Effect of Novosphingobium sp. CuT1 inoculation on the rhizoremediation of heavy metal- and diesel-contaminated soil planted with tall fescue

Rhizoremediation is a promising method based on the synergism between plant and rhizobacteria to remediate soil co-contaminated with heavy metals and total petroleum hydrocarbons (TPHs). A plant growth–promoting (PGP) rhizobacterium with diesel-degrading capacity and heavy metal tolerance was isolated from the rhizosphere of tall fescue (Festuca arundinacea L.), after which the effects of its inoculation on rhizoremediation performance were evaluated in heavy metal– and diesel-contaminated soil planted with tall fescue. The bacterial isolate (Novosphingobium sp. CuT1) was characterized by its indole-3-acetic acid (IAA) production, 1-aminocyclopropane-1-carboxylic acid (ACC) deaminase activity, and siderophore productivity as PGP traits. CuT1 was able to grow on 1/10 LB-agar plates containing 5 mM of Cu or 5 mM of Pb. To evaluate the remediation effect of heavy metal– and diesel-contaminated soil by CuT1 inoculation, the experimental conditions were prepared as follows. The soil was artificially contaminated with heavy metals (Cu and Pb) at a final concentration of 500 ppm. The soil was then further contaminated with diesel at final concentrations of 0, 10,000, and 30,000 ppm. Finally, all plots were planted with tall fescue, a representative hyperaccumulating plant. Compared to the rhizoremediation performance of the co-contaminated soil (Cu + Pb + diesel) without inoculation, the bioavailable Cu concentrations in the soil and the tall fescue biomass were significantly increased in CuT1 inoculation. Additionally, the root growth of tall fescue was also promoted in CuT1 inoculation. Correlation analysis showed that Cu bioavailability and bioconcentration factor were positively correlated with CuT1 inoculation. The diesel removal efficiency showed a positive correlation with CuT1 inoculation, although the diesel removal was below 30%. CuT1 inoculation was positively correlated with IAA and dehydrogenase activity in the soil. Moreover, the dry biomass of the tall fescue’s roots was highly associated with CuT1 inoculation. Collectively, our findings suggest that Novosphingobium sp. CuT1 can be utilized as an applicable bioresource to enhance rhizoremediation performance in heavy metal– and TPH-contaminated soils.


Introduction
Copper (Cu) is an essential mineral that promotes plant growth (Mir et al. 2021). However, excessively high Cu levels inhibit plant development and growth due to physiological disturbances (Mir et al. 2021). Moreover, high Cu concentrations (≥ 100 mg/kg) also have a toxic effect on soil microorganisms and reduce the availability of nutrients such as nitrogen and phosphorus (Mir et al. 2021). Lead (Pb) is a highly toxic non-essential heavy metal (Mitra et al. 2021) that can be bioaccumulated by plants and animals (Acuña et al. 2021). Cu and Pb coexist in contaminated soils, especially in the cropland (Rizwan et al. 2021). Continuous contamination of Cu and Pb increases leaf necrosis and chlorosis symptoms and inhibits plant growth (Khan et al. 2017) and has a limited effect on the diversity and activity of microorganisms living in soil (Abdu et al. 2016). Total petroleum hydrocarbons (TPHs) such as diesel and gasoline Responsible Editor: Kitae Baek are widely used as fuels (Gran-Scheuch et al. 2020). Soil is contaminated with TPHs by spill accidents and leakage from aging storage tanks (Gran-Scheuch et al. 2020). Polycyclic aromatic hydrocarbons (PAHs) were a major impairment to photosystem II and thylakoid activity in plants . Co-contamination of soil with heavy metals and TPHs severely affects the growth and activity of soil organisms and, consequently, the remediation process of the contaminated soil by soil microorganisms (Sandrin and Maier 2003;Zhang et al. 2011).
Rhizoremediation is an environmentally friendly and economical method based on the synergism between plant and rhizobacteria or the remediation of heavy metal-and TPH-contaminated soil (Aransiola et al. 2017;Seo and Cho 2020;Hoang et al. 2021). Plant growth-promoting rhizobacteria (PGPR) improves the rhizoremediation of heavy metals and TPHs (Hong and Cho 2007). Some PGPR have plant growth-promoting traits such as indole-3-acetic acid (IAA) production, dehydrogenase (DH) production, and siderophore production in contaminated soil (Jung et al. 2007;Kim et al. 2010;Lee and Bae 2011;Chakraborty et al. 2021). Additionally, some have heavy metal tolerance, as well as plant growth-promoting traits (Chlebek et al. 2022). Heavy metals in the rhizosphere are adsorbed into plant roots or are converted into insoluble species, which reduces their bioavailability (He et al. 2021). Some PGPR can improve the extraction efficiency of heavy metals into plant tissues by enhancing the bioavailability of heavy metals (He et al. 2021). Zhang et al. (2015) demonstrated that Bacillus megaterium not only increased the growth of Napier grass (Pennisetum hybridum) but also promoted the absorption of manganese. Furthermore, Enterobacter sp. JYX7 inoculation increased rapeseed biomass (Brassica napus L.) and zinc bioavailability (Asad et al. 2019). In particular, the genus Novosphingobium, to which the strain CuT1 used in this study belongs, is reportedly suitable for rhizoremediation. The genus Novosphingobium is not only resistant to heavy metals but also exhibits TPH degradation (Segura et al. 2017;Chettri and Singh 2019;Wang et al. 2021). In addition, its potential as a PGPR that enhances plant growth has been investigated (Islam et al. 2013;Krishnan et al. 2017;Vives-Peris et al. 2018).
Tall fescue (Festuca arundinacea L.) is a widely cultivated perennial grass with excellent heavy metal remediation efficiency due to its high tolerance and accumulation capacity for heavy metals (Li et al. 2021a;Peng et al. 2021). Hong and Cho (2007) reported that the remediation efficiency of diesel-contaminated soil planted only with tall fescue was 50%, and its efficiency improved to 90% upon PGPR inoculation. When Enterobacter cloacae CAL2 and Enterobacter cloacae UW4 were inoculated into the TPH-contaminated soil planted with tall fescue, the TPH removal efficiency was 90% (Zhuang et al. 2007). Additionally, the inoculation of these strains improved the growth of plants in soil contaminated with polycyclic aromatic hydrocarbons (PAHs) (Zhuang et al. 2007). The inoculation of Acinetobacter sp. RA1, Bacillus sp. EH7, and Bacillus sp. RA2 increased the germination ratio of tall fescue to 60% in Cu and Cd at 50 ppm each (Ke et al. 2021). Burkholderia cepacia JB12 could increase the biomass of tall fescue's shoots and roots and promote Pb uptake into tall fescue by increasing Pb bioavailability in soil contaminated with 1000 ppm of Pb (Jin et al. 2013). Previous studies on the effect of PGPR inoculation on remediation were conducted under single heavy metal or single TPH-contaminated soil. More research is required to explore the effect of PGPR inoculation on the remediation of soils co-contaminated with heavy metals and TPHs (Chakraborty et al. 2021).
In this study, a plant growth-promoting rhizobacterium with diesel-degrading capacity and heavy metal tolerance was isolated from the rhizosphere of tall fescue planted in soil contaminated with heavy metals and diesel. Afterward, the effects of inoculating the isolated strain on rhizoremediation performance were evaluated in heavy metal-and diesel-contaminated soil planted with tall fescue.

