Citric acid and AMF Inoculation Combination Assisted phytoextraction of vanadium (V) by Medicago Sativa in V Mining Contaminated Soil

The use of citric acid (CA) chelator to facilitate metal bioavailability is a promising approach for phytoextraction of heavy metal contaminants. However, the role of CA chelator associated with arbuscular mycorrhizal fungi (AMF) inoculation on phytoextraction of vanadium (V) has not been studied. Therefore, in this study, a greenhouse pot experiment was conducted to evaluate the combined effect of CA chelator and AMF inoculation on plant growth and V phytoextraction in the V mining contaminated soil by Medicago sativa Linn. (M. sativa). The experiment was performed via CA (at 0, 5 and 10 mM kg − 1 soil levels) application alone or in combination with AMF inoculation. Plant biomass, root mycorrhizal colonization, P and V accumulation, antioxidant enzyme activity in plant, and soil chemical speciation of V were evaluated. Results depicted (1) a marked decline in plant biomass and root mycorrhizal colonization in 5- and 10-mM CA treatments which were accompanied by a signicant increased V accumulation in M. sativa tissues. The effects could be attributed to the enhancement of bioavailable V by mainly transferring from the reducible to acid-soluble V fraction. (2) The presence of CA signicantly enhanced P acquisition while the ratio of P/V concentration in plant shoots and roots decreased, owing to the increased V translocation from soil to plant. (3) In both CA treated soil, AMF symbiosis signicantly improved dry weight (31.4–73.3%) and P content (37.3-122.5%) in shoot and root of M. sativa, and showed markedly contribution in reduction of malondialdehyde (MDA) content (12.8– 16.2%) and higher antioxidants (SOD, POD and CAT) activities in the leaves, suggesting their combination could promote growth performance and stimulate antioxidant response alleviating V stress induced by CA chelator. (4) Taken together, 10 mM kg − 1 CA application and AMF inoculation combination exhibited higher amount of extracted V both in the shoot and root. Thus, citric acid-AMF-plant symbiosis provides a novel remediation strategy for in situ V phytoextraction by M. sativa in the contaminated soil.


