Cloning and Molecular Characterization of CmOxdc3 Coding for Oxalate Decarboxylase in the Mycoparasite Coniothyrium minitans

Coniothyrium minitans (Cm) is a mycoparasitic fungus of Sclerotinia sclerotiorum (Ss), the causal agent of Sclerotinia stem rot of oilseed rape. Ss can produce oxalic acid (OA) as a phytotoxin, whereas Cm can degrade OA, thereby nullifying the toxic effect of OA. Two oxalate decarboxylase (OxDC)-coding genes, CmOxdc1 and CmOxdc2, were cloned, and only CmOxdc1 was found to be partially responsible for OA degradation, implying that other OA-degrading genes may exist in Cm. This study cloned a novel OxDC gene (CmOxdc3) in Cm and its OA-degrading function was characterized by disruption and complementation of CmOxdc3. Sequence analysis indicated that, unlike CmOxdc1, CmOxdc3 does not have the signal peptide sequence, implying that CmOxDC3 may have no secretory capability. Quantitative RT-PCR showed that CmOxdc3 was up-regulated in the presence of OA, malonic acid and hydrochloric acid. Deletion of CmOxdc3 resulted in reduced capability to parasitize sclerotia of Ss. The polypeptide (CmOxDC3) encoded by CmOxdc3 was localized in cytoplasm and gathered in vacuoles in response to the extracellular OA. Taken together, our results demonstrated that CmOxdc3 is a novel gene responsible for OA degradation, which may work in a synergistic manner with CmOxdc1.


Introduction
Oxalic acid (OA) is a natural organic acid with a low molecular weight. It is characterized by strong acidity, reducibility and calcium-chelating capability [1]. OA can affect human health, causing urolithiasis, hyperoxaluria, renal failure and other diseases [2]. Furthermore, OA is a typical anti-nutrient, it can reduce solubility and utilization of minerals and trace elements, such as Ca 2+ , resulting in nutritional imbalance [3]. Many plants, including purslane, spinach and tea, as well as many fungi, such as Sclerotinia sclerotiorum, Aspergillus niger and Ceriporiopsis subvermispora can produce OA [4][5][6][7].
OA secretion by fungi produces many benefits for mycelial growth and substrate (niche) colonization. It is well-recognized that plant pathogenic fungi utilize OA as a virulence factor or a non-specific phytotoxin in interaction with plants with the following mechanisms [8][9][10]: (i) acidification of plant tissues by OA, thereby improving activity of the cell wall-degrading enzymes; (ii) disequilibration of the redox balance in plant tissues due to the strong reducibility of OA; (iii) stabilization of Ca 2+ by chelation, thereby weakening plant resistance; and (iv) timely inhibition of autophagy and apoptosis. In a word, OA is the key factor regulating pathogenicity of many fungal pathogens, including S. sclerotiorum.

Gene Cloning and Analysis
The genomic DNA (gDNA) and complementary DNA (cDNA) of C. minitans were used as templates to amplify CmOxdc3 as well as CmOxdc1 with the specific primers (Table S1) designed based on the genome of C. minitans [39,40]. Amino acid sequences encoded by CmOxdc3 and other four OxDC genes (CmOxdc1, CmOxdc2, FvOxdc, BsOxdc) were aligned using MEGA 7.0.26 [16,17,41]. The amino acid sequences encoded by 39 OxDC genes from eighteen fungi were used to construct a phylogenetic tree using the neighbor-joining method in MEGA 7.0.26, and a bootstrap test was carried out with 1000 replicates (Table S2). Prediction of the conserved domains and the signal peptides in CmOxdc3 and other OxDC genes were conducted using SMART (Simple Modular Architecture Research Tool) and SignalP 5.0, respectively [42,43].

DNA Extraction and Southern Blotting
Conidial suspension (1 × 10 7 conidia mL −1 ) of C. minitans was inoculated on autoclaved cellophane membranes placed on PDA and the cultures were incubated at 20 • C in dark for 3 d. The mycelial mass was harvested for extraction of gDNA using the CTAB method (CTAB = cetyltrimethylammonium bromide) [44]. Southern blot analysis was done to confirm disruption and complementation of genes in mutants of C. minitans using the Gene Images AlkPhos Direct Labeling and Detection System from GE Healthcare (Amersham Biosciences, Buckinghamshire, UK). The gDNA of the wild type was completely digested with Nco I, and the gDNAs of the disruption and complementation mutants were completely digested with Kpn I and BamH I. The DNA sequence at the left flank of CmOxdc3 was obtained by PCR and labeled as the probe.