Isolation and identification of a heavy metal tolerance and PGP rhizobacteria
Metal-contaminated soil was used to obtain PGP rhizobacteria with heavy metal tolerance. The soil was artificially contaminated with heavy metals (Cu, Pb, and Cd) and diesel. The initial contaminant concentration was 500 mg·kg-soil −1 of each metal and 30,000 mg-TPH·kg-soil −1 . Tall fescues had been grown in the contaminated soil for 4 months. Rhizosphere soil adhering to the tall fescue roots was collected by gentle shaking. The collected soil sample was mixed with sterile water at a 1:9 ratio (w/v), serially diluted, and then spread on a plate of 1/10 Luria-Bertani (LB) medium containing 1.5% agar and 1 mM of Cu. The plates were cultivated at 30 ℃ for 72 h. A yellow colony, which dominated the plates, was isolated and named strain CuT1. To evaluate the Cu tolerance of strain CuT1, the strain was cultivated on a 1/10 LB agar plate containing 0.5-10 mM of Cu. The Pb tolerance of strain CuT1 was also evaluated using a 1/10 LB agar plate containing 0.5-10 mM of Pb. The plates were cultivated at 30 °C for 72 h. Cu and Pb 1 M stock solutions were prepared using CuSO 4 ·5H 2 O and Pb(NO 3 ) 2 , respectively. The stock solution was added to the 1/10 LB medium at final concentrations of 0.5, 1, 5, and 10 mM.
To evaluate the PGP properties of strain CuT1, indole-3-acetic acid (IAA), siderophore, and 1-aminocyclopropane-1-carboxylic acid (ACC) deaminase production were investigated as described by Lee et al. (2021a). Strain CuT1 was identified by 16S rRNA sequencing analysis. To extract the genomic DNA of strain CuT1, a colony was mixed with 30 μL of sterilized water and centrifuged at 11,000 rpm for 5 s. Then, the suspension was heat-treated for 15 min in a heating block at 95 °C and centrifuged at 11,000 rpm for 5 s. This process was repeated three times to obtain the genomic DNA of strain CuT1. The genomic DNA sample was used as a template for PCR. To identify strain CuT1, PCR was performed using the 340F (5′-TCC TAC GGG AGG CAG CAG-3′) and 805R (5′-GAC TAC HVG GGT ATC TAA TCC-3′) primer set, which targets the 16S rRNA gene (Kim et al. 2013). The PCR mixture and thermocycling conditions were described previously in detail (Kim et al. 2013). The PCR products were sequenced by Macrogen (Seoul, Republic of Korea), and the resulting sequence was analyzed using the Basic Local Alignment Search Tool (BLAST) developed by the National Center for Biotechnology Information (NCBI). The sequence was deposited in the NCBI GenBank database under accession number MW407066.

