Phytoremediation of uranium-contaminated soil by perennial ryegrass (Lolium perenne L.) enhanced with citric acid application

Perennial ryegrass (Lolium perenne L.) was planted in uranium-contaminated soil mixtures supplemented with different amounts of citric acid to investigate the defense strategies of perennial ryegrass against U and the enhanced mechanism of citric acid on the remediation efficiency in the laboratory. The uranium content in the plant tissues showed that the roots were the predominant tissue for uranium accumulation. In both root and shoot cells, the majority of U was located in the cell wall fraction. Furthermore, antioxidant enzymes were also stimulated when exposed to U stress. These results suggested that perennial ryegrass had evolved defense strategies, such as U sequestration in root tissue, compartmentalization in the cell wall, and antioxidant enzyme systems, to minimize uranium stress. For an enhanced mechanism, the optimal concentration of citric acid was 5 mmol/kg, and the removal efficiency of U in the shoots and roots increased by 47.37% and 30.10%, respectively. The treatment with 5 mmol/kg citric acid had the highest contents of photosynthetic pigment and soluble protein, the highest activity of antioxidant enzymes, and the lowest content of MDA (malondialdehyde) and relative electrical conductivity. Moreover, the TEM (transmission electron microscope) results revealed that after 5 mmol/kg citric acid was added, the cell structure of plant branches partially returned to normal, the number of mitochondria increased, chloroplast surfaces seemed normal, and the cell wall became visible. The damage to the cell ultrastructure of perennial ryegrass was significantly alleviated by treatment with 5 mmol/kg citric acid. All the results above indicated that perennial ryegrass could accumulate uranium with elevated uranium tolerance and enrichment ability with 5 mmol/kg citric acid.


Introduction
Uranium pollution in soil commonly results from uranium mining and milling activities in the nuclear industry (Sha et al. 2019). Uranium is regarded as radioactive and toxic, which poses a serious threat to ecosystems and human health (Ye et al. 2020). How to remediate large-area and low-concentration uranium-contaminated soil effectively has become a research hotspot. Physical methods such as soil replacement and electric remediation are often costly to treat uranium-contaminated soils, which seriously affect the soil environment (Shi et al. 2018), while chemical methods such as soil washing and solidification often have adverse effects on soil biological activity and soil fertility (Singh and Prasad 2011). Low cost, environmental friendliness, and widespread methods have attracted wide interest.
Pytoremediation, which uses the fixation and extraction functions of plants, can effectively remove uranium from soil and has several advantages, such as environmental friendliness, easy implementation, and cost-effectiveness (Burges et al. 2017;Hu et al. 2019;Li et al. 2019). Over the last decade, some studies have demonstrated that some plants can remediate uranium-polluted soil with high uptake capacity, such as Indian mustard (Brassica juncea (L.) Czern.) and sunflower (Helianthus annuus L.) (Laurette et al. 2012;Qi et al. 2014). However, most of these species are not suitable for commercial phytoremediation mainly because of their low annual harvestable biomass and low growth rate (Nascimento et al. 2020).
Perennial ryegrass (Lolium perenne L.) is a grass from the family Poaceae, which is widely cultivated in China as forage grass with the advantages of rapid growth, ease of management, multiple cutting times each year, and strong regeneration ability (Bin et al. 2019, Grachet andWalker 2016). Recently, the restorative effect of perennial ryegrass on uranium-and other heavy metal-contaminated soils has been investigated (Qi et al. 2019;Zhao et al. 2018). However, the defense strategies of perennial ryegrass for U have not been systematically stated and further investigation is still required.
The efficiency of metal uptake by plants is associated with the bioavailability of metals by root uptake (Newete et al. 2016). Chelating agents such as ethylenediaminetetraacetic acid (EDTA) can produce a water-soluble metal-chelating agent and can alter the existing form of heavy metals in soil, which increases the bioavailability for plant root uptake and translocation to the leaves of the plant (Gunawardana et al. 2011). However, the addition of EDTA can increase the risk of leaching heavy metal pollutants from soil into groundwater due to its low biodegradability (Chen et al. 2020b). Currently, biodegradable chelating agents with no adverse effects on the environment are favorable. Citric acid, as a biodegradable and environmentally friendly chelating agent, can significantly improve plant remediation efficiency and enhance plant resistance to heavy metals at appropriate concentrations (Duquène et al. 2008;Monroy-Figueroa et al. 2015). Studies have shown that citric acid could enhance the Ni concentrations in roots, stems, and leaves up to 138.2%, 54.2%, and 38%, respectively, compared with Ni-only treated plants (Khair et al. 2020). To the best of our knowledge, it is not clear whether citrate can improve uranium enrichment in perennial ryegrass (Liu et al. 2018). Research on the specific mechanisms with an appropriate amount of citric acid is still required to elevate U remediation efficiency.
The objective of this study was to investigate the defense strategies of perennial ryegrass against uranium and the effects of different concentrations of citric acid on the remediation efficiency of perennial ryegrass in uraniumcontaminated soils. Antioxidant enzymes, including superoxide dismutase (SOD), catalase (CAT), peroxidase (POD), and glutathione reductase (GR) in perennial ryegrass may be expressed in response to various environmental stresses and may play important roles in protecting organelles and minimizing tissue injury (Garg and Kaur 2013). The biochemical and physiological indices of perennial ryegrass and the activities of 4 types of antioxidant enzymes in perennial ryegrass were measured. The uranium enrichment ability of perennial ryegrass was determined. Transmission electron microscopy (TEM) was used to observe the changes in cell ultrastructure during the phytoremediation process to understand the mechanism of perennial ryegrass responding to the increased uranium accumulation caused by the addition of citric acid to soil.

