Enhancing the Rice Seedlings Growth Promotion Abilities of Azoarcus sp. CIB by Heterologous Expression of ACC Deaminase to Improve Performance of Plants Exposed to Cadmium Stress

Environmental pollutants can generate stress in plants causing increased ethylene production that leads to the inhibition of plant growth. Ethylene production by the stressed plant may be lowered by Plant Growth-Promoting Bacteria (PGPB) that metabolizes the immediate precursor of ethylene 1-aminocyclopropane-1-carboxylate (ACC). Thus, engineering PGPB with ACC deaminase activity can be a promising alternative to mitigate the harmful effects of pollutants and thus enhance plant production. Here we show that the aromatics-degrading and metal-resistant Azoarcus sp. CIB behaves as a PGP-bacterium when colonizing rice as an endophyte, showing a 30% increment in plant weight compared to non-inoculated plants. The cloning and expression of an acdS gene led to a recombinant strain, Azoarcus sp. CIB (pSEVA237acdS), possessing significant ACC deaminase activity (6716 nmol mg−1 h−1), constituting the first PGPB of the Rhodocyclaceae family equipped with this PGP trait. The recombinant CIB strain acquired the ability to protect inoculated rice plants from the stress induced by cadmium (Cd) exposure and to increase the Cd concentration in rice seedlings. The observed decrease of the levels of reactive oxygen species levels in rice roots confirms such a protective effect. The broad-host-range pSEVA237acdS plasmid paves the way to engineer PGPB with ACC deaminase activity to improve the growth of plants that might face stress conditions.


Introduction
Ethylene is a gaseous plant hormone with essential functions in plant growth and development such as seed germination, cell expansion and elongation of stems and roots, the formation of adventitious roots and root hairs, fruit ripening, photomorphogenesis, and senescence in general [1][2][3][4][5][6]. However, the increase of ethylene concentrations in tissues as a response to environmental stress may cause adverse effects on plant health. A peak of ethylene activates defense systems in plant cells [7][8][9], but in case the stress is maintained, higher amounts of ethylene are produced, which can cause plant growth Table 1. Bacterial strains and plasmids used in this study.

This work
Km r : kanamycin resistant; Gm r : gentamicin resistant; Sm r : streptomycin resistant.

Culture Media and Growth Conditions
Azoarcus sp. CIB was grown aerobically on MC medium (MA basal medium plus trace elements and vitamins) [32] at 30 • C. The composition of MA basal medium was the following, per liter of distilled water: 0.33 g of KH 2 PO 4 , 1.2 g of Na 2 HPO 4 , 0.11 g of NH 4 [32]. As carbon source, 0.2% (w/v) pyruvate was added. When necessary, bacterial strains were also grown on a variant of the VM-ethanol rich medium [28] with the following composition (in 1 L of distilled water): KH 2 PO 4 0.4 g; K 2 HPO 4 0.6 g; NaCl 1.1 g; E. coli was grown at 37 • C in lysogeny broth (LB) medium [49] or in VM-ethanol medium [28] at 30 • C. When appropriate, kanamycin (50 µg mL −1 ) or gentamycin (10 µg mL −1 ) was added to the medium.
To determine the ACC deaminase activity a slightly reformulated DF medium [50] was used (per 1 L of distilled H 2 O): 4.0 g of KH 2 PO 4 ; 6.0 g of Na 2 HPO 4 ; 0.2 g MgSO 4 × 7H 2 O; 2.0 g glucose and 2.0 g of pyruvate. Trace elements and vitamins were added at the same concentrations than used to formulate the MC medium. 3 mM 1-aminocyclopropane-1-carboxylic acid (ACC) was added as the sole nitrogen source.

