Grey mould control by oxalate degradation using non-antifungal Pseudomonas abietaniphila strain ODB36

Grey mould is an important necrotrophic fungal pathogen that causes huge economic losses in agriculture. Many types of bacteria are used for biological control of grey mould via competition for space or nutrients and/or the production of antifungal metabolites. Oxalate is a key component of virulent necrotic fungal pathogens. In this study, we isolated non-antifungal oxalate-degrading bacteria (ODB) from the surfaces of oxalate-rich spinach and strawberries to investigate their ability to control necrotic fungal pathogens such as grey mould. A total of 36 bacteria grown on oxalate minimal (OM) agar plates were tested for oxalate-degrading activity. Five isolates exhibiting the highest oxalate degradation activity were subjected to molecular identification using 16S rRNA gene sequencing. Two isolates exhibiting non-antifungal activity were subjected to disease suppression assays using Arabidopsis–Botrytis systems. The isolate Pseudomonas abietaniphila ODB36, which exhibited significant plant protective ability, was finally selected for further investigation. Based on whole-genome information, the pseudomonad oxalate degrading (podA) gene, which encodes formyl-CoA transferase, was analysed. The podA− mutant did not inhibit Botrytis infection and oxalate toxicity; the defects were recovered by podA complementation. Purified PodA–His converted oxalate to formate and eliminated oxalate toxicity. These results indicate that P. abietaniphila ODB36 and PodA enzyme are associated with various aspects of grey mould disease inhibitory effects.


Results
oDB isolation and oxalate-degrading activity. A total of 36 bacteria grown on oxalate minimal (OM) agar plates (2 g Na 2 C 2 O 4 , 2.7 g K 2 HPO 4 , 0.9 g NaH 2 PO 4 , 0.9 g NH 4 Cl, 0.27 g MgSO 4 ·7H 2 O, 0.009 g CaCl 2 ·2H 2 O, and 0.0024 g FeSO 4 ·7H 2 O in 1 L distilled water with 1.5% agar) were isolated from the surfaces of spinach (27 isolates) and strawberries (9 isolates). The oxalate-degrading activity of these 36 bacteria was tested in OM broth supplemented with 1/10 Bacto tryptic soy broth (TSB). Oxalate in the culture supernatants was determined using an oxalate colorimetric assay kit (Bio Vision Inc., San Francisco, CA, USA). Therefore, the lower the absorbance value at 450 nm (A 450 ), the higher the degradation activity. When the non-inoculated media control (nc) supplemented with 15 mM sodium oxalate showed a value of 0.55, five isolates (ODB5, ODB29, ODB31, ODB35, and ODB36) exhibited significantly reduced values < 0.5 (Fig. 1a). Isolates with values higher than the control (nc) were suspected to have the ability to synthesize oxalate by themselves, and thus exhibited values > 0.55 (Fig. 1a). Each isolate showed a different growth pattern; in particular, the growth of ODB35 was relatively slow compared to other isolates (Fig. 1b). Because the growth of each isolate directly affects oxalate degradation, it is necessary to normalize values. Five isolates (ODB5, ODB29, ODB31, ODB35, and ODB36) had normalized values < 0.2, Figure 1. Oxalate degradation by bacterial isolates. (a) Absorbance measurement at A 450 . Oxalate in the supernatants was determined using an oxalate colorimetric assay kit. Bacterial isolates were grown in oxalate minimal (OM) medium supplemented with 1/10 TSB at 28 °C for 3 days. Cell-free supernatants were subjected to the sample preparation protocols described in Methods. The values report the amount of oxalate remaining in the OM medium. A total of 36 bacteria growing on OM agar plates were isolated from the surfaces of spinach (27 isolates; green) and strawberries (9 isolates; red). (b) Optical density (OD 600 ) of the bacterial growth. (c) Normalised values calculated by multiplying the A 450 value of the oxalate amount by the OD 600 of the culture. The lower the value, the higher the degradation activity. Bacterial isolates ODB5, 29, 31, 35, and 36 exhibited significant oxalate degradation. Values are averages of triplicate assays and error bars represent the range. nc denotes the non-inoculated medium control. Asterisks denote significant differences from the control (*p < 0.05; Student's t test).
indicating that they had considerable oxalate-degrading activity (Fig. 1c). Three isolates (ODB4, ODB17, and ODB21) that did not grow in liquid culture were excluded.