RNA Extraction and Quantification of Gene Expression
Total RNA was extracted from the mycelial masses of the wild type and mutants of C. minitans using E.Z.N.A. Fungal RNA Kit (TaKaRa, Dalian, China) following the manufacturer's instructions. PrimeScript TM RT reagent Kit with gDNA Eraser (TaKaRa) was used to synthesize cDNA, and TB Green TM Premix Ex Taq TM II (TaKaRa) was used for fluorescence quantitative PCR (RT-qPCR) with the primers listed in Table S1. The actin gene (Cmactin) was chosen as internal control [16]. The relative expression level of each gene was calculated by using the 2 −∆∆Ct method [45].

Gene Expression Analysis of CmOxdc3
The expression pattern of CmOxdc3, as well as CmOxdc1 in the wild type (ZS-1), was detected by RT-qPCR with the RNA from the mycelial masses in the following assays: (i) the OA assay, 3-day-old cultures (150 rpm, 20 • C) of ZS-1 in PDB followed with 1-h treatment by amended with OA at 0 (control), 12, 24 or 32 mM; (ii) the acid assay, 3-day-old PDB cultures of ZS-1 (150 rpm, 20 • C) followed with 1-h treatment under pH 3 with OA (12 mM), acetic acid (75 mM), citric acid (50 mM), fumaric acid (50 mM), HCl (6 mM), lactic acid (50 mM), maleic acid (10 mM) or malonic acid (20 mM); and (iii) the pH assay, 2-h treatment of the 3-day-old PDA cultures of ZS-1 on cellophane films (20 • C) under pH 3, 4, 5 or 6 adjusted with 0.1 M citric acid-sodium phosphate buffer, the mycelial mass without extra pH treatment was used as control. There were three replicates for each treatment, and the entire experiment was repeated three times.

Gene Knockout and Complementation
CmOxdc3, as well as CmOxdc1, were separately disrupted using the split marker system [46] as shown in Figure S1A. For disruption of CmOxdc3, the 5 -flank region (772 bp) and the 3 -flank region (810 bp) of CmOxdc3 were cloned and fused with the partial neomycin resistance gene (NEO). The two fused fragments, CmOxdc3-5 -NE (1614 bp) and EO-CmOxdc3-3 (1816 bp), were then simultaneously transformed into the protoplasts of ZS-1. For disruption of CmOxdc1, the 5 -flank region (875 bp) and the 3 -flank region (843 bp) of CmOxdc1 were cloned and fused with the partial hygromycin resistance gene (HYG).
The disrupted CmOxdc3 in the mutant ∆CmOxdc3-7 was complemented using the strategy shown in Figure S2. The 5 -flank region, together with the coding sequence of CmOxdc3, excluding the stop codon (2701 bp) and the enhanced green fluorescent protein gene (eGFP, 745 bp) from the plasmid pCHEG (Drs. Guogeng Yang and Daohong Jiang, unpublished), were separately amplified. They were ligated and inserted into the plasmid pSKTN [35] with the aid of Kpn I and Hind III to generate the intermediate plasmid pSKTN-Oxdc3-eGFP. Then, the hygromycin resistance gene (HYG) from pBluscrikp II KS1004 [16] was inserted into pSKTN-Oxdc3-eGFP at Xba I to generate the final plasmid (pSKTH-Oxdc3-eGFP), which was transformed into the protoplasts of ∆CmOxdc3-7 mediated by PEG [16], followed by protoplast regeneration on TB3 and selection of the complementary mutants. Finally, a complementary mutant ∆CmOxdc3-7C was obtained.
The transformants were screened on PDA amended with hygromycin B (50 µg mL −1 ) or G418 (25 µg mL −1 ) three times and verified by PCR ( Figure S1B) with the primers in Table S1 and/or by Southern blotting.