Molecular Biology Techniques
Standard molecular biology techniques were performed as previously described [51]. All DNA fragments were purified using Gene-Turbo (BIO101 Systems, Irvine, CA, USA). Plasmids and PCR products were purified utilizing a High Pure Plasmid and PCR Product Purifications kits (Roche), respectively. Oligonucleotides were supplied by Sigma Co, and their sequences were: 5 RBSacdS (5 -CCAAGCTTTGACCTAAGGAGGTAAATAATGAACCTGCAACGATTCCCTCGTTAC-3 (HindIII site underlined)); 3 acdS (5 -GGACTAGTTTAGCCGTTGCGGAAAATGAAGCTG-3 (SpeI site underlined)); 5 PtacacdS (5 -TTGGCGCGCCCTGGCAAATATTCTGAAATGAGCTG-3 (AscI site underlined)) and 3 PtacacdS (5 -TTGGCGCGCCTTAGCCGTTGCGGAAAATGAAG-3 (AscI site underlined)). The cloned inserts and the DNA fragments employed were confirmed by DNA sequencing with fluorescently labeled dideoxynucleotide terminators [52] and AmpliTaq FS DNA polymerase (Applied Biosystems, Foster City, CA, USA) in an ABI Prism 377 automated DNA sequencer (Applied Biosystems). Transformations of E. coli were performed by using the RbCl method or by electroporation (Gene Pulser, Bio-Rad, Hercules, CA, USA) [51]. Transformation of Azoarcus sp. CIB was done by biparental conjugation using the strain E. coli S17-1λpir as a donor, following a protocol previously established [44] with the following small modifications: donor cells were grown to A 600 of 5, and the receptor CIB strain was grown on MC medium supplemented with pyruvate 0.2%, and concentrated to reach an A 600 of 35. The transconjugants were selected on MC supplemented with 10 mM glutarate plus the corresponding antibiotic.

Construction of Plasmids pIZacds and pSEVA237acdS
The acdS gene was PCR-amplified by using the oligonucleotides 5 RBSacdS/3 acdS and the genomic DNA template from P. phytofirmans PsJN. The amplified fragment of 1017 bp was first cloned into plasmid pIZ1016 using the HindIII and SpeI restriction sites, rendering the plasmid pIZacdS (Table 1). A 1171-bp Ptac-acdS DNA fragment was PCR-amplified from pIZacdS plasmid using the oligonucleotides 5 PtacacdS/3 PtacacdS. The amplified fragment was cloned into wide-host-range plasmid pSEVA237 using the AscI restriction site, obtaining the plasmid pSEVA237acdS (Table 1).

Inoculation of Rice Seedlings with Bacteria
Dehulled rice seeds (O. sativa L. GLEVA) were surface sterilized by shaking for 30 min in 30 mL 1% (v/v) sodium hypochlorite. After rinsing them three times for 10 min in sterile water, the seeds were incubated in VM-ethanol for 48 h. The germ-free seedlings were selected for inoculation. The germination of the seeds was continued on humidified filter papers for 24 h prior to inoculation with gfp-expressing bacteria cells. The gfp-expressing Azoarcus sp. CIB (pSEVA237) or E. coli CC118 (pSEVA237) bacteria were obtained as described previously [30]. After growing up to the mid-exponential phase, bacterial cells were collected by centrifugation, washed with sterile 0.9% NaCl (w/v) solution, resuspended in 1 mL of sterile distilled water, and the cell suspension was inoculated onto the surface of each seedling in aseptic conditions. After inoculation, the seedlings were grown at 25 • C under natural daylight conditions (10 h of light and 14 h of darkness) in sterile tubes (12 mL volume and one seedling each tube) for 5-10 days.

Examination of the Growth of Inoculated Rice Plants in Climate Chambers
Dehulled surface sterilized rice seeds (O. sativa L. GLEVA) were obtained by shaking them at 25 • C for 30 min in 30 mL 1% (v/v) sodium hypochlorite. Seeds were inoculated with the corresponding strain, as described before. The inoculated seeds were sown in square pots (2 seeds per pot) of 72 × 72 × 100 mm (Kisker Biotech, Steinfurt, Germany) with autoclaved vermiculite (Soprema Iberia S.R.U., Madrid, Spain; grain size of size 0.5-3 mm) as solid substrate. The bacteria (A 600 of 0.6) were additionally inoculated into the liquid Hoagland's No.2 solution (Sigma-Aldrich, San Louis, MS, USA) used to water plants. As a control, non-inoculated seeds were germinated and cultivated in the same conditions. The plants were kept in the growth chamber under the following conditions: photoperiod of 18 h light (day) and 8 h of darkness (night); the temperature of 25 • C during the day and 22 • C during the night; humidity of 75% during the day and 80% during the night. At harvest time, the fresh weight of each plant was determined. Each experiment was performed by triplicate and the total number of plants used was 20.