Bacterial identification and antifungal activity tests.
To confirm the identities of the ODB, we amplified and sequenced the 16S rRNA genes of ODB5, ODB29, ODB31, ODB35, and ODB36, which exhibited significant oxalate-degrading activity. BLAST analysis of the 16S rRNA gene sequence of ODB5 showed 99% identity with P. fluorescens; those of ODB29, ODB31, and ODB36 showed 99% identity with P. abietaniphila; and that of ODB35 showed 99% identity with Methylobacterium zatmanii.

Disease suppression assay.
To investigate disease suppression of ODB35 and ODB36, which exhibited no antifungal activity, we used the standard A. thaliana-B. cinerea (plant-pathogen) system. After 16 days of inoculation, water-sprayed plants exhibited a 65% disease rate against grey mould, whereas ODB36-sprayed plants exhibited significant disease suppression depending on the treatment concentration (Fig. 2). However, ODB35 exhibited no significant disease suppression compared with ODB36 (Fig. 2).
Phylogenetic analysis and sequence alignment. Using whole-genome information from GenBank (accession no. SDG16549), we analyzed a gene encoding an oxalate-degrading enzyme in P. abietaniphila ATCC 700689 17 . Translated amino acid sequences of PodA (42 kDa) showed 98% similarity with P. abietaniphila ATCC 700689 and 82% with P. fluorescens BBc6R8. Figure 3 shows the phylogenetic relationships of several putative formyl-CoA transferases from organisms whose protein sequences were available. As expected, the formyl-CoA transferases from P. abietaniphila and P. fluorescens clustered together. Sequence analysis showed that PodA, a formyl-CoA transferase, belongs to pfam02515, the CaiB-BaiF family of enzymes with diverse functions including fatty acid racemases, carnitine dehydratase (CaiB), bile acid inducible operon protein F (BaiF), benzylsuccinate CoA-transferase (BbsF), and anaerobic toluene catabolic protein in the presence of toluene. Identical residues, shown in red, are located in half of the N-termini of the proteins and partially in the C-termini (Fig. 4). (b) disease rate (%). Three-week-old Arabidopsis plants were sprayed with the ODB suspension. After incubation for 2 days, a B. cinerea spore suspension (5 × 10 5 spores/mL) was prepared using sterilised water. The spore suspension was sprayed until run-off occurred. Inoculated plants were placed in an opaque plastic box lined with saturated paper towels, the lids were removed after 2 days, and disease development was observed for 16 days and measured. ODB36 exhibited significant disease suppression depending on treatment concentration. Values are averages of triplicate assays; error bars represent the range. Asterisks denote significant differences from the water control (*p < 0.05; Student's t test). (2020) 10:1605 | https://doi.org/10.1038/s41598-020-58609-z www.nature.com/scientificreports www.nature.com/scientificreports/ PodA is required for the inhibition of Botrytis infection. To investigate the effect of the wild-type ODB36, podA − (podA-lacZY) mutant, and podA complementation strains on Botrytis infection, we performed disease suppression assays. Wild-type strain exhibited significant disease suppression. The podA − mutant did not inhibit Botrytis infection; the defect was recovered by podA complementation (Fig. 5). These results indicated that podA is critical for the inhibition of Botrytis infection in P. abietaniphila.
PodA is required for the inhibition of oxalate toxicity. In tobacco leaf disc assays, oxalate toxicity, such as bleaching of leaf discs, was observed for the podA − (podA-lacZY) mutant but was weak for the wild-type and podA complementation strains (Fig. 6a). This result was confirmed by the quantification of chlorophyll. Tobacco leaf discs treated with the podA − mutant showed bleaching or a low chlorophyll level. These defects were restored to near wild-type levels by podA complementation (Fig. 6b). Also, oxalate degradation was induced by treatment of the wild-type and podA complementation strains (Fig. 6c). overexpression, purification, and oxalate-degrading and formate conversion activities of podA protein. We overexpressed and purified P. abietaniphila ODB36 PodA from Escherichia coli BL21(DE3) carrying pLY205 (pET21b::podA). The podA gene encodes a protein of 396 amino acid residues with a calculated molecular weight of 42 kDa. To confirm that the podA gene encodes an oxalate-degrading enzyme, it was expressed under control of the T7 promoter in E. coli. Upon induction of E. coli BL21(DE3) harbouring pLY205 with IPTG, a His-tagged protein with a molecular mass of around 42 kDa was produced; approximately half of the PodA-His protein was expressed in soluble bodies at pH 8.0-9.0 (Fig. 7a). PodA-His was purified to near homogeneity using Ni-NTA affinity chromatography (Fig. 7b). The oxalate colorimetric assay kit revealed that PodA-His (0.8 mM) exhibited 75% oxalate-degrading activity (Fig. 7c). Using a formate colorimetric assay kit, we measured the formate conversion activity of PodA-His; the amount of degradation confirmed that oxalate was consistently converted to formate (Fig. 7c).