Materials
The seeds of perennial ryegrass (Lolium perenne L.) were provided by the specimen laboratory of Northwest Agriculture and Forestry University (Yangling, China). Full and uniform seeds without damage by moths were used in this study.

Experimental design
Natural soil was collected from a horizon 0-20 cm depth from the university campus. The natural soil samples were air-dried, ground, and passed through a 2-mm nylon sieve. The soil mixture was prepared with an equal volume of river sand, vermiculite, and natural soil, which could avoid anaerobic conditions during plant cultivation. The background uranium content in the natural soil sample was negligible. The basic properties of the natural soil were as follows: organic matter content was 13.01 g/kg, available phosphorus was 40.24 mg/kg, available potassium was 66.24 mg/ kg, available alkali hydrolysable nitrogen was 221.25 g/kg, cation exchange capacity was 12.37 cmol/kg, clay 51.5%, silt was 24.2%, and sand was 24.3%.
Flowerpots (150 × 190 × 160 mm) containing 1.0 kg of the soil mixture were used. The uranyl nitrate solution was evenly sprayed onto the soil mixture to obtain a uranium-soil mixture with a concentration of 5 mg/kg. In addition, the soil mixtures were incubated for 30 days, and the moisture of the soil was kept at 65% during this period by adding distilled water to compensate for the lost water due to evaporation. Then, 30 perennial ryegrass seeds were planted in each flowerpot; 25 plants with uniform growth vigor were selected 1 week after seed germination. These plants were watered 1-2 times per day to keep the soil moisture at 65% . Modified Hoagland's nutrient solution (see Supplementary Material 1), a standard medium made from distilled water enriched with specified nutrients for plant growth studies, was added once a week.
Four series of experiments were performed to measure the effects of different citric acid concentrations on the efficiency and physiological characteristics of perennial ryegrass under uranium-induced stress: Con + 10 (10 mmol/ kg citric acid), Con + 5 (5 mmol/kg citric acid), Con + 1 (1 mmol/kg citric acid), and Con + 0 (0 mmol/kg citric acid). The solution pH of citric acid at different concentrations was adjusted to 6.5 with NaOH and HCl solutions and added to the experimental treatments a week before harvest. Equal volumes of distilled water (pH 6.5) were added to the control group (Con + 0). In sum, 12 planted pots which included a control and 3 treatments with three replicates were used in this experiment (visual inspection of plant cultivation in Supplementary Material 2).
After 60 days, the plants were harvested and separated into shoots (aboveground parts) and roots. The shoots were first washed with tap water and subsequently with distilled water twice. The roots were washed with abundant tap water until they were free of soil particles and subsequently washed with distilled water three times Nezami et al. 2016). After natural drying, the stems and roots were placed in a drying oven and dried at a constant temperature of 80 °C for 12 h. Then, the plant samples were ground and filtered through an 80-mesh sieve, and stored in small sample bags for later use.
The related physiological and biochemical indices and uranium enrichment in the shoots and roots of perennial ryegrass were investigated in the following experiments.

