoots. Compared with controls, O 2 ·− inhibition ability in the shoots was promoted significantly when the rice plants were exposed to As(III). Negative values of O 2 ·− inhibition ability in Fig. 3g represented more residual O 2 ·− were detected after reacting with the root tissues than with the reference ascorbic acid. Inoculation with As(III)-oxidizing strain promoted the inhibition ability of O 2 ·− in the rice roots. In comparison with exposure to As(III) alone, inhibition ability of · OH in the shoots was improved while that in the roots was restricted when inoculated with As(III)-oxidizing strain. For H 2 O 2 content, significant difference was only found in the roots between A(III) + B group and other groups. The SOD activity in both shoot and root tissues was enhanced significantly when exposed to As(III). In contrast, CAT activity in the shoots from As(III) and As(III) + B groups weakened obviously compared with control group. There was no significant difference in MDA contents of the rice tissues between three treatment groups.

Ionomic Profile of the Rice Plants
Total element concentrations in the rice plants are summarized in Fig. 4. The concentrations of macro elements including Mg, P, K, and Ca in the rice shoots from As(III) group were higher than those from control group. Accumulation of the macro elements (Mg, P, K, and Ca) in the rice shoots from As(III) + B group were significantly reduced relative to the shoots from As(III) group. The same phenomena were also observed for some nutrient elements such as B, Mn, Co, Cu, Zn, and Mo in the rice shoots. With regard to the rice roots, the average concentrations of Mg, P, K, Ca, Mn, Co, Cu, and Zn decreased in the order of control group > As(III) group > As(III) + B group, except for B and Mo. Element Sb exhibited similar distribution patterns with As in the rice tissues. Fe in the rice shoots represented a descending order of control group > As(III) group > As(III) + B group, though there was no significant difference between Fe in the rice roots from different groups.
The ionomic diversity of the rice plants in different treatment groups were investigated through PCA based on element concentration matrix. As shown in Fig. 5a, the first two principal components (PCs) explained 74.7% of total variance. The rice samples were separated into two regions by PC1. The shoots clustered on positive axis with positive loading of B, Mg, P, K, Ca, and Mn, and the roots clustered on negative axis due to Al, V, Cr, Fe, Co, Ni, Cu, Zn, As, Sr, Mo, Cd, Sb, and Pb with negative loading on PC1. Figure 5b shows the variations of ionomes in the rice shoots from different treatment groups. The shoots were classified into three distinct clusters according to the first two PCs generated by PCA. Three clusters were corresponded to three treatment groups. Similar to the shoots, the PCA approach could also distinguish the rice roots from different treatment groups based on the ionomic matrix. As shown in Fig. 5c, the ionomic profiles of the rice roots from control group were clustered at negative PC2, while the profiles of the roots from As(III) + B group were clustered in the area of positive PC2 and negative Ionomic network and the connections between ionomes and oxidative activities in the rice samples were investigated using Pearson's correlation analysis. The detailed information on coefficient and significance is summarized in Table S4. Significant correlations between two oxidative stress indices, oxidative stress index and element, and As and other element were selected and the correlation networks generated by Cytoscape are shown in Figure S1. In the rice shoots, As was positively correlated with B, Mg, P, K, Ca, Mn, Co, Cu, Mo, and Sb. Whereas, only Sb was positively correlated with As in the rice roots. Constant positive relationships were observed between As and SOD in the rice plants. Except negative correlation with CAT in the roots, no significant correlation was found between As and other oxidative stress indices in the rice plants.