Determination of Sensitivity to OA
The mutants (∆CmOxdc1-1, ∆CmOxdc1-25, ∆CmOxdc3-2, ∆CmOxdc3-7, ∆CmOxdc3-7C, ∆CmOxdc1&3-13, ∆CmOxdc1&3-45) and the wild type (ZS-1) were separately incubated for 30 days (20 • C) on PDA amended with OA at 0 (control), 8, 16 or 24 mM in Petri dishes. There were three dishes (replicates) for each strain in the control and each OA treatment, and the entire experiment was repeated three times. Diameter of the colony in each dish was measured at 10-day and 30-day post inoculation (dpi). Sensitivity of a strain to OA at a given concentration was calculated based on the colony diameters of that strain in the control (0 mM OA) and each OA treatment [35].

Assaying OA Degradation and Mycoparasitic Activities
The mutants and the wild type (ZS-1) mentioned above were separately inoculated in 150 mL flasks, each containing 50 mL PDB amended with OA to the final concentrations of 0 (control), 12 or 24 mM. Aliquots of the conidial suspension (1 × 10 7 conidia mL −1 ) of each mutant or ZS-1 were inoculated in the media, 100 µL in each flask, 3 flasks for each mutant and ZS-1. The flasks were shake-incubated at 20 • C and 150 rpm for 12 days. The cultures were separately filtered, and the cultural filtrates were loaded in high performance liquid chromatography (HPLC) to determine the OA concentration [47]. The oxalate degradation rate was calculated using the method described by Zeng et al. [16]. There were three replicates for each treatment, and the entire experiment was conducted three times. Mycoparasitic activity of the wild type and mutants of C. minitans to the hyphae and sclerotia of S. sclerotiorum was performed using the method described by Lou et al. [35]. Infection of the hyphae was evaluated by the aggressiveness of a mutant (or ZS-1) in the invasion of the colonies of S. sclerotiorum in the dual cultures of that mutant (or ZS-1) and S. sclerotiorum, and infection of sclerotia by a mutant (or ZS-1) was evaluated by sclerotial rot index [35]. This experiment was also repeated three times.

Confocal Microscopy
To visualize the localization of CmOxDC3, the complementary mutant ∆CmOxdc3-7C with the transformed with the eGFP gene (eGFP = enhanced green fluorescent protein) and the eGFP control strain ZS-1-E1 were inoculated in PDB (50 mL per flask) with 1 × 10 7 conidia in each flask, and the cultures were shake-incubated at 20 • C and 150 rpm for 3 days. Then, OA was added to the flasks to the final concentrations of 0 (control), 8 mM, 16 mM or 24 mM, and the cultures were further incubated for 1 h. The hyphae were collected from each flask and observed under Leica SP8 Inverted Confocal Microscope (Germany).

Statistical Analysis
One-way analysis of variance (ANOVA) in the GraphPad Prism 8.0.2 (https://www. graphpad.com/; accessed on 6 July 2020) was used to determine significant differences among mutants and the wild type (ZS-1) of C. minitans in OA degradation rates, sclerotial rot index and relative gene expression levels. Treatment means were separated using Duncan's multiple range test or Student's t test.

Identification and Characteristics of CmOxdc3
The amino acid sequences encoded by CmOxdc1 (GenBank Acc. No. JF718548) and CmOxdc2 (GenBank Acc. No. JF718549) were used as queries to search the homologs in the genome of C. minitans ZS-1 (GenBank Acc. No. VFEO01000000). A homologous gene was finally found, and the coding region of that gene was 1565 bp long with one intron (50 bp long) and two extrons (333 and 1182 bp long). A polypeptide with 504 amino acids (aa) encoded by the homologous gene was predicted, and it shared 35.76% and 31.56% identity to the polypeptides CmOxDC1 and CmOxDC2 encoded by CmOxdc1 and CmOxdc2, respectively. Moreover, the 504-aa polypeptide harbored a bicupin structure similar to that in CmOxDC1 and CmOxDC2 ( Figure 1A). However, different from CmOxDC1 and CmOxDC2, the 504-aa polypeptide did not have a signal peptide. Therefore, the homologous gene was probably a novel oxalate decarboxylase gene, herein designated as CmOxdc3 (GenBank Acc. No. MN688991). SWISS-MODEL (https://swissmodel.expasy.org/; accessed on 5 March 2022) analysis showed that the predicted structure of CmOxDC3 shared 55.59% identity to that of oxalate decarboxylase in Bacillus subtilis ( Figure S3). Phylogenetic analysis showed that the 39 OxDCs were grouped into five clades (A to E), CmOxDC3 belonged to Clade A, whereas CmOxDC1 and CmOxDC2 belonged to Clade D and Clade E, respectively ( Figure 1B).