Examination of the Growth of Inoculated Rice Plants under Greenhouse Conditions
Plants were inoculated with bacteria as described above. Twenty-four hours from the inoculation, the seedlings were introduced in test tubes (12 mL of volume) with semisolid (0.4% agar) Hoagland's No.2 basal salt mixture (Sigma-Aldrich) supplemented, when required, with 50 µM CdCl 2 . Plants were watered every 5 days with sterilized Hoagland's No.2 basal salt mixture. When required, 50 µM CdCl 2 and bacterial suspensions (A 600 = 0.6) were added to the irrigation solution. The growing plants were kept in a greenhouse at 30 • C for 15 days. After 15 days, root and shoot samples of 50-100 mg each were taken for metal extractions. The determined dry weights of these samples were used to calculate average root and leaf Cd concentrations per g dry weight. Each experiment was performed in triplicate and the number of plants used was 5.

ACC Deaminase Enzymatic Assay
The ACC deaminase activity was measured spectrophotometrically following a protocol based on the accumulation of α-ketobutyrate [53], with slight modifications. The bacterial culture was grown until the mid-exponential phase, the cells were gathered by centrifugation and the pellet was resuspended in DF medium supplemented with 3 mM 1-aminocyclopropane-1-carboxylic acid (ACC) as the sole nitrogen source. After 24 h incubation at 30 • C, the cells were pelleted and resuspended in 1.5 mL Tris-HCl pH 7.0 buffer. After one wash of the cells with the same buffer, they were lysed by adding 30 µL of toluene. For the enzymatic assay, 200 µL of cell lysate was incubated 15 min at 30 • C with 20 µL ACC 0.5 M. The reaction was stopped by the addition of 1 mL HCl 0.56 M. The mix was centrifuged for 5 min at 20,000× g, and 1 mL supernatants were incubated at 30 • C with 800 µL HCl 0.56 M and 300 µL 2,4-dinitrophenylhydrazine. After 30 min of incubation, 2 mL NaOH 2M were added, and the A 540 was measured in a Shimadzu UV-260 spectrophotometer, Kioto, Japan. The α-ketobutyrate produced was determined using a calibration curve built with known concentrations of α-ketobutyrate. The protein concentration of the cell extracts was determined using the Bradford method [54].

Cadmium Concentration in Plants Tissues
Root samples were washed for 15 min with 10 mM Pb(NO 3 ) 2 and immersed for another 15 min in water to exchange surface-bound elements. Leaves were washed twice with Millipore water. Samples were dried at 105 • C for 1 week, weighted again to determine dry weight, and digested for 3 rounds in HNO 3 suprapur (70-71%) using a heat block (VWR, Leuven, Belgium). For the fourth digestion cycle, the sample was digested in 1 mL HCl suprapur (37%). Finally, the samples were dissolved in 500 µL of 20% HCl, and completed till 5 mL with Millipore water. The metal concentration was determined by inductively coupled plasma atomic emission spectrometry (Perkin-Elmer, 1100B, Waltham, MA, USA). Blanks (only HNO 3 ) and a standard sample (NIST Spinach (1570a)) were analyzed for reference purposes.

Superoxide Dismutase (SOD) Enzymatic Assay
SOD activity was estimated according to McCord and Fridovich, 1969. Root samples (50 mg) were collected and immediately frozen in liquid nitrogen and stored at −80 • C. The frozen tissue was homogenized in 0.1 M Tris-HCl buffer pH 7.8, 1 mM EDTA, 1 mM DTT and 4% polyvinylpyrrolidone. After homogenization, the samples were centrifuged for 10 min at 20,000× g and 4 • C. The sample solution contained: 50 mM KH 2 PO 4 pH 7.0 buffer, 1 mM EDTA, 0.1 mM cytochrome c, 0.5 mM xanthine and 5 U/mL xanthine oxidase, supplemented with 100 µL homogenized rice roots, or 100 µL of homogenization buffer (standard solution). The inhibition of the increment in the A 550 measured in the sample solution, compared to the standard, was used to estimate the SOD enzyme activity. One unit of enzyme activity was defined as the quantity of SOD necessary to inhibit 50% of the cytochrome c reduction [55].