Inhibition of oxalate toxicity by PodA-His protein. To investigate the inhibitory effect of PodA-
His protein on oxalate toxicity, we used a standard A. thaliana-oxalate (plant-toxin) system. Oxalate-sprayed plants exhibited an 81% oxalate toxicity rate, while PodA-His-sprayed plants showed significant and concentration-dependent inhibition of oxalate toxicity (Fig. 8).

Discussion
Grey mould causes damage during the cultivation, storage, and transportation of fresh vegetables and soft-pulling fruits 1-3 . Efforts have been made to control grey mould using antifungal bacteria, but this is complicated by the stability of antimicrobial substances in fresh vegetables and the emergence of resistance 4,5 . For this reason, we investigated the possibility of suppressing grey mould using non-antifungal ODB, similar to the use of a non-antagonistic bacterium to control Salmonella enterica, which causes food poisoning 18 . www.nature.com/scientificreports www.nature.com/scientificreports/ In this study, we isolated ODB from the surfaces of spinach and strawberries and evaluated their disease control activity. We also explored potential correlations between oxalate-degrading ability and disease inhibitory effects of the ODB isolates. Previous studies have used antagonistic ODB 13 or ODB with unknown antagonistic effects 12 . Therefore, unlike these previous studies, we attempted to exclude antagonistic effects by excluding P. fluorescens ODB5, which exhibited antifungal activity; we selected bacterial strains M. zatmanii ODB35 and P. abietaniphila ODB36 for further testing of dose response to disease suppression. Significant suppression of P. abietaniphila ODB36 against grey mould is shown in Fig. 2. We expect that biological control of fungal pathogens can be achieved by neutralising pathogenic factors and inhibiting the occurrence of resistant individuals by applying a method that does not kill pathogens.
The disease suppression effect of ODB35, which has the highest oxalate-degrading ability, did not reach that of ODB36. This result is likely due to the slow growth rate of ODB35. As shown in Fig. 1, oxalate-degradation activity was calculated by multiplying the A 450 value by bacterial growth. ODB35 exhibited relatively slow growth, resulting in a high value. These results suggest that the growth rate of ODB on the plant surface is important to its efficacy as a disease control agent. The aim of this study was to evaluate the oxalate-degrading ability of bacteria in After incubation for 2 days, a B. cinerea spore suspension (5 × 10 5 spores/mL) was prepared using sterilised water. The spore suspension was sprayed until run-off occurred. Inoculated plants were placed in an opaque plastic box lined with saturated paper towels, the lids were removed after 2 days, and disease development was observed for 12 days and measured. Wild-type ODB36 strain exhibited significant disease suppression. The podA − mutant did not inhibit Botrytis infection; the defect was recovered by podA complementation. Values are averages of duplicate assays; error bars represent the range. Asterisks denote significant differences from the water control (*p < 0.05; Student's t test).   www.nature.com/scientificreports www.nature.com/scientificreports/ disease control; we finally selected ODB36 as the optimal agent among those examined. Growth on plant surfaces was not investigated in this study.
Interestingly, the wild-type and podA complementation strains eliminated the toxicity of 40 mM oxalate, such as bleaching of tobacco leaf discs, but the podA − mutant showed bleaching due to no oxalate toxicity. The remaining oxalate concentration in the wild-type and podA − complementation strains was lower than that of the podA − mutant strain. These data suggest that podA contributes to the degradation of oxalate. These results are in good agreement with the finding that grey mould control is by oxalate degradation in P. abietaniphila.