Analytical methods
The contents of photosynthetic pigments, soluble proteins, and MDA were determined by using the ethanol extraction method, Coomassie brilliant blue method, and TBA method, respectively (Chaturvedi et al. 2015). Enzyme activities were measured as previously reported (Dixon et al. 2009).
The cellular and subcellular distributions of uranium in plant epidermal cells were observed by utilizing the 5-Br-PADAP method using a Hitachi H-7650 transmission electron microscope (TEM). Briefly, the plant tissues were fixed with 4% glutaraldehyde, and treated with acetone dehydration (30%, 50%, 70%, 80%, 90%, and 100%). The samples were embedded in Epon 812 and cut into 70-nm-thick slices by a Leica microtome (Leica Microsystems, Co., Ltd., Germany). ICP-MS (Agilent 150 Technologies 7500 Series, Santa Clara, USA) was used to detect the uranium content in perennial ryegrass. The detection limit was 0.03 ng L −1 for U.

Calculation formula
The translocation factors (TFs) were calculated using Eq. (1), where C root is the uranium content in the root (mg kg −1 in dry weight), and C shoot is the uranium content in the shoot (mg kg −1 in dry weight) (Al Mahmud et al. 2018;Chen et al. 2020c).

Data analysis
All experimental data (mean ± SD) were analyzed with GraphPad Prism for Windows version 5.0 (GraphPad Software, La Jolla, USA). The single factor analysis of variance was performed. Values of p < 0.05 were considered significant.

Biomass of perennial ryegrass and uranium enrichment with different concentrations of citric acid
The repair efficiencies of uranium contaminated soil with perennial ryegrass were positively correlated with the biomass and uranium enrichment in the plants. The uranium concentrations in the roots of all treatments were higher than those in shoots, which demonstrated that the translocation factor of uranium from roots to shoots was less than 0.165 (Table 1). Recently, Li et al. reviewed uranium sources, speciation, uptake, toxicity, and bioremediation strategies in soil-plant systems and indicated that root cell walls could provide negatively charged sites to bind UO 2 2+ ions penetrating into the root parts, restricting the translocation of uranium to the aerial parts. Moreover, phosphate groups on the root cell walls were the main coordination sites for U complexation, precipitation, and mineralization, which were responsible for alleviating uranium phytotoxicity (Chen et al. 2021). Therefore, uranium was usually insoluble in water and remained in the roots of most plants, and its transfer from roots to stems was usually limited (Nezami et al. 2016;Qi et al. 2019).
The biomass and uranium enrichment in the shoots and roots of perennial ryegrass increased when citric acid was not more than 5 mg/kg (Table 1). The highest values of biomass in the shoots and roots of perennial ryegrass were detected in the Con + 5 treatment, which increased by 17.41% (P < 0.05) and 12.21% (P < 0.05), respectively, compared with those in the control group (Con + 0). However, some researchers have pointed out that the addition of certain chelating agents to soil can increase the concentration of heavy metal ions in soil solution, thus inhibiting plant growth and reducing biomass (Begum et al. 2012;Hseu et al. 2013;Xin et al. 2009). The addition of organic acids in the phytoremediation process promoted the uptake of heavy metals and biomass production in other studies (Han et al. 2016;Najeeb et al. 2009;Wang et al. 2019). The effects of chelating agents on heavy metal uptake and biomass production might be related to the adaptation tolerance of the plants, as well as the concentration and types of chelating agents.
In addition, the shoots and roots of the Con + 5 treatment had the highest uranium concentrations, which increased by 47.37% (P < 0.05) and 30.10% (P < 0.05), respectively. As a result, the transfer coefficient significantly increased in the Con + 5 treatment (p < 0.05). The reason for this may have been that the application of citric acid influenced the sorption of uranium by the soil mixture, and enhanced the mobility and bioavailability of uranium, thus increasing the capability of plants to transfer U from roots to shoots Li et al. 2014). The introduction of citric acid and oxalic acid in the phytoremediation process increased the uptake of 226 Ra by 1.5 times compared with the control with corn (Zea mays L.) (Nezami et al. 2016). It was also confirmed that citric acid could promote the absorption of 241 Am by Barley (Hordeum vulgare L.) roots and its transport in the plants (Wang et al. 2016). However, a higher citric acid concentration (10 mmol/kg) would decrease the accumulation of uranium, although citric acid could substantially enhance the bioavailability of uranium by improving the solubilization of soil-bound uranium (Chen et al. 2020c). Thus, the toxic effects of citric acid may damage the physiological structure of perennial ryegrass and cause a decrease in plant biomass when the citric acid concentration reached 10 mmol/kg (Duquène et al. 2008;Monroy-Figueroa et al. 2015).
These results suggested that 5 mmol/kg citric acid was the most effective concentration of citric acid to increase biomass production (in both shoots and roots) and uranium enrichment in perennial ryegrass. To illustrate the enhanced mechanism of the citric acid-assisted uptake of uranium in perennial ryegrass, the effects of different concentrations of citric acid on the subcellular distribution of uranium, physiological characteristics of perennial ryegrass, activities of antioxidant system enzymes, and cellular ultrastructure under uranium stress will be analyzed in the following sections.