Statistical Analysis
The data were analysed with the program IBM SPSS statistics 24 using analysis of the variance ANOVA. One-way or two-way ANOVA were applied to determine significant differences between the mean obtained in the assays performed. The post hoc Bonferroni test was employed to compare the different conditions. Statistical differences were presented as p < 0.001 (***), p < 0.01 (**) or p < 0.05 (*).

Azoarcus sp. CIB Is Able to Promote Plant Growth
We investigated whether the CIB strain was able to effectively promote the growth of rice plants as endophyte. To do so, rice seeds were germinated and inoculated with Azoarcus sp. CIB (pSEVA237) ( Table 1), using E. coli CC118 (pSEVA237) as a non-endophyte reference bacterium (Table 1). Plants inoculated with Azoarcus sp. CIB showed a 30% increment in weight compared to those that were not inoculated ( Figure 1). Since the weight of plants inoculated with E. coli CC118 did not increase compared to that of non-inoculated plants (Figure 1), we could exclude that the Azoarcus sp. CIB-mediated growth promotion was the consequence of a bacterial fertilizer effect. It has been previously reported that Azoarcus sp. CIB is able to colonize rice roots as an endophyte, showing several PGP traits such as nitrogen fixation, production of indoleacetic acid, and solubilization of phosphate [30]. However, it is known that only a limited number of bacteria demonstrating in vitro PGP traits are effectively able to promote plant growth [56]. Here, our results confirm that Azoarcus sp. CIB can promote the growth of rice plants, hence acting as a PGP-bacterium. In this sense, Azoarcus sp. CIB, that is likely a member of the new Aromatoleum genus [31], behaves as typical Azoarcus strains that colonize plants as endophytes and promote plant growth giving rise to a 50% rice weight increase, e.g., Azoarcus sp. DQS-4 [23], or 15% weight increase, e.g., Azoarcus sp. BH72 [25,28]. Further studies are in progress to determine the factors and mechanisms involved in the weight increase of rice plants after inoculation with the CIB strain.
confirm that Azoarcus sp. CIB can promote the growth of rice plants, hence acting as a PGP-bacterium. In this sense, Azoarcus sp. CIB, that is likely a member of the new Aromatoleum genus [31], behaves as typical Azoarcus strains that colonize plants as endophytes and promote plant growth giving rise to a 50% rice weight increase, e.g., Azoarcus sp. DQS-4 [23], or 15% weight increase, e.g., Azoarcus sp. BH72 [25,28]. Further studies are in progress to determine the factors and mechanisms involved in the weight increase of rice plants after inoculation with the CIB strain.