In this study, we did not quantify the formate concentration in ODB because P. abietaniphila strains contain several genes encoding formate dehydrogenase subunits, which catalyse the oxidation of formate to carbon dioxide. However, we performed protein overexpression and enzyme activity assays and confirmed that PodA-His converted oxalate to formate, which plays a role in the biosynthesis of many compounds in energetic metabolism and signal production related to stress in plants 23 [24][25][26] . In enzymatic analysis of oxalate degradation of bacteria, formyl-CoA transferase (EC 2.8.3.16) catalyzes the chemical reaction, which is demonstrated in Fig. 9 15,16 . The two substrates, oxalate and formyl-CoA, are converted into the two products oxalyl-CoA and formate. Formyl-CoA is required for the conversion of oxalate to oxalyl-CoA by formyl-CoA transferase 16 . However, the inhibitory effect of PodA on oxalate toxicity in Arabidopsis suggests the presence of formyl-CoA as a substrate. It is possibile that formyl-CoA derived from Arabidopsis-resident microorganisms acts as a substrate for PodA.
Previous studies have explored disease control in transgenic crops using plant-derived oxalate-degrading genetic resources. Barley oxalate oxidase transgenic peanuts showed enhanced resistance against Sclerotinia minor 27 . Soybean plants expressing wheat oxalate oxidase were resistant to Sclerotinia sclerotiorum 28,29 . These preliminary studies suggest that developed oxalate degradative genetic resources can be introduced into crops and inspire the development of transgenic plants with related abilities.
In conclusion, to overcome the competing processes of biological control and occurrence of resistant species, we investigated non-antimicrobial genetic resources for biological control of plant pathogens. Oxalate, a key pathogenic factor of necrotrophic fungi, was targeted to disarm, but not kill, plant pathogens. Oxalate-degrading bacteria were isolated and identified, and their activity was evaluated. Using disease suppression assays, www.nature.com/scientificreports www.nature.com/scientificreports/ P. abietaniphila ODB36 was finally selected for further investigation. PodA from ODB36, which encodes formyl-CoA transferase, was overexpressed and assayed using oxalate degradation tests. Bacterial cells of both ODB36 and enzyme PodA were found to be directly applicable in the field to control grey mould. The podA gene can also be applied for the development of transgenic resistant cultivars against grey mould. Methods isolation of oDB. Surfaces of strawberries and spinach were swabbed with sterilised swabs, which were then streaked onto OM medium with 1.5% agar and incubated at 28 °C. After 7 days, single colonies were re-streaked onto OM agar medium for purification.
Bacterial oxalate-degrading activity. Oxalate-degrading activity was evaluated using an oxalate colorimetric assay kit according to the supplier's protocol. Optical density (OD 600 ) was measured in bacterial cultures grown in OM broth plus 1/10 TSB at 28 °C for 3 days. After centrifugation, the supernatants were transferred into 96-well plates and mixed with 2 μL oxalate converter, which was supplied with the kit. The mixtures were incubated at 37 °C for 1 h, and 50 μL prepared reaction mix (46 μL buffer; 2 μL oxalate enzyme mix, and 2 μL oxalate probe; all reagents were supplied with the kit) was added to each well containing standard or samples. Then the plates were incubated at 37 °C for 1 h and absorbance was measured at 450 nm (A 450 ). The values reporting the amount of oxalate remaining in the sample were normalized to culture growth by multiplying the A 450 value for the non-degraded oxalate amount by the OD 600 of the culture. The lower the value, the higher the degradation activity.
16S rRNA gene sequencing. To confirm the identities of the ODB, the 16S rRNA gene was amplified and sequenced using the primers 27mF (5′-AGAGTTTGATCMTGGCTCAG-3′) and 1429mR (5′-GGYTACCTTGTTACGACTT-3′). Total DNA was extracted using the Wizard Genomic DNA Purification Kit (Promega, Madison, WI, USA) following the manufacturer's instructions. Polymerase chain reaction (PCR) was performed using a T100 thermal cycler (Bio-Rad, Hercules, CA, USA) with PCR polymerase (AccurPower PCR Premix; Bioneer, Daejeon, South Korea), 1 μL target DNA, and 1 mM each primer at 98 °C for 2 min, followed by 30 cycles of denaturation at 98 °C for 30 s, annealing at 55 °C for 30 s, and extension at 70 °C for 1 min, followed by a final extension at 72 °C for 4 min. The amplified products were separated by electrophoresis in 0.8% (w/v) agarose gels. PCR amplification yielded a single visible DNA product, whose band was cleaved from the ethidium bromide (EtBr)-stained gel and purified using a 200-p Expin Gel SV kit (GeneAll biotechnology, Seoul, Korea), following the manufacturer's instructions. Purified PCR products were sequenced by Macrogen Services (Daejeon, Korea) in both directions using previously described primers 30 . DNA sequences were analyzed using the BLASTn program. DNA sequences of the 16S rRNA gene were compared with those in the National Center for Biotechnology Information (NCBI) GenBank database (http://www.ncbi.nlm.nih.gov/blast/). Antifungal activity test. Among the ODB, the antifungal activity of the three isolates (ODB5, ODB35, and ODB36) showing the highest oxalate-degrading activity was tested against B. cinerea, A. alternata, and S. cerevisiae. Botrytis cinerea and A. alternata spores were spread on half potato dextrose agar with half protease peptone (PDP) and 1.5% agar. Saccharomyces cerevisiae cells were embedded in PDP agar. The ODB suspension (10 μL) was dropped onto the plate, which was then incubated at 28 °C for 48 h and the magnitude of inhibition of fungal growth was monitored.