Introduction
Vanadium (V) is a transition metallic element and widely distributed in the lithosphere (Hao et al., 2018).
Generally, V concentration in Earth's crust is 150 mg kg − 1 , but varied with soil types in the range of 2-310 mg kg − 1 (Rehder, 1991). As a valuable strategic resource, it is widely used in machinery manufacturing, aerospace, railways and other elds (Moskalyk and Alfantazi, 2003;Schlesinger et al., 2017). High concentration in soil and water environments can lead to detrimental effect on plant growth and all living organisms, although it in trace amounts is an essential for human beings and animals (Crans et al., 2004;. Thus, the United Nations Environment Programme (UNEP) has put vanadium on the priority list of environmental hazardous elements in 1980s (Hindy, 1990). In China, about 53% of vanadium minerals have been produced from varieties of vanadium-titanium magnetite mines and account for the current global output , but the mining and smelting activities have . Additionally, the characteristics of nutrient de ciency, extreme pH, and decreased microbial diversity, accompanied in V mining contaminated soil, consequently cause an impoverished habitat hindering plant establishment (Xiao et al., 2015). Hence, it is urgent to develop sustainably and economically e cient techniques for remediation of vanadium polluted sites.
Phytoextraction, as a subgroup of phytoremediation, can utilize speci c plants to enrich heavy metals in the aerial parts to remove metals in soil (Freitas et al., 2013). This technology can be used in large area of mine reclamation and heavy metal contaminated sites owing to its cost-effectiveness, no secondary damage and high e ciency (Sarwar et al., 2017). For a better performance, metal bioavailability in soil is a key factor in controlling the success of phytoextraction. Based on BCR (Community Bureau of Reference) fractionation, soil heavy metals can be categorized into different forms including acid-soluble, reducible, oxidizable and residual fraction (Hao et al., 2018). Among them, soil acid-soluble fraction represents the most mobile fraction that can be absorbed by plants. However, V element exist mainly in residual fraction in soil (Xiao et al., 2015;Hao et al., 2018), and the low metal bioavailability strongly weakens phytoextraction e ciency .
In recent years, many studies have focused on the effects of chelator on controlling the solubility of metals in soil such as EDTA (ethylenediaminetetraacetic acid) and some low molecular organic acids (LMWOA) including citric acid, oxalic acid or malic acid, those of which effectively enhance metal mobility and diffusion to root surface, thereby boosting phytoextraction (Blaylock et  . Importantly, citric acid (CA), as a natural chelating agent, has been reported a better substitute to synthetic chemical chelator for phytoextraction because of its low cost and effortless degradation without leaching metal-chelator compounds (Farid et al., 2017), and currently numerous studies associated with CA on phytoextraction of Cr (Farid et al., 2017), Cu (Zaheer et al., 2015), Cd (Sinhal et al., 2010), Pb (Shakoor et al., 2014) and As (Almaroai et al., 2012) for decontaminating polluted soils have been reported intensively, but less attention has been paid on V phytoextraction. However, higher concentration of CA and/or elevated bioavailable metals in phytoremediation also result in severe phytotoxicity symptoms such as plant growth inhibition (Turgut et al., 2004;, interfering with nutrients uptake (Cao et al., 2009) and inducing overgeneration of reactive oxygen species (ROS) (Imtiaz et al., 2015). Higher induced ROS production could further cause oxidative damage and retard plant growth by disturbing physiological and biochemical activities.
Arbuscular mycorrhizal fungi (AMF), which form mutualistic symbioses with most terrestrial plants, can improve plant tolerance against both biotic and abiotic stresses such as that from heavy metal (Dhawi et al., 2016). AMF symbiosis can alleviate oxidative damage by increasing the activity of antioxidants such as superoxide dismutase (SOD), peroxidase (POD) and catalase (CAT) to scavenge the ROS (Mollavali et al., 2016). Furthermore, some studies have reported the mechanisms of AMF symbiosis in mediating metal transportation and accumulation in soil-plant system for boosting metal tolerance (Chen et al., 2004). Thus, AMF-plant symbiosis has been treated as the potential candidate for phytoremediation (Miransari, 2011;Ma et al., 2019;Wang et al., 2019). In addition, AMF symbiosis take an active part in improved absorption of mineral nutrition (particularly P), and this improvement of plant P nutrition associated with AMF symbiosis has been generally regarded as plant tolerance mechanism against heavy metals' toxic effect . However, the role of AMF symbiosis in V accumulation and physiological responses against vanadium stress has not been studied intensively.  CA concentration was suitable for the release of chemically extractable V from the contaminated soil ( Fig. S1 in the Supporting Information).
The pot experiment adopted a 2×3 factorial with AMF inoculation and CA application in a completely

Plant and soil sampling
At destructive harvest, plant shoots and roots were separated and carefully washed. Before processing, a handful of fresh plant leaves and roots were collected and washed with deionized water for the determination of antioxidant enzyme activities and root mycorrhizal colonization, respectively. The loose soil attached to the individual plant root was shaken off, and collected and mixed uniformly to form a rhizosphere soil sample. Then, the remaining plant samples were rstly deactivated at 105℃ for 30 min, and then dried at 70℃ in the oven to the constant weight and weighed for plant biomass. Soil samples were sieved with a 2-mm mesh to remove coarse or ne root, and other visible impurities, and then airdried and used for measurement of soil chemical properties.

Measurement of mycorrhizal colonization, P and V concentration in plant
Mycorrhizal root colonization was determined by the method of Phillips and Hayman (1970). Brie y, the washed roots were su ciently softened with 10% KOH for 24 h, and then stained with 0.05% trypan (w/v) in lactic acid, glycerol and water solution (1:1:1 v/v/v) for 12 h after rinsing in water, and then decolorized with 50% acid glycerol solution. We randomly selected two groups of 15 root segments (1.5-cm-long) for individual plant root, and then deposited them on the slides for microscopy. The percentage of root mycorrhizal colonization (%) was calculated by the grid-line intersect method (Giovannetti and Mosse,1980).
For plant P and metal analysis, 0.2 g milled plant samples of shoot (stem and leaves) and root were respectively digested in a mixture of concentrated HNO 3 /HCl (6:2 v/v) on the microwave system (Mars5, CEM Corporation, USA). Then, the wet digestion samples were ltered and diluted with 1% HNO 3 , and analyzed for P and V in plant samples by ICP-MS. V accumulation amount (mg pot − 1 ) in plant was measured by multiplying V concentration in shoot or root organ (mg kg − 1 ) with the corresponding organ dry weight (kg pot − 1 ). Additionally, the bioaccumulation factor (BCF), translocation factor (TF) and metal extraction amount were introduced to evaluate phytoextraction e ciency of V in plant shoot and root. BCF was calculated as the ratio of metal concentration of plant shoot to the initial soil. TF was determined by the metal concentration of shoot divided by that of root.