Engineering an ACC Deaminase-Producing Azoarcus sp. CIB Strain
To engineer a CIB strain harbouring ACC deaminase activity, the acdS gene from Paraburkholderia phytofirmans (formerly Burkholderia phytofirmans PsJN) was cloned under control of the heterologous Ptac promoter generating the promiscuous plasmid pSEVA237acdS ( Table 1). The new recombinant Azoarcus sp. CIB (pSEVA237acdS) strain did not show any significant difference in growth rate in the culture medium compared with that of the wild type CIB strain (data not shown), suggesting that expression of acdS did not reduce bacterial fitness. However, whereas the wild-type CIB strain did not exhibit detectable ACC deaminase activity, the recombinant Azoarcus sp. CIB (pSEVA237acdS) strain displayed a significant deaminase activity (6716 nmol mg −1 h −1 ), almost two times higher than that observed in the parental P. phytofirmans PsJN strain (3716 nmol mg −1 h −1 ) ( Figure 2). This high ACC deaminase activity in Azoarcus sp. CIB (pSEVA237acdS) could be due to the fact that the acdS gene is in multicopy doses and expressed under the strong Ptac promoter. The ACC deaminase activity level is in the range of that previously reported for Rhodococcus sp. strain 4N-4 (12,970 nmol mg −1 h −1 ), Rhodococcus sp. strain Hp2 (7320 nmol mg −1 h −1 ), Variovorax paradoxus 5C-2 (4322 nmol mg −1 h −1 ) or Acidovorax facilis 4p-6 (3080 nmol mg −1 h −1 ) strains [57], but much higher than that reported for most microorganisms equipped with ACC deaminase enzyme [6].
Endophytic bacteria that promote plant growth often harbour the acdS gene [58], which codes for an ACC deaminase that can lead to lower levels of the plant growth-inhibiting ethylene [11]. Interestingly, none of the Azoarcus strains described so far, including Azoarcus sp. CIB, possess the acdS gene [13,59]. Therefore, it could be interesting to examine the performance of an Azoarcus/Aromatoleum strain expressing an exogenous acdS gene. The heterologous expression of ACC deaminase genes and their biological effects have been reported previously in some bacteria. For example, the impact of ACC deaminase overproduction on plant growth and enhancement of nodulation abilities by Cupriavidus taiwanensis STM894 [60], Serratia grimesii BXF1 [61] and Mesorhizobium ciceri strain LMS-1 [19,62] has been described. Also, it has been reported the positive effects on plant growth promotion and plant copper tolerance of a genetically engineered Sinorhizobium meliloti strain overproducing ACC deaminase [20] have been reported, as well as the enhancement of the symbiotic performance of Mesorhizobium ciceri strains overproducing ACC deaminase when they were used as inoculants of chickpea plants cultivated under saline conditions, thus alleviating the negative effects caused by salinity [21]. However, to the best of our knowledge, our results evidence the first example of successful suggesting that expression of acdS did not reduce bacterial fitness. However, whereas the wild-type CIB strain did not exhibit detectable ACC deaminase activity, the recombinant Azoarcus sp. CIB (pSEVA237acdS) strain displayed a significant deaminase activity (6716 nmol mg −1 h −1 ), almost two times higher than that observed in the parental P. phytofirmans PsJN strain (3716 nmol mg −1 h −1 ) (Figure 2). This high ACC deaminase activity in Azoarcus sp. CIB (pSEVA237acdS) could be due to the fact that the acdS gene is in multicopy doses and expressed under the strong Ptac promoter. The ACC deaminase activity level is in the range of that previously reported for Rhodococcus sp. strain 4N-4 (12,970 nmol mg −1 h −1 ), Rhodococcus sp. strain Hp2 (7320 nmol mg −1 h −1 ), Variovorax paradoxus 5C-2 (4322 nmol mg −1 h −1 ) or Acidovorax facilis 4p-6 (3080 nmol mg −1 h −1 ) strains [57], but much higher than that reported for most microorganisms equipped with ACC deaminase enzyme [6]. Endophytic bacteria that promote plant growth often harbour the acdS gene [58], which codes for an ACC deaminase that can lead to lower levels of the plant growth-inhibiting ethylene [11]. Interestingly, none of the Azoarcus strains described so far, including Azoarcus sp. CIB, possess the acdS gene [13,59]. Therefore, it could be interesting to examine the performance of an Azoarcus/Aromatoleum strain expressing an exogenous acdS gene. The heterologous expression of ACC deaminase genes and their biological effects have been reported previously in some bacteria. CIB (yellow column), P. phytofirmans PsJN (brown column) and Azoarcus sp. CIB (pSEVA237acdS) (dark grey column) was determined according to a colorimetric method described in Section 2. The intensity of the color varies from the yellow (negative) to dark brown (positive). The spectrophotometric method determines the presence of α-ketobutyrate (nmol per mg of total protein −1 h −1 ) produced from ACC. Means of three independent experiments are presented and the standard deviation is indicated as error bars.