Characterization of the heavy metal tolerance and diesel degrading capabilities of strain CuT1
To evaluate the effect of heavy metals (Cu and Pb) on the growth of strain CuT1, the strain was pre-cultured using a 1/10 LB medium (Hwang and Song 2020;Lee 2009). The CuT1 pre-culture was incubated at 30 °C under constant shaking (120 rpm) for 48 h. CuT1 cells were harvested by centrifuging at 20,784 g for 10 min, after which the resulting pellet was washed with sterile water. Finally, the pellet was re-suspended in sterile water to obtain the desired density (OD 600 of 0.6). Next, 10 mL of CuT1 suspension (OD 600 of 0.6) was mixed with 100 mL of 1/10 LB medium containing heavy metals (Cu and Pb) at final concentrations of 0, 10, 50, 100, 250, and 500 μM. The culture medium was incubated at 30 °C and 120 rpm. The absorbance (OD 600 ) was measured using a UV/Vis spectrophotometer (Libra S22, Biochrom Ltd., Cambridge, England) at 3-h intervals. The specific growth rate and doubling time of strain CuT1 in the presence of heavy metals were calculated according to the methodology described in previous studies (Koo and Cho 2007;Kim et al. 2015).
To investigate the diesel-degrading ability of strain CuT1, 0.5 mL of CuT1 suspension (OD 600 of 0.6) was inoculated in 5 mL of inorganic salt medium containing 3% (v/v) sterilized diesel in a sterilized test tube. The inorganic salt medium contained MgSO 4 ·7H 2 O (0.25 g/L), Na 2 HPO 4 ·12H 2 O (17.09 g/L), KH 2 PO 4 (3 g/L), NaCl (0.5 g/L), CaCl 2 (0.01 g/L), and NH 4 Cl (1 g/L). The test tubes were incubated at 30 °C and 120 rpm for 2 weeks. Then, 5.5 mL of culture medium was mixed with 5 mL of hexane and incubated at 30 °C and 200 rpm for 30 min. Afterward, the mixtures in the test tubes were kept at room temperature for 30 min to obtain a supernatant. The diesel concentration extracted from the supernatant was measured via a gas chromatography (GC 6980 N system, Agilent Technologies, CA, USA) equipped with a flame ionization detector (FID; Agilent Technologies) and an HP-5 capillary column (30 m × 0.320 mm × 0.25 μm, Agilent Technologies). All procedures were performed as described in our previous report (Lee et al. 2021a). The diesel removal efficiency was calculated as a percentage by comparing the initial diesel concentration with the residual diesel concentration after 2 weeks.

Soil preparation
Pot experiments were conducted to investigate the effect of CuT1 inoculation on rhizoremediation performance in co-contaminated soil (heavy metals and diesel). Cultivated soil was collected from a garden on the rooftop of the New Engineering Building, Ewha Womans University, Seoul, South Korea (37°57′N, 126°95′E). Weeds and stones were removed from the soil, and compost was added at a 0.5:9.5 (w/w) ratio to provide nutrients (N and P) as previously described (Lee et al. 2022). The soil was then artificially contaminated with heavy metals (Cu and Pb) and diesel (Li et al. 2021b;Banda et al. 2022;Kujawska and Pawłowska 2022). Next, 500 mg/L Cu and Pb stock solutions were prepared using CuSO 4 ·5H 2 O and Pb(NO 3 ) 2 , respectively, after which each solution was sterilized using a sterile syringe filter (0.45 μm). The initial contaminant concentrations were established based on the Korean Soil Pollution Concern Standards (Korean Ministry of Environment 2018a). A total of 50 mL of each stock solution was mixed with 24.3 kg of the soil to achieve a final concentration of Cu and Pb of 500 mg·kg-soil −1 . The metalcontaminated soils were divided into three experimental groups; heavy metals + non-diesel (HM treatment), heavy metals + 10,000 mg-diesel·kg-soil −1 (HM + 10D treatment), and heavy metals + 30,000 mg-diesel·kg-soil −1 (HM + 30D treatment). The final contaminated soils were stored in a shaded location on the rooftop for 24 h before use.
The pot experiments were conducted in 120-mm-diameter and 115-mm-height pots. Coarse sand was evenly placed at the bottom of each pot for drainage. For each treatment, 450 g of contaminated soil was added on top of the coarse sand in the pots. A total of 54 pots were prepared with three treatments (HM, HM + 10D, and HM + 30D) and 18 replicates.

Plant cultivation and inoculation of strain CuT1
Tall fescue seeds were sown in a garden on the rooftop of the New Engineering Building at Ewha Womans University for 2 months. After, two tall fescue seedlings with healthy shoot lengths of 30-40 cm were planted in each experimental pot. The pots were then inoculated with a CuT1 cell suspension (OD 600 of 0.6) prepared as described above ("Characterization of the heavy metal tolerance and diesel degrading capabilities of strain CuT1" section). Of the 54 pots prepared in total, 30 mL of the strain CuT1 suspension was inoculated around the tall fescue roots in 27 pots (CuT1 inoculation group). In the control groups, 30 mL of sterilized water was dispensed into the remaining 27 pots. The pot experiment was conducted on the rooftop of the New Engineering Building, Ewha Womans University, for 60 days (August 12, 2021-October 11, 2021). The pot soil was watered twice a week to maintain an average water content of approximately 25-35% during the experimental period. On days 25 and 55, identical volumes of the CuT1 suspension were re-inoculated in the pots due to heavy rain.