Evaluation of physiological indices
Malondialdehyde (MDA) content in plant leaves was determined with minor modi ed method of thiobarbituric acid (TBA) reaction by Heath and Packer (1968). Brie y, 0.2 g of shoot sample was added with 2 mL of 10% trichloro acetic acid (TCA) to grind into tissue homogenate. The homogenization was centrifuged for 10 min at 12,000×g. Then, the assay solution was mixed by addition of centrifuged supernatant to an equal volume of 0.67% TBA. The resultant mixture was suddenly cooled after boiling in water bath for 30 min. After centrifugation for 10 min at 12,000×g, the obtained supernatant was determined spectrophotometrically followed by absorbance at 450, 532 and 600 nm wavelength, and ddH 2 O was used as the blank.
Antioxidant enzymes including superoxide dismutase (SOD, EC 1.15.1.1), peroxidase (POD, EC 1.11.1.7) and catalase (CAT, EC 1.11.1.6) in the leaf samples were evaluated in this study. Known weights of fresh leave samples were rstly frozen in liquid nitrogen, and put into a pre-cooled mortar with 2.0 mL of 50 mM phosphate buffer (pH 7.8). The mixture was grinded into homogenate on ice bath, and then centrifugated for 20 min at 4℃ and 12, 000×g, and the supernatant was collected in tube for antioxidant enzymes assays.
SOD activity was measured by Total Superoxide Dismutase (T-SOD) assay kit (Nanjing Jiancheng Bioengineering Institute, Nanjing, China) according to the manufacturer's instructions. One unit of SOD was de ned as the amount of SOD when inhibiting the reduction of SOD by 50% per gram fresh tissue in 1 mL reacted solution, as monitored at 550 nm with a UV-1600 spectrophotometer . POD activity was detected according to Maehly and Chance (1954). The assay mixture was comprised of 30 µL enzyme extract, 3 mL 0.2 M phosphate buffer (pH 6.0), 76 µL guaiacol and 0.112 mL H 2 O 2 (30%).
One unit of POD was de ned as the increase in absorbance by 0.01 per minute and per gram fresh tissue at 470 nm. CAT activity was determined by the method of Knörzer et al. (1996). The assay mixture contained 100 µL enzyme extract, 3 mL 0.15 M phosphate buffer (pH 7.0) and 0.3092 mL H 2 O 2 (30%).
One unit of CAT was de ned as the decrease in absorbance by 0.01 per minute and per gram fresh tissue, as monitored at 240 nm.

Determination of soil chemical speciation of vanadium and pH
Soil chemical speciation of vanadium (acid-soluble, reducible, oxidizable and residual) was analyzed according to the Community Bureau of Reference (BCR) sequential extraction procedure with some amendments (Hao et al., 2018). The detailed procedures of modi ed BCR sequential extraction were described in Table 2. In addition, V concentration of the residual fraction and total in soil samples were digested with an acid mixture (HNO 3 /HCl/HF; 6:3:1 v/v/v) through the microwave system. Digestion solution was transferred into polytetra uoroethylene (PTFE) tubes to evaporate nearly dry, and then dissolved and diluted to 50 mL with 1% HNO 3 . Both total and sequential extracted V ltrates were determined with ICP-MS. The certi ed standard reference materials (GBW 07453) obtained from the China National Center for Standard Reference Materials was measured for metal concentration by 86.3%-98.2%. Soil pH was measured in a 1:2.5 (m/v) ratio of soil: water by pH meter (Sartorius, PB-10).