Azoarcus sp. CIB-Expressing acdS Gene Promotes Rice Growth under Stress Conditions
To check whether the expression of the acdS gene in Azoarcus sp. CIB confers an additional PGP trait to this recombinant bacterium, we took advantage of the ability of strain CIB to tolerate moderate concentrations of the toxic metal Cd (up to 1 mM CdCl 2 ) without significant effects on cell growth [36]. As expected, non-inoculated rice seedlings exposed to CdCl 2 (50 µM) showed a significant weight reduction ( Figure 3A). However, the weight of Cd-exposed plants inoculated with Azoarcus sp. CIB (pSEVA237acdS) (0.23 ± 0.03 g) was 65 ± 3% higher compared to that of non-inoculated control plants or plants inoculated with the CIB (pSEVA237) control strain (0.15 ± 0.02 g) ( Figure 3B). Thus, a protective effect was obvious in case of Cd-exposed plants inoculated with Azoarcus sp. CIB (pSEVA237acdS) expressing the ACC deaminase gene. Hence, these results demonstrated that the expression of the acdS gene in Azoarcus sp. CIB (pSEVA237acdS) colonizing rice roots may protect the plant against the toxic effects of Cd exposure.
Endophytic bacteria expressing ACC deaminase activity have been associated with the protection of the plants growing in hostile environments that provoke ethylene-induced stress [6], and the relationship between metals and ethylene has been well established in transgenic plants expressing a bacterial acdS gene [2,63]. Moreover, it has been described previously that some bacteria with ACC deaminase activity can promote plant growth under metal stress [64] or drought stress [65]. For example, the protective effects of Burkholderia sp. J62 on Pb (lead)-and Cd -exposed Lycopersicon esculentum [66], Pseudomonas putida UW4 on Ni (nickel)-exposed Brassica napus [67], Pseudomonas putida UW4 on Cu (copper)-exposed Medicago lupulina [20], Variovorax paradoxus on Cd-exposed Brassica juncera [57] have been reported. Concerning rice, it has been shown that plants inoculated with mutant bacteria lacking the acdS gene and grown in soil polluted with zinc were not able to develop normally, in contrast to plants inoculated with wild-type bacteria expressing the acdS gene [68]. As mentioned above, Cd is a very harmful metal, even at low concentrations, that can accumulate in soils leading to phytotoxic effects and hence decrease agricultural production [69]. It is well-known that Cd provokes stress in plants by increasing levels of reactive oxygen species (ROS) and ethylene production [9,70,71]. In fact, under laboratory conditions, rice seedlings were unable to grow when exposed to 500 µM CdCl 2 , and a significant (about 50%) reduction of growth was observed at 50 µM ( Figure 3A). Therefore, to the best of our knowledge, this is the first report about PGPB-expressing ACC deaminase protecting rice plants to stress induced by Cd exposure.
To check whether the expression of the acdS gene in Azoarcus sp. CIB confers an additional PGP trait to this recombinant bacterium, we took advantage of the ability of strain CIB to tolerate moderate concentrations of the toxic metal Cd (up to 1 mM CdCl2) without significant effects on cell growth [36]. As expected, non-inoculated rice seedlings exposed to CdCl2 (50 µM) showed a significant weight reduction ( Figure 3A). However, the weight of Cd-exposed plants inoculated with Azoarcus sp. CIB (pSEVA237acdS) (0.23 ± 0.03 g) was 65 ± 3% higher compared to that of non-inoculated control plants or plants inoculated with the CIB (pSEVA237) control strain (0.15 ± 0.02 g) ( Figure 3B). Thus, a protective effect was obvious in case of Cd-exposed plants inoculated with Azoarcus sp. CIB (pSEVA237acdS) expressing the ACC deaminase gene. Hence, these results demonstrated that the expression of the acdS gene in Azoarcus sp. CIB (pSEVA237acdS) colonizing rice roots may protect the plant against the toxic effects of Cd exposure.  Presented are mean values for 20 plants with the corresponding errors. The asterisks indicate the statistical significance observed when two conditions were compared applying the Two-ways ANOVA method and the Bonferroni test. p < 0.001 (***); p < 0.01 (**).