Disease suppression assay. Arabidopsis thaliana Columbia (Col-0) was grown in a growth chamber at 20 °C under a 16-h photoperiod. Bacterial suspensions were prepared by resuspending the strains using sterilised water to adjust the OD 600 values to the corresponding ranges (0.7-1.0). We sprayed 3-week-old Arabidopsis plants with the ODB suspensions (5 mL/plant). After 2 days of incubation, a B. cinerea spore suspension (5 × 10 5 spores/ mL) was prepared using sterilised water. The spore suspension was sprayed until run-off. Inoculated plants were placed in an opaque plastic box lined with saturated paper towels, the lids were removed after 2 days, and disease development was observed for 16 days and measured according to the following calculation: (no. diseased leaves/ no. inoculated leaves) × 100.
Sequence alignment and phylogenetic analysis. Protein sequences obtained from NCBI (https:// www.ncbi.nlm.nih.gov/) were aligned and utilized to generate an unrooted phylogenetic tree using the neighbor-joining method (CLUSTALX software). construction of the podA − mutant and podA complementation strains. To generate the podA − mutant, plasmid pLY201, carrying the internal fragment of the podA gene was constructed using ODB36 genomic DNA as the PCR template and PodAE (5′-AAGATACTGGGTGAGTTT-3′) and PodAK (5′-GGTACCCATCACCAGTTTCGGATT-3′) as primers. The amplified region (291 bp) was purified from an agarose gel and ligated into the pGEM-T Easy Vector System (Promega, Mannheim, Germany) to generate pLY201, which was confirmed by sequencing. pLY201 fragments generated by digestion with the restriction enzymes EcoRI and KpnI (TaKaRa Bio Inc., Kusatsu, Japan) were purified after electrophoresis from an agarose gel and inserted into the suicide vector pVIK112 31 , generating pLY207. The resulting construct, pLY207, was transferred into E. coli S17-1 λpir and introduced into P. abietaniphila ODB36 by conjugation, generating podA-lacZY. The lacZY reporter gene fusion insertion mutants were selected based on the kanamycin-resistance phenotype and confirmed by PCR with primers that annealed upstream of the truncated fragments of podA, PodPro (5′-ATGCAAAAGCCTTTGCAAGGG-3′) and LacFuse (5′-GGGGATGTGCTGCAAGGCG-3′).
To construct the podA complementation strain, wild-type coding sequences with Plac were cloned into the broad-host vector pLAFR3. Coding sequences were amplified by PCR from ODB36 genomic DNA, using the Oxalate degradation and formate conversion activities of PodA-His. The oxalate-degrading and formate conversion activities of purified PodA-His were evaluated using an oxalate colorimetric assay kit and a formate colorimetric assay kit, respectively, according to the supplier's protocol (BioVision Inc., Milpitas, CA, USA).

Inhibition of oxalate toxicity by PodA-His protein.
Arabidopsis thaliana Columbia (Col-0) was grown in a growth chamber at 20 °C under a 16-h photoperiod. Protein suspensions were prepared by resuspending PodA-His in potassium phosphate buffer (pH 8.5) to 1, 2, and 4 mg/mL. We sprayed 6-week-old Arabidopsis plants with PodA-His protein suspensions (5 mL/plant). After incubation for1 day, the plants were sprayed with a sodium oxalate suspension (20 mM; 1-mL/plant). Oxalate toxicity was observed for 5 days and calculated according to the following equation: (no. of bleached leaves/total no. of leaves) × 100.