Discussion
Only inorganic As forms were detected in all the samples including rice tissues and culture solutions. Numerous studies have confirmed that rice seedlings could not convert inorganic As to methylated or any other organic As species [12,34]. Oxidation of As(III) was observed in the culture solutions without inoculation of the As(III)oxidizing strain. This can be attributed to the root exudates which played an important role on accelerating As(III) oxidation [13]. In previous study, As(III) was found to be stable in the culture solution without rice cultivation [13]. Dominant proportion of As(III) in the rice tissues can be attributed to high levels of As(V) reductases in rice seedlings [35]. After uptake, As(V) can be reduced to As(III) readily with the involvement of As(V) reductases Values represent means ± S.D. (n = 6). Different letters indicate significant differences (p < 0.05) according to one-way ANOVA for detoxification. Then As(III) in the rice roots would be sequestered by complexing with phytochelatins, translocated to the shoots or excluded into the culture solutions [17]. That means the constant As(III) concentrations (around 0.2 mg/kg) in the culture solutions as shown in Fig. 2a were mainly from the efflux of the rice roots.
Exposure to As(III) had more impacts on the rice shoots than roots. Biomass and length of the rice shoots were significantly reduced under As(III) stress alone compared with controls. Alleviation of As(III) stress was obvious when As(III) coexisted with As(III)-oxidizing strain SMS11. No matter biomass or length of both shoots and roots from As(III) + B group were promoted remarkably in comparison with those from As(III) group. Total As concentrations in the rice shoots were also affected significantly with the addition of bacteria. Considering comparable total As in the roots, much lower total As in the shoots from As(III) + B group than As(III) group indicated that bacterial inoculation could restrict As translocation from root to shoot. This might due to the formation of various As compounds induced by bacteria [36]. Moreover, total As in the culture solutions declined with the existence of bacteria after 6-days exposure. The reduced As in the culture solutions can also be explained by the formation of As compounds, which resulted in relatively higher total As in the rice roots from As(III) + B group than As(III) group as shown in Fig. 2b. In combination of published pathways [37,38], schematic overview of ROS produce and scavenging in the rice seedlings is plotted in Figure  S2. The oxidative stress indices increased and decreased significantly were marked as up-regulated and downregulated, respectively. SOD as a metalloenzyme is an ubiquitous enzymatic ROS antioxidants in all subcellular compartments [37]. As shown in Figure S2, SOD played a Fig. 5 Combination of score plot and loading plot generated from PCA for a all the rice tissue samples, b the rice shoots from three treatment groups, and c the rice roots from three treatment groups critical role on scavenging O 2 ·− induced by As(III) stress. Exposure to As(III) alone or coexisting with bacteria could constantly enhance SOD activity in the rice shoots and roots. Correlation analysis also obtained positive relationships between As and SOD in the rice tissues. Therefore, promoted O 2 ·− inhibition ability in the rice plants here was mainly attributed to the upregulation of SOD. The intracellular H 2 O 2 can be converted into H 2 O and O 2 regulated by CAT or reduced to · OH accompanied with Fenton reaction [38]. Weakened activity of CAT in the rice shoots from As(III) and As(III) + B groups elevated the risk of · OH formation. Among the ROS O 2 ·− , H 2 O 2, and · OH, · OH is the most reactive to attack all molecules in surroundings [39]. Inoculation with bacteria promoted the inhibition ability of · OH in the rice shoots, reducing the damage to DNA and proteins from · OH.
The ionomics of the rice plants revealed significant variations with tissues and environmental conditions. The difference between two tissues (shoot and root) was distinguished clearly by PCA based on the ionomic matrix, indicating that plant organ would be the crucial factor to drive the ionome [28]. Moreover, environmental condition was proved to make a substantial contribution to the ionomic balance in the rice plants. The rice tissue samples from different groups were separated into distinct clusters through the PCA approach. The results represented ionomic profile as a promising tool to identify rice plants under different external conditions. Uptake of As simultaneously elevated the levels of macro elements such as Mg, P, K, and Ca in the rice shoots as shown in Fig. 4. The ionomic network is mediated by the deeper genetic network [26]. Overexpression of the genes correlated with ion binding and transport in rice shoots under As(III) stress was the direct reason for accumulation of nutrient elements [27]. Upregulation of calcium binding and transport genes, phosphate transporter genes and potassium channel genes were observed in the rice shoots under As(III) stress [27,40], resulting in elevated concentrations of P, K, and Ca. Moreover, zinc transporter genes were upregulated in response to external As(III) stress [27]. This can explain higher Zn contents in the rice shoots under As(III) stress in comparison to the rice shoots under control condition. It is common knowledge that elemental species with similar structure and chemical property can be bound and transported by the same proteins [19,41]. Elevated Mg level in the rice shoots under As(III) stress could be attributed to its similar chemical properties to Ca and Zn. When co-cultivated with the As(III)-oxidizing strain, the concentrations of nutrient elements were declined in the rice shoots simultaneously with alleviation of As stress. The variation tendency of nutrient elements in the rice roots was contrary to shoots. The contents of most nutrient elements in the rice roots from both As(III) and As(III) + B groups were decreased compared to control group.

Conclusion
The toxicity of As(III) to the rice seedlings has been alleviated by the As(III)-oxidizing strain SMS11 efficiently. Rice shoot was considered as a more sensitive tissue to external conditions than root. Inoculation with the bacteria restricted As accumulation in the rice shoots under As(III) stress, consequently enhancing the growth of rice plants. Investigation of oxidative stress indices indicated that co-cultivated with the bacteria could promote the inhibition ability of · OH in the rice shoots, reducing the damage to DNA and proteins. The ionome research revealed that ionomic profiles of the rice plants were varied with tissues and environmental conditions, representing a promising tool to identify rice plants under different external conditions. Most nutrient elements in the rice shoots were increased with As increasing. The levels of most nutrient elements were declined with alleviation of As stress in the rice shoots from As(III) + B group. It implied co-cultivated with the As(III)-oxidizing bacteria could help rice plants with ionomic balance.