Enhanced Cd Concentration in the Shoots by Strain CIB Expressing the acdS Gene
To determine the effects of Azoarcus sp. CIB inoculation on Cd concentrations and its allocation in rice, the total metal concentrations in roots and shoots were determined. No significant differences in root Cd concentrations were observed when comparing treatments with or without bacterial inoculations ( Figure 4A).
However, the shoots of plants inoculated with Azoarcus sp. CIB (pSEVA237acdS) showed significantly higher Cd concentrations (38.24 ± 6) than the non-inoculated Cd-exposed controls (19.3 ± 1.4) ( Figure 4B). Interestingly, the Azoarcus sp. CIB (pSEVA237) strain, which does not express the acdS gene, also increased the metal concentration in the shoots compared to the non-inoculated control, leading to Cd concentrations that were not significantly different from those obtained with the recombinant CIB (pSEVA237acdS) strain. Hence, independent of the acdS gene expression, Azoarcus CIB can alter metal concentrations in tissues, which is an interesting finding for phytoextraction applications. Although the leaf Cd concentrations in the range of 15-40 µg g −1 are toxic to rice [72], as seen by the plant growth inhibition (Figure 3), Azoarcus sp. CIB (pSEVA237acdS) inoculated plants were growing better (Figure 3), which suggests that for the same concentration of metals in the leaves, the plants inoculated with CIB expressing acdS can cope better with the metals. grow when exposed to 500 µM CdCl2, and a significant (about 50%) reduction of growth was observed at 50 µM ( Figure 3A). Therefore, to the best of our knowledge, this is the first report about PGPB-expressing ACC deaminase protecting rice plants to stress induced by Cd exposure.

Enhanced Cd Concentration in the Shoots by Strain CIB Expressing the acdS Gene
To determine the effects of Azoarcus sp. CIB inoculation on Cd concentrations and its allocation in rice, the total metal concentrations in roots and shoots were determined. No significant differences in root Cd concentrations were observed when comparing treatments with or without bacterial inoculations ( Figure 4A). It has been described that rhizosphere residing bacteria equipped with ACC deaminase activity are able to enhance the uptake of inorganic pollutants through modification of the root architecture and the root uptake system of the plant, increasing the accumulation of the metal in plant tissues [57,73]. In addition, other mechanisms by which PGPB can influence Cd-uptake have been extensively reviewed [74] and include for example, microbe induced changes to the rhizosphere pH, Cd solubilisation, organic acid secretion, by which Cd can be more readily available in the rhizosphere for passive or active uptake by root cells [74,75]. Similar mechanisms could explain the effect of the endophytic Azoarcus sp. CIB strain expressing the exogenous acdS gene using rice as proof of concept of phytoextraction and survival at high concentrations of Cd. Although rice is not commonly recommended for cultivation on Cd-polluted paddy soils since Cd can be readily incorporated by plants [76], and easily found in rice grains [77] and incorporated in the food chain [78], some reports showed that specific varieties of rice, such as TCM213, can be possible candidates for effective Cd phytoextraction [41]. The use of rice has been especially suggested for paddy soils, environments that are more adapted for rice cultivation [42,43]. In this sense, it has also been also described that in soil polluted with zinc, some endophytes were able to increase the rice growth and Zn concentration in rice tissues [69]. Finally, it is worth noting that plants growing in oil hydrocarbons polluted soils are often subjected to strong stress linked to the production of an excess of ethylene [79]. The ability of Azoarcus sp. CIB to degrade aromatic compounds, such as toluene and m-xylene, paves the way to employ the CIB strain expressing the acdS gene as a strategy to promote rice plant growth both by eliminating the aromatic pollutants from the paddy soils by controlling the ethylene levels in the plant.

Azoarcus sp. CIB Expressing acdS Gene May Increase the ROS Quenching Capacity in Rice Roots
To further explore the mechanisms behind the plant growth promotion by Azoarcus sp. CIB (pSEVA237acdS) cells, we estimated the levels of oxidative stress in the plants using the activity of superoxide dismutase (SOD) as an indicator of superoxide levels in roots. As expected, exposure of non-inoculated plants to 50 µM CdCl 2 caused an important increase of the SOD activity (66 ± 6%) ( Figure 5). However, when plants were inoculated with strain CIB (pSEVA237), SOD activity was lower, being the lowest when the strain used for inoculation was equipped with the acdS gene, i.e., Azoarcus sp. CIB (pSEVA237acdS) ( Figure 5). It is well known that plant responses to metal-exposure are directly linked to the ROS produced and the levels of ethylene [9]. The results presented here show that ACC deaminase activity is mitigating the SOD activity, which suggests a reduction of ROS levels in CIB (pSEVA237acdS) inoculated rice. These results are in agreement with previous work reporting decreases of the SOD activity in plants exposed to different concentrations of metals and aromatic compounds when inoculated with PGPB-expressing ACC deaminase genes [80,81]. Therefore, the growth-promoting effect observed with strain Azoarcus sp. CIB (pSEVA237acdS) under Cd exposure can, at least partly, be attributed to its modulating effects on the levels of ethylene produced by the plant and the reduction of ROS levels.