Soil sampling and physicochemical characteristics analysis
The soil was randomly sampled (i.e., destructive sampling) from the pots at 20-day intervals. The soil adhering to the roots was collected by gentle shaking. Approximately 80 g of the collected soil was air-dried at room temperature for half a day. The air-dried soil was sieved through a 2-mm mesh and stored at 4 °C before analysis. Some of the collected soil was oven-dried at 110 °C for 4 h for heavy metal analysis, whereas the remaining sample was freeze-dried for diesel analysis.
The pH and water content of the soil were measured according to the Korean Standard Soil Analysis Methods (ES.07302.1b and ES.07301.1b) (Korean Ministry of Environment 2018b). Briefly, soil pH was measured by mixing 5 g of soil and 25 mL of distilled water for 30 min. Then the pH values were measured by a pH meter (Orion model 420A, Thermo Scientific Inc., Tokyo, Japan). The soil was dried in a constant temperature dryer (VS-4150ND, Vision Scientific Co., Ltd., South Korea) at 110 °C for more than 4 h to measure the water content. Thereafter, the water content was calculated by measuring the weight of the soil before and after drying. The organic matter content was measured based on the Korean Standard Waste Analysis Method (ES.06303.1c)) (Korean Ministry of Environment 2017). After drying the soil at 110 °C, it was dried in a furnace (DH.WF11.27 Furnace, DAIHAN Scientific Co., Ltd., South Korea) at 550 °C for 1 h.
The organic matter content was calculated by measuring the weight of the soil before and after drying at 550 °C.

Instrumental analysis
Total contents of heavy metals (Cu and Pb) in soils were measured following a modified version of an Environmental Protection Agency protocol (EPA 3050B) and other methods described in previous studies (Mapanda et al. 2005;Huang et al. 2007;Long et al. 2021;Yahaya et al. 2021). The ovendried soil samples were finely ground using mortar. Briefly, 0.1 g of ground soil was sequentially acid-treated with HNO 3 -HF-HClO 4 (4:4:2, v/v/v), HF-HClO 4 (1:1, v/v), and aqua regia (HCl-HNO 3 , 3:1, v/v) at 180 °C for 8 h. After each acid treatment step, the samples were washed with distilled water, and the acid fumes were heat-treated at 130 °C for 4 h in a fume hood. After the fuming step, the samples were mixed with 50 mL of 2% (v/v) HNO 3 solution. The total contents of heavy metals (Cu and Pb) were measured with an atomic absorption spectrophotometer (AAnalyst 400, Perkin Elmer Inc., USA). A calibration curve was prepared with 0.5-10 ppm Cu and Pb standard solutions to calculate the heavy metal concentration.
The bioavailable contents of heavy metals in soils were extracted using 0.01 M of CaCl 2 solution (Rajapaksha et al. 2012;Bandara et al. 2017). Briefly, 2 g of ground soil was mixed with 10 mL of 0.01 M CaCl 2 solution. The mixture was incubated at 25 °C and 150 rpm for 2 h. The suspension was then centrifuged at 5580 g for 5 min and filtered through a PTFE syringe filter (pore size 0.45 μm). The supernatant was analyzed via atomic absorption spectrometry using an AAnalyst 400 system (Perkin Elmer).
The residual diesel content in the soil was measured as follows. A total of 3 g of freeze-dried soil was mixed with 5 mL of hexane-acetone (1:1, v/v) solution as an extractant and then incubated at 30 °C and 120 rpm for 30 min. After incubation, the supernatant was obtained and the concentration of diesel extracted from the supernatant was measured using a gas chromatograph (GC 6980 N system, Agilent Technologies, CA, USA) equipped with a flame ionization detector (FID; Agilent Technologies) and an HP-5 capillary column (30 m × 0.320 mm × 0.25 μm, Agilent Technologies). The operating conditions for the gas chromatography system are described in our previous study (Lee et al. 2021b). The temperatures of the injector and detector were 300 °C and 320 °C, respectively. The oven temperature was maintained at 60 °C for 3 min, increased to 260 °C at a 4 °C⋅min −1 rate, to 310 °C at an 84 °C⋅min −1 rate, and finally maintained at 310 °C for 5 min. The removal efficiency was calculated according to Eq. (1): where C 0 is the initial diesel concentration and C i is the concentration on the ith day.