Statistical analysis
All the data was presented as means of four replicates, and complied with normal distribution and homogeneity prior to conducting signi cance analyses. Two-way ANOVA was conducted to analyze the impact of CA application and AMF inoculation on each variable studied using SAS 8.02 software (SAS Institute, Cary, NC). Least Signi cant Difference (LSD) test was used to determine signi cant differences between treatments at the levels of 5% when ANOVA was signi cant. Pearson's correlation analysis was used to evaluate the relationship between soil pH and acid-soluble V, plant biomass and P content in shoot, plant biomass and P content in root, shoot biomass and root mycorrhizal colonization, root biomass and root mycorrhizal colonization. Graph plotting were performed with SigmaPlot version 14.0.

Mycorrhizal colonization and plant biomass
In this study, mycorrhizal colonization in M. sativa root ranged from 49.2 to 60.8% throughout all samples. These high colonization rates demonstrate the symbiosis between Funneliformis mosseae and host plant was successfully established (Fig. 1a). In all CA treatments, presence of AMF symbiosis brought by the successful colonization signi cantly promoted the dry weight of both plant shoot and root by 31.4-43.5% and 63.2-73.3%, respectively, when compared to associated sole CA application treatments. Though low colonization rates (0.83%) in plant root were observed in the control and 10 mM sole CA treatments, these low rates were probably due to the indigenous mycorrhizal fungi infection from the slightly incomplete sterilized soil.
However, addition of CA signi cantly decreased root mycorrhizal colonization rate as compared to those respective AMF-inoculated treatments without CA addition (Fig. 1b). Similarly, addition of CA treatment signi cantly reduced dry weight of plant shoot (17.0-27.4%) and root (26.2-37.1%), regardless of the existence of AMF inoculation. In addition, this decrease on plant dry weight was associated with increased CA concentration, suggesting CA chelator at concentration of 5-and 10-mM kg − 1 soil showed some inhibitory effect on plant growth.

Phosphorus concentration in samples
In Fig. 2a, it displays P concentrations in both shoot and root of M. sativa that were cultivated in the vanadium contaminated soil gradually increased with increasing CA addition levels. Compared with the control plants, P concentrations in shoot and root were signi cantly increased in both 5-and 10-mM CA treatment (43.78-67.9% and 18.9-48.9%, respectively) (Fig. 2a), suggesting CA application would facilitate P transportation and accumulation from soil to plant tissue. Furthermore, AMF inoculated plants presented higher concentration of P both in shoot and root as compared to respective treatments without AMF inoculation. This is possibly as a consequence of AMF promotion on P absorption in plant.
In root, the P concentrations were signi cantly in uenced by the interaction of AMF inoculation and CA application, according to the two-way ANOVA analysis (Table 3). In shoot, AMF inoculation gradually increased P content (mg per pot) in shoot with increasing levels of CA addition when plants were exposed to 5-and 10-mM CA relative to control plants (Fig. 2b). The highest P content in shoot was achieved in the group with combined treatment of 10 mM CA and AMF inoculation.  with AMF inoculation along with 5-and 10-mM CA application, as compared to the respective CA only treatment. This result explicitly presents that AMF symbiosis could suppress V absorption when the bioavailable V in soil was increased by CA chelator.
Concerning the P/V ratio in plant issues, both CA addition treatments (5 and 10 mM) signi cantly decreased the ratio in both shoot and root of M. sativa, as compared to non-CA treated plants (Fig. 4). When inoculated with AMF, P/V concentration ratio in shoot was enhanced by 29.7% and 47.1% under 5and 10-mM CA treatments, respectively, as compared to the respective CA-only treated plants (Fig. 4a).
Similarly, in plant root, AMF inoculation exhibited signi cant positive effect on the P/V concentration ratio regardless of CA concentrations (Table 3; Fig. 4b).