Conclusions
The results of the present study allow us to conclude that an aromatics-degrading and metal-resistant strain closely related to endophytic Azoarcus, i.e., Azoarcus sp. CIB, is able to produce a 30% plant weight increase when colonizing rice as an endophyte, hence revealing that it behaves also as a PGP-bacterium. In addition, the heterologous expression of an exogenous ACC deaminase gene in the CIB strain generates a PGP that increases the weight of rice plants growing in a Cd-polluted semi-solid medium, indicating its ability to protect rice against the Cd-induced stress, and increasing the concentration of Cd in rice tissues. To the best of our knowledge, this is the first report about the use of PGPB-expressing ACC deaminase to protect rice plants to the stress induced by Cd exposure. The pSEVA237acdS plasmid developed in this work represents a new genetic tool that could be implemented to provide an additional PGP trait in those PGPB that lack ACC Figure 5. Superoxide dismutase activity in rice roots. Rice plants were grown in Hoagland's medium in the absence (0 µM) or presence (50 µM) of CdCl 2 after inoculation with Azoarcus sp. CIB (pSEVA237) (CIB pS) (orange columns), Azoarcus sp. CIB (pSEVA237acdS) (CIB pSacdS) (red columns), or without bacterial inoculation (CT) (yellow column) in greenhouse conditions as detailed in Section 2. After harvesting, the superoxide dismutase activities were estimated in roots. Data show the mean of 10 independent measurements; the standard deviations are presented as error bars. Asterisks show the statistical significance after applying the two-way ANOVA analysis and the Bonferroni test. p < 0.01 (**).
The results presented here show that ACC deaminase activity is mitigating the SOD activity, which suggests a reduction of ROS levels in CIB (pSEVA237acdS) inoculated rice. These results are in agreement with previous work reporting decreases of the SOD activity in plants exposed to different concentrations of metals and aromatic compounds when inoculated with PGPB-expressing ACC deaminase genes [80,81]. Therefore, the growth-promoting effect observed with strain Azoarcus sp. CIB (pSEVA237acdS) under Cd exposure can, at least partly, be attributed to its modulating effects on the levels of ethylene produced by the plant and the reduction of ROS levels.

Conclusions
The results of the present study allow us to conclude that an aromatics-degrading and metal-resistant strain closely related to endophytic Azoarcus, i.e., Azoarcus sp. CIB, is able to produce a 30% plant weight increase when colonizing rice as an endophyte, hence revealing that it behaves also as a PGP-bacterium. In addition, the heterologous expression of an exogenous ACC deaminase gene in the CIB strain generates a PGP that increases the weight of rice plants growing in a Cd-polluted semi-solid medium, indicating its ability to protect rice against the Cd-induced stress, and increasing the concentration of Cd in rice tissues. To the best of our knowledge, this is the first report about the use of PGPB-expressing ACC deaminase to protect rice plants to the stress induced by Cd exposure. The pSEVA237acdS plasmid developed in this work represents a new genetic tool that could be implemented to provide an additional PGP trait in those PGPB that lack ACC deaminase activity, hence fostering plant production even under stress conditions such as cultivation in polluted soils. In addition, we open here a new possibility to explore the use of rice for phytoremediation of Cd-polluted paddy soils. Since Cd-polluted soils often also contain oil hydrocarbons, the application of Azoarcus sp. CIB (pSEVA237acdS) as a tool to improve the growth of rice, the phytoextraction of Cd and the removal of aromatic pollutants will be investigated.