Soil enzyme activity (IAA, siderophore, and DH production)
Indole-3-acetic acid (IAA) production, a plant growth-promoting hormone, was evaluated through a colorimetric method (Benitez et al. 2004) as described in previous reports (Benitez et al. 2004;Lee et al. 2021c). A total of 2 g of soil was placed in a 50 mL flask. In total, 6 mL of phosphate buffer (pH 7.5) with glucose (1 g glucose 100 mL −1 phosphate buffer) and 4 mL of 0.06% (w/v) L-tryptophan were added. Sterilized water instead of L-tryptophan was used as a control. The mixture was incubated at 37 °C for 24 h in the dark. The mixture was centrifuged at 4000 rpm for 5 min, after which 3 mL of supernatant was transferred to a test tube. A total of 2 mL of Salkowski's reagent solution (2 mL of 0.2 M FeCl 3 ·6H 2 O and 98 mL of 35% (v/v) HClO 4 ) was added, and the mixture was incubated for 30 min at 25 °C in the dark. The absorbance was measured using a UV/Vis spectrophotometer (Biochrom Ltd.) at 535 nm. The IAA was determined as the amount of IAA (in μg) produced per 1 g of dry soil per day (μg-IAA·g-dry-soil −1 ·d −1 ).
The production of siderophores, a type of heavy metal chelating agent in soil, was performed as described in previous studies (Nagarajkumar et al. 2004;Hwang and Song 2020;Lee et al. 2021c). Briefly, 2 g of air-dried soil was added to a 50 mL flask, followed by 20 mL of sterilized nutrient broth. The mixture was then incubated at 30 °C and 120 rpm for 24 h, after which 14 mL of supernatant was transferred to a fresh test tube. The pH of the supernatant was adjusted to 2 using 1 M of HCl, after which an equal volume of ethyl acetate was added. Next, 2 mL of Hathway reagent was added to each test tube and incubated at 30 °C for 30 min in the dark. Siderophore production was measured with a UV/Vis spectrophotometer (Biochrom Ltd.) at 700 nm. A standard curve was prepared using dihydroxy benzoic acid (Nagarajkumar et al. 2004), and the quantity of siderophore produced was expressed as μM-benzoic acid per 1 g of dry soil per day (μM-benzoic acid·g dry soil −1 ·d −1 ).
Dehydrogenase (DH) activity was measured via the triphenyl tetrazolium chloride (TTC) method (Liu et al. 2018), described in detail in previous reports (Lee et al. 2021c;Liu et al. 2018). A total of 1 g of the soil was mixed with 2 mL of Tris-HCL buffer (pH 7.6) and 1 mL of 1% (w/v) triphenyl tetrazolium chloride (TTC) solution in a test tube and incubated in the dark at 37 °C for 24 h (Liu et al. 2018). Then, 96% ethanol (10 mL) was added and mixed by vortexing at 1200 rpm for 20 s to extract the 1,3,5-triphenyltetrazolium formazan (TPF). The mixture was centrifuged at 4000 rpm for 5 min to obtain the supernatant. Finally, triphenyl formazan (TPF) concentrations were measured with a UV/Vis spectrophotometer (Biochrom Ltd.) at 485 nm. DH was determined as the amount of TPF (in μg) produced per 1 g of dry soil per day (μg-TPF·g-dry-soil −1 ·d −1 ).

Plant roots sampling and heavy metal uptake analysis
Tall fescue roots were sampled after removing the rhizosphere soil from the roots. Tall fescue roots were lightly washed with tap water to completely remove the residues. All root samples were air-dried in a wellventilated place to prevent additional contamination, after which the root lengths and weights were measured. The roots were oven-dried at 70 °C for 72 h (Lee et al. 2016) and the dry weights of the roots were then measured.
The amount of heavy metal uptake by tall fescue roots was measured using a 0.01 M CaCl 2 solution. The ovendried root samples were ground into fine powder for the measurement and then mixed with a 0.01 M CaCl 2 solution at a 1:10 (w/v) ratio. The procedure was identical to the analysis of bioavailable heavy metals content in soils ("Instrumental analysis" section). To evaluate heavy metal uptake by tall fescue, the bioconcentration factor (BCF) was calculated using Eq. (2) (Oh et al. 2012;Ali et al. 2013;Liu et al. 2022): where C r is the bioavailable heavy metal concentration (mg·kg −1 ) in the roots and C s is the bioavailable heavy metal concentration in the soil (mg·kg −1 ).

Statistical analysis
All measurements were conducted in triplicate and average results were presented. The correlations between CuT1 inoculation, bioavailable heavy metals contents, soil enzyme activates, and plant biomass were visualized using the "corrplot" package in R (v. 4.0.1). T-tests were also conducted using R (v. 4.0.1) at a 0.05 p-value significance threshold. Standard deviation (SD) values were computed as an average of three replications. Significant effects were identified as follows. A normality check was performed using the Shapiro-Wilk test. Based on the results of normality tests, one-way ANOVA followed by multiple comparison tests (Tukey's test) was also conducted using R (v. 4.0.1) at a 0.05 p-value significance threshold. The graphical work was performed in SigmaPlot program (v. 12.5).