Subcellular distribution of uranium in perennial ryegrass with different concentrations of citric acid
Understanding the subcellular distribution of heavy metals in organisms is fundamental to understanding the process of heavy-metal uptake, storage, and detoxification (Nie et al. 2015). The distributions of uranium in the cell wall fraction, organelle fraction, and cytosolic fraction in the shoots and roots of perennial ryegrass were further investigated. As shown in Table 2, low concentrations (1 mmol/kg and 5 mmol/kg) of citric acid could promote the enrichment of uranium in the subcellular structure of perennial ryegrass, and 5 mmol/kg citric acid was the best concentration. In the Con + 5 treatment, the uranium contents in the cell wall, organelle fraction, and cytosolic fraction increased by 51.68%, 36.25%, and 42.24% in the shoots and by 32.96%, 18.36%, and 34.24% in the roots, respectively. In contrast, the accumulation of uranium decreased in both roots and shoots in the Con + 10 treatment. Although citric acid has been favored by researchers due to its high biodegradability and its ability to stimulate metal uptake in plants ( Moslehi et al. 2019), citric acid also has adverse effects on plant growth and development, as it can cause etiolation, withering, and even death when present in excess amounts (Hasan et al. 2019). The biomass changes in perennial ryegrass with different concentrations of citric acid also indicated that the growth of perennial ryegrass was hindered gradually when the citric acid concentration exceeded 5 mmol/kg (Table 1). These results further proved that 5 mmol/kg citric acid was the optimal concentration to enhance the enrichment of perennial ryegrass. The subcellular partitioning of uranium in plants reflects its internal processes during uranium accumulation, which can provide more mechanistic information about uranium tolerance and the interaction process between uranium and perennial ryegrass (Nie et al. 2014). The distribution proportion of uranium in different parts of the same cell in roots was ordered as cell wall > organelle > cytosol, while the order in shoots was cell wall > cytosol > organelle (Table 2). These results demonstrated that a greater proportion of uranium was stored in the cell wall, more than 60% in both roots and stems; therefore, the cell wall plays an important role in uranium tolerance. This trend was consistent with the results proposed by other researchers (Nie et al. 2015). It was reported that plant cell walls were principally composed of polysaccharides and proteins that would provide negatively charged sites for uranium binding, and the transport of uranium into the cytosolic part would be hindered as a result (Chen et al. 2021;Huang et al. 2017). The phosphate groups on the cell walls would complex or precipitate with uranium ions, which further limited the transport of uranium from cell walls to the inner parts of the cells (Baker et al. 2019). Meanwhile, the storage capacity of uranium in the cell wall and the barrier-protecting effect of cytosol improved when low concentrations (1 mmol/kg and 5 mmol/kg) of citric acid were added. Consequently, all distribution proportions of uranium in the cell wall were higher than those in the control group in both roots and shoots, and all values in organelles were lower than those in the control group. However, the proportions of uranium in the cytosol of the shoots contrasted with those in the roots when low concentrations of citric acid were added, which might be closely related to the buffer capacity of the cytosol (phosphate in the cytosol bioprecipitated with uranium) and uranium concentrations in the cell wall of roots and shoots (Pan et al. 2015). A high concentration (10 mmol/kg) of citric acid decreased the accumulation of uranium in roots and shoots, and the distribution trends of uranium in different parts were consistent with those in the 5 mmol/kg citric acid treatment.