Expression Pattern and Subcellular Localization of CmOxdc3
The results of RT-qPCR indicated that the wild type ZS-1 of C. minitans had a distinct expression pattern of CmOxdc3 in response to OA and other acids as well as to ambient pH ( Figure 2). In the OA assay, expression of CmOxdc3 was significantly (p < 0.01) up-regulated in the presence of OA, compared to the control treatment, the OA treatments (12, 24 and 32 mM) had the average relative expression values (REVs) of CmOxdc3 being increased by 21, 34 and 7 folds, respectively (Figure 2A). CmOxdc1 showed a similar expression pattern, the average REVs under the OA treatments were increased by 2 to 10 folds, compared to the control treatment.
Besides OA, malonic acid and HCl could also trigger a significant (p < 0.01) upregulated expression of CmOxdc3 with the average REVs at 81 and 8, respectively, in the presence of the two acids ( Figure 2B). In contrast, the remaining five acids (fumaric acid, maleic acid, lactic acid, acetic acid and citric acid) yielded the average REVs of CmOxdc3 lower than 4.5, which did not significantly (p > 0.05) differ from that in the control treatment. This expression pattern appeared different from that of CmOxdc1, which was significantly (p < 0.01) up-regulated by OA as well as by malonic acid, HCl, fumaric acid, maleic acid and lactic acid. The average REVs in these six acids ranged from 9 to 68. nificantly (p < 0.01) up-regulated by OA as well as by malonic acid, HCl, fumaric acid, maleic acid and lactic acid. The average REVs in these six acids ranged from 9 to 68.
In the pH assay, expression of CmOxdc3 did not have a significant (p > 0.05) change under pH 3 to 6. Results also showed that expression of CmOxdc1 was significantly (p < 0.01) increased by 9 and 13 folds under pH 3 and 4, respectively; however, it was not upregulated under pH 5 and 6 ( Figure 2C).  In the pH assay, expression of CmOxdc3 did not have a significant (p > 0.05) change under pH 3 to 6. Results also showed that expression of CmOxdc1 was significantly (p < 0.01) increased by 9 and 13 folds under pH 3 and 4, respectively; however, it was not up-regulated under pH 5 and 6 ( Figure 2C).
To understand the subcellular localization of the polypeptides (CmOxDC3) encoded by CmOxdc3, we constructed the complementary mutant ∆CmOxdc3-7C, which contained the coding sequence (without stop codon) of CmOxdc3 and the enhanced green fluorescent protein (eGFP)-coding gene under the native promoter of CmOxdc3. Microscopic observation showed that the CmOxDC3-eGFP was visible in the cytoplasm of the transformed hyphae in the absence of OA. With the increase of oxalic acid concentration (from 8 mM to 24 mM), CmOxDC3-eGFP signal was greatly enhanced and gradually gathered. After 1-h 24 mM OA treatment, CmOxDC3-eGFP was obviously gathered to vacuoles in the cytoplasm of the transformed hyphal cells (Figure 3). However, the control eGFP signal was diffusely distribution in the cytoplasm of the control strain (ZS-1-E1) hyphae with or without OA treatment.
scopic observation showed that the CmOxDC3-eGFP was visible in the cytoplasm of transformed hyphae in the absence of OA. With the increase of oxalic acid concentrat (from 8 mM to 24 mM), CmOxDC3-eGFP signal was greatly enhanced and gradua gathered. After 1-h 24 mM OA treatment, CmOxDC3-eGFP was obviously gathered vacuoles in the cytoplasm of the transformed hyphal cells (Figure 3). However, the c trol eGFP signal was diffusely distribution in the cytoplasm of the control strain (ZS E1) hyphae with or without OA treatment.