Characteristics of Novosphingobium sp. CuT1
Strain CuT1 isolated from the tall fescue's rhizosphere in the soil contaminated with heavy metals and diesel was identified as Novosphingobium sp. (Fig. 1). The CuT1 isolate was able to grow on 1/10 LB-agar plates containing 5-mM Cu or 5-mM Pb and could degrade diesel (Table 1). A previous study reported that Novosphingobium sp. was tolerant to heavy metals and promoted the growth of rice in Cd-contaminated soil ). This bacterium also expressed pahA and pahT, which are functional genes that contribute to the degradation of PAHs (Segura et al. 2017). Chettri and Singh (2019) reported that Novosphingobium panipatense P5:ABC was tolerant to various heavy metals, such as Fe, Al, Cr, Pb, Ni, and Zn, and it could degrade aliphatic and aromatic hydrocarbons.  Strain CuT1 had plant growth-promoting traits such as ACC deaminase activity and IAA and siderophore production (Table 1). Similar to our finding, Novosphingobium pokkalii could produce IAA and siderophore (Krishnan et al. 2017). Novosphingobium sp. RFNB21 significantly increased the growth of tomatoes and red peppers due to its N 2 -fixing ability, IAA production, and ACC deaminase activity (Islam et al. 2013). Novosphingobium sp. HR1a could produce plant hormones such as abscisic acid, salicylic acid, and IAA and promote the growth of citrus by minimizing salt stress (Vives-Peris et al. 2018).
The effects of Cu and Pb on strain CuT1's growth are shown in Fig. 2. The strain CuT1 hardly grew in the 1/10 LB medium containing 250 μM Cu (Fig. 2a) and 500 μM Pb (Fig. 2b). The doubling time in microbial growth tends to increase with increasing heavy metal concentrations (Cha et al. 2003). In this study, as the concentration of heavy metals increased, the doubling time for the strain CuT1 growth increased (Fig. 2c). The specific growth rate of strain CuT1 was 0.098 h −1 in the 1/10 LB medium without heavy metals, which decreased with increasing heavy metal concentrations (Fig. 2d). The EC 50 values (i.e., the concentration at which the specific growth rate decreases by half) for Cu and Pb were 210 and 280 μM, respectively.
The doubling time of a rhizobacterium isolated from the rhizosphere of Phragmites inhabiting Cu mines was 80-90 min in non-contaminated conditions. However, the doubling increased to more than 100 min in the presence of 500-μM Cu (Kunito et al. 2001). Similarly, the doubling time of Yarrowia lipolytica NCIM 3589 was 6 h in noncontaminated conditions but increased to more than 84 h and 96 h when the soil was contaminated with 1-mM Cr and Cu, respectively (Bankar et al. 2018). Furthermore, the specific growth rate of Beauveria bassiana at 1000 ppm Cu was 1/5 of the rate in the absence of Cu (Gola et al. 2016). Table S1 summarizes the total Cu and Pb concentrations in soil and bioavailable Cu concentrations in soil and tall fescue roots during the rhizoremediation of soil co-contaminated with heavy metals and diesel. Some studies have suggested that the bioavailable concentration of heavy metals is a better indicator for risk assessment compared to total heavy  (Ma et al. 2002;Wu et al. 2006). In this study, to analyze the bioavailable heavy metal concentration, 0.01 M of CaCl 2 solution was used as an extraction solvent. Nevertheless, the bioavailable Pb concentration was not accurately analyzed with this solvent (Krzyżak et al. 2017;Nenova et al. 2018). Therefore, the effect of CuT1 inoculation on the behavior of bioavailable metal concentration was evaluated only for Cu in the present study (Table S1). When strain CuT1 was inoculated into the heavy metal-contaminated soil without diesel (HM sample), the bioavailable Cu concentration was 1.25 times higher than the control (non-inoculation) (Table S1). In the soil co-contaminated with heavy metals and 10,000 ppm of diesel (HM + 10D sample), the bioavailable Cu concentration was increased by 1.03-fold upon inoculation with strain CuT1. In the soil co-contaminated with heavy metal and 30,000 ppm of diesel (HM + 30D sample), Cu bioavailability increased 1.19-fold times by the inoculation. Furthermore, CuT1 inoculation increased the bioavailable Cu concentration in the tall fescue's root by 1.2-4.0-fold (Table S1).