Physiological characteristics of perennial ryegrass with different concentrations of citric acid
Photosynthetic pigment content, which is a direct indicator of plant photosynthesis, can be used as a tolerance criterion for heavy metals in plants (Sebastian and Prasad 2018). As shown in Fig. 1A, 5 mmol/kg citric acid enhanced the photosynthesis of perennial ryegrass, which indicates an elevated tolerance to uranium. This phenomenon was also reported in a previous study (Chen et al. 2020c). However, the enhanced effects on the uranium tolerance of perennial ryegrass with 1 mmol/kg or 10 mmol/kg citric acid were not significant compared with the 5 mmol/kg citric acid treatment. Compared with the control group (Con + 0), the levels of chlorophyll-a (chl-a), chlorophyll-b (chl-b), and carotenoids in the 5-mmol/kg treatment (Con + 5) increased by 28.29%, 44.16%, and 28.99%, respectively. In contrast, the levels of chl-a, chl-b, and carotenoids did not significantly change in the Con + 1 and Con + 10 treatments (p > 0.05). The data indicated that the chl-b content in perennial ryegrass was significantly correlated with the citric acid concentration (p < 0.05). However, the chl-a and carotenoid levels were not significantly correlated with the citric acid concentration (p > 0.05). Further research is required to elucidate the enhanced mechanisms of the associations among the pigment contents, citric acid concentrations, and uranium tolerance of perennial ryegrass.
The permeability of the cell membrane reportedly increased when it was exposed to uranium, which caused the leakage of intracellular electrolytes and increased the relative electric conductivity (REC) (Dai et al. 2017). Simultaneously, lipid peroxidation of the cell membrane produced MDA, which reacted with proteins and nucleic acids and Table 2 Effect of citric acid on the uranium subcellular distribution of perennial ryegrass Values are given as the mean ± SD, n = 3. *P < 0.05: significant difference compared with the control (Con + 0) group.