Effect of Disruption of CmOxdc3 on Sensitivity to OA
To explore the specific function of CmOxdc3, the gene deletion mutants ΔCmOxd 2 and ΔCmOxdc3-7 were generated using the split-marker strategy ( Figure S1A). In ad tion, the complementary mutant CmOxdc3-7C and the double disruption muta ΔCmOxdc1&3-13, ΔCmOxdc1&3-45 were also obtained. Disruption of CmOxdc3, as w as CmOxdc1 and complementation of CmOxdc3, were confirmed by PCR and/or Sou ern blotting ( Figure S4). The wild type (ZS-1) and the mutants of CmOxdc3 as well as mutants of CmOxdc1 (ΔCmOxdc1-1, ΔCmOxdc1-25) were compared for sensitivity to on PDA amended with OA at 8, 16 or 24 mM (PDA-OA 8, PDA-OA 16, PDA-OA 24, spectively). The results showed that after incubation at 20 °C for 10 days, both the w type and all the mutants formed large colonies on PDA alone (control), however, th formed small colonies on PDA amended with OA at 8, 16 and 24 mM (Figure 4 Therefore, OA can inhibit mycelial growth of the tested strains of C. minitans. The resu also showed that the colony size varied with the strains of C. minitans on PDA-OA 8 and 24. On PDA-OA 8, ΔCmOxdc3-2 and ΔCmOxdc3-7 had a significantly (p < 0.05) hi

Effect of Disruption of CmOxdc3 on Sensitivity to OA
To explore the specific function of CmOxdc3, the gene deletion mutants ∆CmOxdc3-2 and ∆CmOxdc3-7 were generated using the split-marker strategy ( Figure S1A). In addition, the complementary mutant CmOxdc3-7C and the double disruption mutants ∆CmOxdc1&3-13, ∆CmOxdc1&3-45 were also obtained. Disruption of CmOxdc3, as well as CmOxdc1 and complementation of CmOxdc3, were confirmed by PCR and/or Southern blotting ( Figure  S4). The wild type (ZS-1) and the mutants of CmOxdc3 as well as the mutants of CmOxdc1 (∆CmOxdc1-1, ∆CmOxdc1-25) were compared for sensitivity to OA on PDA amended with OA at 8, 16 or 24 mM (PDA-OA 8, PDA-OA 16, PDA-OA 24, respectively). The results showed that after incubation at 20 • C for 10 days, both the wild type and all the mutants formed large colonies on PDA alone (control), however, they formed small colonies on PDA amended with OA at 8, 16 and 24 mM ( Figure 4A). Therefore, OA can inhibit mycelial growth of the tested strains of C. minitans. The results also showed that the colony size varied with the strains of C. minitans on PDA-OA 8, 16 and 24. On PDA-OA 8, ∆CmOxdc3-2 and ∆CmOxdc3-7 had a significantly (p < 0.05) higher mycelial growth-inhibition (MGI) rate (31%) than the wild type (22%) and CmOxdc3-7C (27%), however, the value was significantly (p < 0.05) lower than those (46% to 54%) for ∆CmOxdc1-1, ∆CmOxdc1-25, ∆CmOxdc1&3-13, ∆CmOxdc1&3-45 ( Figure 4B). This result suggests that disruption of CmOxdc3 causes less sensitivity to OA at 8 mM than disruption of CmOxdc1.