Effect of Novosphingobium sp. CuT1 inoculation on bioavailable Cu dynamics
Based on the results in Table S1, the linear correlation coefficients between the total Cu concentration and the bioavailable Cu concentration were compared between samples with and without CuT1 inoculation (Fig. 3). The correlation coefficients in the non-inoculation and inoculation conditions were 0.32 and 0.62, respectively (p < 0.05), indicating that CuT1 inoculation increased Cu bioavailability. Several studies have reported that the bioavailability of heavy metals in rhizosphere soil can be enhanced by PGPR inoculation. The bioavailable Cd concentration in the soil was significantly increased by 1.4-fold by inoculation with Bacillus sp. QX8 . A previous study reported that coinoculation with Azotobacter chroococcum HKN-5, Bacillus megaterium HKP-1, and Bacillus mucilaginosus HKK-1 increased the bioavailable Pb concentration in the soil by 1.3-fold (Wu et al. 2006).
The bioconcentration factor (BCF) is an index that determines the uptake and transport of heavy metals from soil to plants (Kwag 2019) and clarifies the mobility of heavy metals in the soil (Kwag 2019). To evaluate the effects of CuT1 inoculation on the BCF values of bioavailable Cu, the BCF values were calculated for each experimental condition (Fig. 4). In the HM sample, the BCF values in the inoculated soil were significantly greater than those in the non-inoculated soil on days 20 and 40, whereas the BCF values between the two conditions were almost identical on day 60 ( Fig. 4(a)). In the HM + 10D sample, the BCF values in the inoculated soil were greater than those in the non-inoculated soil after the 20th day, although the BCF value in the non-inoculated soil was greater than in the inoculated soil on day 20 (Fig. 4(b)). In the HM + 30D sample, the value was only greater on day 20 in the inoculated soil compared with the BCF values in the non-inoculated soil (Fig. 4(c)).
The metabolic activity of rhizobacteria can increase the transfer ratio of heavy metals from the soil into plants. Heavy metal uptake from the soil into the root of Chrysopogon zizanioides was enhanced 1.2-1.5-fold by Serratia marcescens sp. SNB6 inoculation (Wu et al. 2020). Similarly, Bacillus cereus T1B3 inoculation doubled the metal transfer from soil to the roots and shoots of Cymbopogon citratus (Nayak et al. 2019). Previous studies have evaluated the effect of PGPR inoculation on BCF values based on the total heavy metal concentration. Our study characterized the effect of PGPR inoculation on the BCF value based on the bioavailable metal concentration, as these concentrations are becoming increasingly important in the development of soil remediation technology (Bouhadi et al. 2021;Xu et al. 2021). Effect of Novosphingobium sp. CuT1 inoculation on tall fescue growth and PGP activity Figure S1 shows the plant growth-promoting activity (IAA, siderophore, and DH production) of the soil samples during the rhizoremediation of the contaminated soil. In the HM sample, the IAA concentrations in the inoculated soil were similar to or less than those in the non-inoculated soil ( Fig. S1A-a). However, IAA concentrations were significantly enhanced by CuT1 inoculation in the HM + 10D and HM + 30D samples (Fig. S1A-b & c). The siderophore concentrations in the inoculated soil were similar to or lower than those in the non-inoculated soil (Fig. S1B). Furthermore, the DH concentrations in the soil inoculated with strain CuT1 were similar to or higher than those in the soil without CuT1 inoculation. In contrast, the DH concentrations in the inoculated soil tended to increase over time (Fig. S1C). Figure 5 shows the relationship between the root length of tall fescue and the PGP activity of the soil. IAA, a plant growth-promoting hormone, can enhance the growth of plant roots in the rhizosphere (Moon and Yoon 2019;Wang et al. 2022). Even though IAA concentration in the non-inoculated soil showed a negative correlation (r = − 0.65) with the root length (Fig. 5a), a positive correlation (r = 0.30) was observed in the CuT1-inoculated soil (Fig. 5b). These results suggest that the positive effects of CuT1 inoculation on the root growth of tall fescue are largely attributable to the IAA production activity of the strain. IAA can improve plant growth by stimulating root growth via nodule formation (Chlebek et al. 2022). Moreover, IAA can increase the number of root exudates by slackening the cell walls (Chlebek et al. 2022). Root exudates can prevent heavy metals  (Chlebek et al. 2022).

Total Cu concentration (mg-
Siderophore concentration in the inoculated soil had a positive correlation (r = 0.20) with the root length of tall fescue, although a negative correlation (r = − 0.13) was observed in the non-inoculated soil ( Fig. 5c and d). Siderophores are highaffinity Fe-chelating proteins that promote Fe uptake by plants (Govarthanan et al. 2016). Fe is an essential micronutrient for plants and microorganisms involved in various important biological processes, such as DNA synthesis, respiration, and photosynthesis (Chlebek et al. 2022). Siderophores not only stimulate plant growth but can also form stable complex compounds with other metals such as Al, Cd, Cu, and Pb. These complexes promote the absorption of heavy metals into plants (Chlebek et al. 2022).
Regardless of CuT1 inoculation, DH concentration had a positive relationship with the root length of the tall fescue ( Fig. 5e and f), and the correlation coefficient in the inoculated soil (r = 0.59) was greater than that in the non-inoculated soil (r = 0.05). DH is one of the most essential soil enzymes (Wolf et al. 2020). DH activity showed a positive correlation with TPH degradation and was one of the most highly active enzymes in TPH-contaminated soil (Ramirez et al. 2021). DH activity is an important indicator to evaluate the activity of soil microorganisms (Małachowska-Jutsz and Matyja 2019) and an index to measure the adaptability of rhizobacteria in heavy metal-contaminated soil . Therefore, Novosphingobium sp. CuT1 may have promoted the root elongation of tall fescue by increasing soil DH activity through interactions with native rhizobacteria. Figure S2 illustrates the time profile of residual diesel concentrations during the rhizoremediation of co-contaminated soil. Although diesel degradation was hardly shown during the experimental period (60 days), the diesel degradation in the inoculated soil was superior to that in the non-inoculated soil contaminated with heavy metals and 30 g·L −1 of diesel (HM + 30D sample). Zhuang et al. (2007) reported that the remediation efficiency of the co-contaminated soil may not be significantly improved by introducing a single PGPR. A microbial consortium composed of microorganisms with various functions can be used to overcome this dilemma (Huang et al. 2005;Divya and Kumar 2011). Abtahi et al. (2020) also reported that the application of TPHs degrading bacterial consortium could improve TPH degradation and bioremediation. Meanwhile, previous studies have reported that Cu inhibits the TPH-degrading activity of microorganisms (Almeida et al. 2013;Ha et al. 2021). When heavy metals and TPH coexist in the soil, the presence of toxic heavy metals hinders the activity of TPH-degrading bacteria (Sprocati et al. 2012). Therefore, the extremely low diesel degradation rates observed in our study were presumably due to the negative effects of heavy metal and diesel co-contamination on the TPH degradation activity of indigenous soil microorganisms and the inoculated CuT1 bacterium. Figure 6 shows the correlation between strain CuT1 inoculation, Cu bioavailability, BCF, diesel removal efficiency, PGP activity (IAA, siderophore, and DH), and tall fescue root biomass. Cu bioavailability and BCF were positively correlated with CuT1 inoculation, as shown in Figs. 3 and 4. Our study is the first to demonstrate that Novosphingobium sp. can enhance Cu bioavailability. Diesel removal efficiency exhibited a positive correlation with CuT1 inoculation (Fig. 6). Furthermore, CuT1 inoculation had positive correlations with IAA and DH activity. Additionally, CuT1 inoculation increased the root elongation of tall fescue (Fig. 5), consistent with the above-described results. Krishnan et al. (2017) reported that Novosphingobium pokkalii sp. L3E4 T , which produces IAA, promoted rice growth. Gogoleva et al. (2019) reported that Serratia proteamaculans B1 used DH to produce cytokinin, an important PGP hormone, and could thus promote plant growth.