Treatments
Uranium content (mg/kg) and ratio (%) in different sbucellulars affected cell function (Chen et al. 2020a;Khair et al. 2020). Therefore, relative conductivity, MDA content, and soluble protein content are the main indices of cell membrane permeability, and are closely related to U tolerance. As shown in Fig. 1B, C, and D, the root conductivity and MDA content were generally higher than those in the leaves. Thus, the degree of damage to the root cells in perennial ryegrass was more severe than that to the leaf cells. Moreover, the Con + 5 treatment had the lowest relative electrical conductivity and MDA. The mean electrical conductivity values of the shoots and roots in the Con + 5 treatment decreased by 20.19% (p < 0.05) and 20.26% (p < 0.05), respectively, compared with the control group (Fig. 1B). Similarly, the mean MDA values of the shoots and roots in the Con + 5 treatment decreased by 22.16% (p < 0.05) and 23.63% (p < 0.05), respectively, compared with the control group (Fig. 1C). Thus, citric acid (5 mmol/kg) can significantly decrease the electrical conductivity and MDA in the shoots and roots of plants, which are negatively correlated with uranium tolerance . Furthermore, the contents of soluble proteins in plants of all treatments were investigated. As illustrated in Fig. 1D, the contents of soluble proteins in the plant shoots of the Con + 1, Con + 5, and Con + 10 treatments were 1.31-, 1.90-, and 1.49-fold higher, respectively, than those in the control group. The contents of soluble proteins in the plant roots of the Con + 1, Con + 5, and Con + 10 treatments were 1.03-, 1.39-, and 1.05-fold higher, respectively, than those in the control group. All contents of soluble protein in the shoots and roots of plants in the Con + 5 treatment were higher than those in the Con + 1 and Con + 10 treatments, which suggested that citric acid (5 mmol/kg) could significantly increase the contents of soluble proteins in the shoots and roots of plants (P < 0.05). It has been proven that uranium stress damages the cell membrane systems, causing leakage of soluble substances in cells and disrupting metabolism, which results in metabolic disorders of the cells and decreases protein synthesis ). The addition of citric acid at an appropriate amount might enhance nutrient uptake by plants (Najeeb et al. 2011) or the synthesis of phytochelation (PCs) in plants (Muhammad et al. 2009). PCs are the most vital type of metal chelators and have a positive effect on U and chelation of other metals, thereby alleviating the adverse effects of U stress in plants (Yu et al. 2021). Therefore, the protein content in the plants increased in all treatment groups. However, citric acid also has adverse effects, such as etiolation, withering, or even death on plant growth and development when present in excess amounts (Hasan et al. 2019). As a result, citric acid alleviated uranium stress, and the protein content in the Con + 10 treatment decreased compared with that in the Con + 5 treatment.
All of these results revealed that 5 mmol/kg citric acid could alleviate the cell damage of perennial ryegrass exposed to uranium stress. Fig. 1 Measurements of the physiological and biochemical indices of perennial ryegrass. A Photosynthetic pigment content (mg/g, FW) in perennial ryegrass with added citric acid. B Relative conductivity (%) in perennial ryegrass with added citric acid. C MDA content (mg/g, FW) in perennial ryegrass with added citric acid. D Soluble protein content (mg/g, FW) in perennial ryegrass with added citric acid. FW, plant fresh weight

Effects of different citric acid concentrations on the activities of antioxidant system enzymes
As shown in Fig. 2, all activities of POD ( Fig. 2A), SOD (Fig. 2B), CAT (Fig. 2C), and GR (Fig. 2D) in the shoots and roots increased with different concentrations of citric acid in the Con + 1, Con + 5, and Con + 10 treatments compared with those in the control group. However, only the values in the Con + 5 treatment were significantly affected compared with those in the control group (P < 0.05). Therefore, 5 mmol/kg citric acid had the best enhancement effect on the activities of four antioxidant enzymes in perennial ryegrass.
When subjected to uranium stress, plant cells may produce H 2 O 2 to reduce the fixation efficiency of CO 2 in cells, while H 2 O 2 (Haber-Weiss) combines with superoxide anion (O 2 − ) to generate reactive oxygen species (ROS), which is harmful to the plants. In (Geebelen et al. 2002;Smeets et al. 2005). GR catalyzes the transformation of oxidized glutathione into glutathione to resist the generation of ROS in combination with SOD (Slomka et al. 2008). It is obvious that adding citric acid could increase the activities of antioxidant enzymes compared with the control group, which could contribute to reactive oxygen species (ROS) neutralization and enhance plant adaptation to uranium-contaminated soil environments . Interestingly, the activities of antioxidant enzymes in perennial ryegrass in the Con + 10 treatment decreased compared with those in the Con + 5 treatment. Antioxidant activity showed dynamic behavior; that is, it increased under mild and/or moderate heavy metal or chelate stress and decreased under higher metal or chelate stress.
Our results indicated that all changes in the activities of POD, SOD, CAT, and GR in the shoots and roots of perennial ryegrass were significantly correlated with citric acid concentrations (P < 0.05), which suggested that the addition of citric acid can enhance the antioxidant enzyme activity of perennial ryegrass. The results in this study were consistent with those in some previous reports (Afshan et al. 2015;Khair et al. 2020;Slomka et al. 2008).