Effect of Disruption of CmOxdc3 on Mycoparasitism on the Host Hyphae
The dual-cultural assay was used to determine mycoparasitic infection of the wild type and the mutants of C. minitans on the hyphae of S. sclerotiorum (20 °C, 30 days). The assay was established on PDA amended with bromophenol blue (a pH indicator) to monitor the production of OA by S. sclerotiorum as shown by the change of the medium color from blue to yellow ( Figure 6A). The result showed that the wild type and all the mutants could invade the colonies of S. sclerotiorum in the dual cultures, thereby destroying the hyphae of S. sclerotiorum. However, the mutants differed from the wild type in invasion aggressiveness, while the wild type invaded the colonies of S. sclerotiorum from Zone I (close to the inoculation point of C. minitans) to Zone IV (close to the inoculation point of S. sclerotiorum), the mutants ΔCmOxdc3-2 and ΔCmOxdc3-7, as well as ΔCmOxdc1-1, invaded the colonies of S. sclerotiorum from Zone I to Zone III, and the remaining mutants ΔCmOxdc1-25, ΔCmOxdc1&3-13 and ΔCmOxdc1&3-45 invaded the colonies of S. sclerotiorum from Zone I and Zone II. This result suggests that disruption of CmOxdc3, as well as CmOxdc1, can delay invasion into the colonies of S. sclerotiorum possibly due to reduced capability for eliminating the toxicity of OA from S. sclerotiorum.
The sclerotial assay was used to determine the mycoparasitic infection of the wild type and the mutants of C. minitans on the sclerotia of S. sclerotiorum ( Figure 6B). The result showed that after incubation at 20 °C for 30 days, the wild type and the mutants could infect the sclerotia of S. sclerotiorum, causing blackening and rot of the sclerotia. The average sclerotial rot index (0 to 100) was as high as 82 for the wild type, however, the value was significantly (p < 0.05) reduced to 62-69 for the mutants ( Figure 6C). Therefore, disruption of CmOxdc3, as well as CmOxdc1, can reduce the capability of C. minitans to infect and destroy the sclerotia of S. sclerotiorum.

Effect of Disruption of CmOxdc3 on Mycoparasitism on the Host Hyphae
The dual-cultural assay was used to determine mycoparasitic infection of the wild type and the mutants of C. minitans on the hyphae of S. sclerotiorum (20 • C, 30 days). The assay was established on PDA amended with bromophenol blue (a pH indicator) to monitor the production of OA by S. sclerotiorum as shown by the change of the medium color from blue to yellow ( Figure 6A). The result showed that the wild type and all the mutants could invade the colonies of S. sclerotiorum in the dual cultures, thereby destroying the hyphae of S. sclerotiorum. However, the mutants differed from the wild type in invasion aggressiveness, while the wild type invaded the colonies of S. sclerotiorum from Zone I (close to the inoculation point of C. minitans) to Zone IV (close to the inoculation point of S. sclerotiorum), the mutants ∆CmOxdc3-2 and ∆CmOxdc3-7, as well as ∆CmOxdc1-1, invaded the colonies of S. sclerotiorum from Zone I to Zone III, and the remaining mutants ∆CmOxdc1-25, ∆CmOxdc1&3-13 and ∆CmOxdc1&3-45 invaded the colonies of S. sclerotiorum from Zone I and Zone II. This result suggests that disruption of CmOxdc3, as well as CmOxdc1, can delay invasion into the colonies of S. sclerotiorum possibly due to reduced capability for eliminating the toxicity of OA from S. sclerotiorum.

Discussion
Oxalic acid (OA) plays various roles for the OA-producing fungal pathogens to infect plants [5,6], and microbial degradation of OA is a valuable research topic regarding The sclerotial assay was used to determine the mycoparasitic infection of the wild type and the mutants of C. minitans on the sclerotia of S. sclerotiorum ( Figure 6B). The result showed that after incubation at 20 • C for 30 days, the wild type and the mutants could infect the sclerotia of S. sclerotiorum, causing blackening and rot of the sclerotia. The average sclerotial rot index (0 to 100) was as high as 82 for the wild type, however, the value was significantly (p < 0.05) reduced to 62-69 for the mutants ( Figure 6C). Therefore, disruption of CmOxdc3, as well as CmOxdc1, can reduce the capability of C. minitans to infect and destroy the sclerotia of S. sclerotiorum.