BCF, TF and V extraction amount by plant from soil
As shown in   their respective non-AMF inoculated plants (Fig. 5a). Three (SOD, POD and CAT) key antioxidant enzymes activities were signi cantly enhanced in CA application alone or in combination with AMF inoculation.
Regarding to the combined effects of AMF inoculation and CA application, AMF inoculation signi cantly increased SOD activity by 13.0-17.3% in the leaves when compared with the respective CA treated plants without AMF inoculation (Fig. 5b). POD activities in the leaves were signi cantly higher than the control treatment, either in combination of AMF inoculation and CA application or alone in V contaminated soil (Fig. 5c). AMF inoculation improved the activities of CAT with no signi cant differences at three (0, 5 and 10 mM kg − 1 ) levels of CA application when compared with their respective controls (Fig. 5d).
3.6 Soil chemical speciation of V and pH  17.3% and 54.3% in the original V-polluted soil, respectively. Generally, the combination of 10 mM CA application and AMF inoculation treatment tended to be more e cient for reducing oxidizable and residual V fractions. Two-way ANOVA analysis validated acid-soluble and residual V fractions to be signi cantly in uenced by their interaction (Table 3). In terms of V removal, M. sativa performed surprisingly well in the 10 mM CA application and AMF inoculation combined treatment, which total V concentration in M. sativa rhizosphere soil decreased by 6.8% relative to that of the initial soil V concentration (1705 mg kg − 1 ). Additionally, soil pH gradually decreased with increasing CA addition levels in vanadium contaminated soil. Either 10 mM CA treatment alone or in combination with AMF inoculation signi cantly decreased soil pH as compared to the other treatments.  As a natural chelating agent, citric acid plays an important role in controlling the phytoavailability of heavy metals in the soil (Turgut et al., 2004;Farid et al., 2017). In the presence of CA, elevated V bioavailability was re ected by observations of increasing V uptake in plant tissues and V morphology change in the soil, and was more obvious with increased CA concentration. Both 5 mM and 10 mM CA application were e cient in solubilizing V from the soil and inducing its uptake by M. sativa. This may be partly because that CA chelator contains negatively charged hydroxyl or carboxyl groups and can form stable chelating compounds with positively charged V, which was more conducive to V transportation and accumulation by M. sativa (Wang et al., 2019). Meanwhile, increased V accumulation signi cantly promoted the BCF and TF values of M. sativa at both levels of CA addition, which is bene cial to V phytoextraction from soil to plant tissues.
In addition, the present study showed a signi cant effect on V morphology resulting from a large amount of acid-soluble V releasing into the soil solution and the overall decrease of reducible V in both CA treated soil, which was in line with an enhancement of the ratio of acid-soluble/reducible V fraction (Table 5). Importantly, soil pH is a key factor affecting metal fraction and bioavailability in soil (

AMF inoculation promotes the growth of M. sativa against the toxicity of CA and V
Plant biomass is considered highly sensitive growth characteristics in response to environmental stresses (Imtiaz et al., 2015). In this study, CA application signi cantly decreased the dry weight of shoot and root in AMF inoculated or non-AMF inoculated plants (Fig. 1b), indicating CA addition to the V contaminated soil induced toxicity in M. sativa's growth. Generally, V concentration is in low concentration for most plant shoot and root , while in this present study plant V concentration was much higher than that of M. sativa reported by Yang et al. (2011) and Gan et al. (2020). The increased bioavailable V in both CA treated soil induced a signi cant enhancement of V concentration in shoot and root of M. sativa than that of in the non-CA treated plants (Fig. 3), accompanied with more MDA production with increasing CA addition, and consequently caused oxidative damage to M. sativa (Fig. 5a). A similar reduction of plant biomass was also observed in the study of Turgut et al (2004), which 1.0 or 3.0 g kg − 1 CA used in Cd, Cr and Ni polluted soil decreased total weight of Helianthus annuus when compared to the control soil without CA chelator. However, the effect of CA application on plant growth performance remains con icting results, and its promotion effect was also observed in some studies (Zaheer et al., 2015;Farid et al., 2017;Wang et al., 2019). This difference may be explained by the varieties of factors such as soil structure, the presence of toxic metal-chelator complexes and chelator dosage (Begum et al., 2012;Farid et al., 2017).
In the present study, combing plants with AMF inoculation could improve the growth of M. sativa under CA application, owing to AMF-plant symbiotic bene ts in P nutrient uptake and antioxidant defense system. AMF-plant symbiosis signi cantly improved P concentration both in shoot and root of M. sativa across three CA addition levels. The improved P acquisition may be explained by the contribution of extensive extraradical mycelium networks by AMF-plant symbiosis . There were signi cantly positive correlations between biomass and P content in shoot (R 2 = 0.35, P < 0.01, Fig. 6b) and in root (R 2 = 0.79, P < 0.01, Fig. 6c). Certainly, the improvement of P uptake associated with AMF inoculated plants was disturbed under CA chelator use, leading to the lower ratio of P/V concentration at both levels of CA addition (Fig. 4). Previous studies suggest that heavy metals show inhibition effects in nutrient absorption and metabolism in soil-plant system (Cao et al., 2009;Aihemaiti et al., 2019).
Interestingly, vanadate has a similar structure and charge to phosphate, and is absorbed into plant through P uptake system which results in competition in plant uptake and assimilation between them  (Fig. 5), which suggested that AMF symbiosis could enhance V stress resistance induced by CA addition and contribute to scavenge ROS and thus alleviate oxidative damage. Meanwhile, the lower MDA production in AMF inoculated plants also clearly demonstrated that, the adverse effect induced by oxidative damage was to some extent counterbalanced by AMF symbiosis.