Cellular ultrastructure with different citric acid concentrations
According to the above results, the cellular ultrastructure was affected by the con + 5 treatment. As shown in Fig. 3A, the cellular ultrastructure of shoots in perennial ryegrass was unchanged with normal mitochondria and an evenly distributed matrix in the control group. Chloroplasts were obviously observed and normally surrounded Fig. 2 Detection of the activities of antioxidant system enzymes in perennial ryegrass. A The activity (Ug −1 min −1 ) of POD in perennial ryegrass with citric acid. B The activity (Ug −1 ) of SOD in perennial ryegrass with citric acid. C The activity (Ug −1 min −1 ) of CAT in perennial ryegrass with citric acid. D The activity (Ug −1 min −1 ) of GR in perennial ryegrass with citric acid by a double membrane. In contrast, the cell structure of plant shoots with 5 mg/kg uranium was changed with a significantly reduced number of mitochondria, expanded chloroplasts, damaged cell walls, and disrupted double chloroplast membranes (Fig. 3B). In the treatment with 5 mmol/kg citric acid (Fig. 3C), the cell structure of plant shoots partially returned to normal. Interestingly, in plant shoot cells, the mitochondrial number increased, chloroplasts were seemingly normal, and the cell wall was observable. The effects of uranium and citric acid on the cell structure in the plant roots were identical to those in the plant shoots. In the treatment without uranium, the cell structure of the plant roots was normal (Fig. 3D). The cell structure was partly disrupted by treatment with 5 mg/kg uranium (Fig. 3E). The effect of uranium on the cell ultrastructure was partially alleviated when 5 mmol/ kg citric acid was added (Fig. 3F). Uranium stress often stimulates the biosynthesis of free radicals and ROS, which disrupt cell structures and organelles. ROS also cause ionic imbalance and protein injury and thus lead to oxidative damage in plants. Citric acid application at an appropriate concentration (5 mmol/kg) neutralizes ROS and plays a promising role in alleviating the phytotoxic effects in plants caused by uranium. If the concentration of citric acid is not enough (1 mmol/kg), the activities of key antioxidants are not stimulated sufficiently, and the detoxification effect is not obvious. When the concentration of citric acid is too high (10 mmol/kg), the oxidative stress induced by chelates dominates and damages the ultrastructure of the plants.

Conclusions
This study showed that perennial ryegrass has evolved several defense strategies to minimize uranium-induced oxidative stress, such as U sequestration in root tissue, compartmentalization in the cell wall, and antioxidant enzyme systems. Perennial ryegrass could strongly enrich uranium in plants, and 5 mmol/kg citric acid could significantly increase the tolerance for U and enhance the enrichment and transportation of uranium in perennial ryegrass. The activities of antioxidant enzymes (POD, CAT, SOD, and GR) were upregulated with 5 mmol/kg citric acid. The relative electrical conductivity and MDA content of perennial ryegrass decreased, and the contents of photosynthetic pigment and soluble protein increased with 5 mmol/kg citric acid. TEM images of the ultrastructure in root and shoot cells of perennial ryegrass proved that 5 mmol/kg citric acid could lighten the degree of damage under uranium stress, which is a promising bioremediation strategy for uranium-contaminated soil.
Author contribution All authors contributed to the study conception and design. Method implementation and optimization, analysis, and data evaluation were performed by Lishan Rong and Shiqi Zhang. The first draft of the manuscript was written by Shiyou Li and Guohua Wang. Shuibo Xie, Jiali Wang, and Guohua Wang commented on previous versions of the manuscript. All authors read and approved the final submitted manuscript.
Funding This work was financially supported by the National Natural Science Foundation of China (11475080, 51904155), Education Department Fund of Hunan Province of China (19C1588), and Hengyang's Science and Technology Planning Projects (2018KJ130). national natural science foundation of china,11475080,Shuibo Xie,51904155,Guohua Wang,education department of hunan province,19C1588,Lishan Rong,hengyang's science and technology planning projects,2018KJ130,Guohua Wang Data availability The datasets generated and/or analyzed during the current study are property of Lishan Rong (University of South China, China); they are available from the corresponding author who will inform Lishan Rong that the data will be released on reasonable request.

Declarations
Ethics approval Not applicable.

Consent to participate Not applicable.
Consent for publication Not applicable.