Correlation analysis between key parameters and strain CuT1 inoculation
In contrast, CuT1 inoculation was negatively correlated with siderophore activity, although the siderophore concentration showed a positive correlation with the tall fescue's root length (Fig. 5). Representative siderophore-producing PGPR include Acetobacter, Azotobacter, Bacillus, Burkholderia, and Pseudomonas (Kumar et al. 2018). Our findings suggest that the contribution of indigenous soil bacteria to siderophore production is higher than that of the inoculated bacterium.
The dry weight of the tall fescue roots had a highly positive correlation coefficient (0.49) with CuT1 inoculation (Fig. 6). The interaction between plant roots and inoculated bacterial strains often enhances plant biomass (Deng and Cao 2017). Tall fescue is a representative plant with welldeveloped roots that can promote the activity of rhizosphere bacteria due to the large surface area of its roots, in addition to degrading pollutants and adsorbing heavy metals (Liu et al. 2013). However, previous studies have reported that pollutants such as TPHs and heavy metals decrease the biomass of tall fescue (Liu et al. 2013;Zhong et al. 2020). Similarly, when two types of heavy metals are simultaneously present in the soil, they adversely affect plant growth compared to when one type of heavy metal is present due to its antagonistic effects (Ghani 2010). Additionally, the binding of two types of heavy metals in the soil reduces the nitrogen and protein contents in plants (Ghani 2010). Our findings indicate that the growth-inhibiting effects of TPHs and heavy metals on tall fescue can be overcome by inoculating the soil with Novosphingobium sp. CuT1.

Conclusions
Rhizoremediation is a promising technology for the remediation of soil co-contaminated with heavy metals and diesel. To improve the performance of rhizoremediation using the synergy between plants and rhizobacteria, it is essential to identify microbial strains with pollutant resistance and removal ability and PGP ability is crucial. Novosphingobium sp. CuT1, which was isolated from the rhizosphere of tall fescue, was not only tolerant to heavy metals but could also degrade diesel and exhibited key PGP traits such as IAA, ACC deaminase, and siderophore productivity. When Novosphingobium sp. CuT1 was introduced into soil co-contaminated with heavy metals and diesel, the Cu bioavailability, BCF, and root biomass of the tall fescue were higher than those in the soil without CuT1. Diesel removal showed a positive relationship with CuT1 inoculation, but the removal efficiency was below 30%. These results suggest that Novosphingobium sp. CuT1 is a promising microbial strain to enhance the rhizoremediation performance of soil co-contaminated with heavy metals and TPHs. However, the diesel degradation ability of strain CuT1 was unsatisfactory. Further study is required to improve the diesel degradation efficiency by strain CuT1 in a limited environment, as in this study. The activity of microorganisms introduced into the soil is affected by physicochemical factors such as root exudates, pH, dissolved oxygen, and contaminant properties and concentrations and by the interaction with indigenous soil microorganisms. Therefore, follow-up studies are needed to identify the optimal physicochemical conditions under which the activity of the inoculated microorganism can be maximized in the contaminated soil. Additional research on the genetic information of strain CuT1, such as heavy metal resistance genes or TPHs degradation genes, is necessary. Further studies are required to elucidate the behavior of inoculated microorganisms in the contaminated soil and their interaction with indigenous soil microorganisms.
Author contribution Soo Yeon Lee, Yun-Yeong Lee, and Kyung Suk Cho contributed to the study conception and design. Material preparation and data collection were performed by Soo Yeon Lee and Yun-Yeong Lee. Data analysis was performed by Soo Yeon Lee and Kyung Suk Cho. The first draft of the manuscript was written by Soo Yeon Lee, and Kyung Suk Cho commented on the previous version of the manuscript. All the authors read and approved the final manuscript.
Funding This study was supported by the Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Science and ICT (NRF-2019R1A2C2006701).

Data availability
The datasets generated and/or analyzed during the current study are available from the corresponding author upon reasonable request.

Declarations
Ethics approval This research does not involve human or animal subjects.

Competing interests
The authors declare no competing interests. Fig. 6 Correlation coefficient matrix of eight parameters. Positive coefficients are represented by blue squares, which indicate a direct relationship between variables in the matrix, and negative coefficients are shown as red squares, which reflect an inverse relationship