Citric mediated AMF-plant symbiosis promotes plant V phytoextraction
Soil V contamination has become a serious environmental problem and requires sustainable and environmental-friendly strategies of remediation. In this present study, we rstly demonstrated that the combined CA application and AMF inoculation treatment had more pronounced promotion in extracted amount of V in M. sativa, accompanied by good growth performance and enhanced antioxidant enzyme activities against V stress (Table 4), suggesting this combination could be recommended for assisting V-phytoextraction in the studied site. In the combined plants, AMF symbiosis could counterbalance the negative effects on plant biomass induced by CA chelating agent, suggesting AMFplant symbiosis has a noteworthy potential in growth promotion against CA-induced phytotoxicity.
However, CA application also provoked inhibitory effects on mycorrhizal colonization of M. sativa root. This was probably due to high bioavailable V toxicity which inhibited the mycorrhizal colonization rate in the presence of CA chelator. It has been shown that mycorrhizal colonization could be negatively affected thus acidifying the rhizosphere soil (Dehghanian et al., 2018). In this study the decreased soil pH may be account for the increase of soil acid-soluble V fraction in the CA and AMF symbiosis combination soil.
However, plant shoot and root V concentration was lower in the combined treatment than that of in single CA application treatment (Fig. 3). The role of AMF symbiosis on metal uptake in contaminated soil-plant system has been systematically studied, with some studies reporting improvements, some reporting suppression and others indicating no discernible effect on metal uptake. These contradictory results may be ascribed to some factors such as AMF species, soil properties, plant species and metal type and concentration in the soil (Miransari, 2011 (Table 3). Although the combined plants had lower BCF, it showed notable potential of V translocation from root to shoot compared with the CA alone treatment. The enhancement of TF in combined plant might have resulted from the markedly decreased V concentration in the root, which may be a defense strategy adopted by AMF-plant symbiosis through the inhibition absorption of high concentration of metals in the soil (Kanwal et al., 2015). Moreover, higher V concentration in shoot and root of M. sativa showed no obvious toxic symptoms. This indicates that M. sativa is V-tolerant plant, indicative of strong potential to assist V phytoextraction associated with the advantage of successive cultivation without repeated planting at each new cycle.

Conclusions
In conclusion, CA application could improve the bioavailability of both V and P in the soil and their transportation and accumulation in soil-plant system, but at the same time, CA could induce phytotoxicity to plant growth, AMF-plant symbiotic relationship as well as the decreased P/V concentration ratio. In our pot-culture system, CA application combined with AMF inoculation could improve M. sativa-mediated V mobility, enhance plant biomass and P accumulation, and alleviate V toxicity to plants by stimulate antioxidant enzyme activity. These effects of the combined CA application and AMF inoculation enhance the extracted amount of V from soil. Specially, in this study, 10 mM CA application combined with AMF inoculation bene ted most the V phytoextraction by M. sativa grown in V contaminated soil. Therefore, this study provides a new understanding of citric acid-AMF-plant symbiosis in phytoextraction of V as well as the competitive absorption between P and V. The ndings also shed light on the potential application of citric acid-AMF-plant symbiosis associated with suitable plants in in situ phytoextraction of V or other heavy metal(loid)s.

Data availability
All data related to this publication are made available from the corresponding author on reasonable request.