Discussion
Oxalic acid (OA) plays various roles for the OA-producing fungal pathogens to infect plants [5,6], and microbial degradation of OA is a valuable research topic regarding the initiation of novel control strategies against the OA-producing fungi [6,11,[15][16][17][18][19][20]. C. minitans is a promising biological control agent of S. sclerotiorum [28][29][30][31][32]. It is of great significance to study the role and molecular mechanisms for OA degradation by C. minitans. In this study, we cloned a novel OA degradation-associated gene (CmOxdc3) from C. minitans, which may work in a synergistic manner with CmOxdc1. CmOxdc3 and CmOxdc1 share the conserved bicupin structure, and the expression of both genes can be triggered by OA. However, CmOxdc3 differs from CmOxdc1 in two aspects. First, CmOxdc3 does not have a signal peptide-coding region, whereas CmOxdc1 has a signal peptide-coding region. As a consequence, CmOxDC3 is located in hyphal cytoplasm as observed in this study. In addition, CmOxdc3 had no detectable response to fumaric acid, maleic acid, lactic acid, acetic acid and citric acid, as well as the ambient acidic pH signaling, whereas CmOxdc1 could be triggered for expression by these acids and acidic pH signaling.
In the process of evolution, microorganisms usually retain multiple genes to cope with the same biochemical processes, such as microbial degradation of OA. This polygenic phenomenon usually has a positive impact on the survival of microorganisms. Previous studies showed that the presence of multiple oxalate decarboxylases (OxDCs) is very common in fungi and bacteria [16], this phenomenon implies that OA degradation by fungi or bacteria is a complex biochemical process. Zeng et al. (2014) cloned two oxalate decarboxylase genes (CmOxdc1 and CmOxdc2) from C. minitans and proved that CmOxdc1 is important for C. minitans to degrade OA [16]. Two oxalate decarboxylase genes (Ss-odc1 and Ss-odc2) have been reported in S. sclerotiorum, and only Ss-odc2 is responsible for degradation of OA [11]. In this study, we found a novel oxalate decarboxylase gene (CmOxdc3) in C. minitans, which had a distinct structure and expression pattern from CmOxdc1 and CmOxdc2. This study found that CmOxdc3 also plays a certain role in degradation of OA. This result, together with our previous observation about the role of CmOxdc1, suggests that CmOxdc1 and CmOxdc3 might be responsible for OA degradation in C. minitans.
Previous studies indicated that fungal oxalate decarboxylases have diverse subcellular localization [23][24][25][26]. This study showed that CmOxDC3 was localized in the cytoplasm of hyphal cells; it moved to vacuoles in response to extracellular OA. Vacuoles in fungi are similar to lysosomes in mammalian cells. They are acidified organelles, which have the function of maintaining cell pH homeostasis, storage and recycling of nutrients, degradation of macromolecules and membrane transport [48,49]. In C. minitans, disruption of the vacuolar morphogenesis-related CmVps39 affected osmotic adaptation, pH homeostasis and cell wall integrity [50]. Therefore, vacuoles may be the main site for intracellular degradation of OA. Intracellular localization of CmOxDC3 suggests that it may differ from CmOxDC1 in degradation of OA, whereas CmOxDC1 has a signal peptide region, and CmOxDC1 can be transported outside of the hyphal cells (Method S1, Figure S5) to degrade extracellular OA.
Based on the results so far achieved, we proposed a model for the synergistic degradation of OA by CmOxDC1 and CmOxDC3 in the interaction between C. minitans and S. sclerotiorum. S. sclerotiorum produces OA, which is transported outside. Partial OA may enter into the hyphal cells of C. minitans (possibly gathers in vacuoles) by an unknown way. Meanwhile, C. minitans senses OA via unknown receptors, and the resulting signals are then transmitted to nuclei, thereby activating the transcription of CmOxdc1/3. CmOxDC3 degrades the intracellular OA, whereas CmOxDC1, with the signal peptide, is secreted to the extracellular matrix to degrade extracellular OA.
In summary, this study cloned a novel oxalate decarboxylase gene (CmOxdc3) in C. minitans. It did not harbor the signal peptide region, it was found to locate inside the hyphal cells, gathering in vacuoles. Transcription of CmOxdc3 was triggered by OA, malonic acid and HCl, but failed to respond to ambient acidic pH from 3 to 6. Disruption of CmOxdc3 increased sensitivity of C. minitans to OA as well as reduced the efficiency of C. minitans to degrade OA and to infect S. sclerotiorum. Therefore, CmOxdc3 plays a different role from that of CmOxdc1, and both CmOxdc1 and CmOxdc3 may work in a synergistic manner to eliminate the toxicity of OA.  Figure S5: Investigation of secretory property of CmOxDC1 signal peptides by a yeast signal trap system; Method S1: Yeast signal sequence trap system [51]; Table S1: Primers used in this study [16]; Table S2: Sequences of 39 OXDCs to construct phylogenetic tree.