An Amidase Contributes to Full Virulence of Sclerotinia sclerotiorum

Sclerotinia sclerotiorum is one of the most notorious and ubiquitous soilborne plant pathogens, causing serious economic losses to a large number of hosts worldwide. Although virulence factors have been identified in this filamentous fungus, including various cell-wall-degrading enzymes, toxins, oxalic acids and effectors, our understanding of its virulence strategies is far from complete. To explore novel factors contributing to disease, a new pipeline combining forward genetic screening and next-generation sequencing was utilized in this study. Analysis of a hypovirulent mutant revealed that a mutation in an amidase-encoding gene, Sscle_10g079050, resulted in reduced virulence. This is a first report on the contribution of an amidase to fungal virulence, likely through affecting oxalic acid homeostasis.


Introduction
Sclerotinia sclerotiorum (Lib.) de Bary is an economically devastating fungal pathogen with an extremely broad host range. Infection is initiated through multiple orchestrated strategies that ultimately incite rapid host tissue maceration, causing diseases in over 600 plant species [1]. Largely through reverse genetic analysis, genes involved in apothecia and ascospore development [2][3][4], mycelial growth [5,6], appressorium establishment [7,8], sclerotial formation and development [9][10][11], and fungal virulence [7,12,13] have been identified over the last decades. Many virulence factors were confirmed to play critical roles in host tissue disruption and degradation, such as cell-wall-degrading enzymes (CWDEs) and toxins [14]. In addition, some small secreted proteins functioning in both host defense suppression and inducing host cell death have been identified [15,16].
The full virulence of S. sclerotiorum requires controlled oxalic acid (OA) accumulation. OA levels are regulated through its biosynthesis and degradation [7,12,13]. Defective OA synthesis leads to severely compromised virulence, implicating the importance of OA in Sclerotinia pathogenesis. Proposed roles of OA include creating a low pH environment available for hydrolytic enzyme activities [17][18][19], contributing to calcium chelation and cell wall breakdown, reducing calcium exposure toxicity during invasion of growing hyphae [20,21], eliciting apoptotic programmed cell death to allow fungal pathogenicity [22], and disruption of chloroplast function [23]. Moreover, OA suppresses and promotes host reactive oxygen species (ROS) at different stages of infection to achieve successful disease development [22,24,25].
With the increasing availability of sequenced whole genomes of fungal pathogens, reverse genetic analysis for genes of interest are available for functional investigation. In terms of novel gene discovery, it is inevitably limited by bias. Recently, Xu et. al. [26] provided an effective pipeline, which combines forward genetic screening with high-throughput nextgeneration sequencing (NGS) to enable quick gene discovery and confirmed the reliability of this method in S. sclerotiorum.
Here, we performed a forward genetic screen for virulence-related genes using the same mutagenized population acquired from Xu et. al. [26]. Mutants defective in virulence were isolated and NGS was used to identify the candidate genes. Specifically, we describe a recovered mutant with a mutation in an amidase-encoding gene. Knockout and transgene complementation experiments confirmed that this amidase mutation is responsible for the observed virulence phenotype. Thus, our study uncovered an important role amidase plays in fungal virulence, probably by affecting OA homeostasis.

Screening for S. sclerotiorum Mutants with Reduced Virulence
Using a UV-mutagenesis based forward genetics pipeline [26], we generated a mutagenized population originated from single haploid ascospores. Our initial screen aimed for mutants deficient in sclerotial formation and melanization. The same population was then grown to search for mutants which exhibit reduced virulence on lettuce leaves. Here, we report on the characterization of mutant 5-10. On lettuce leaves, wild-type (WT) S. sclerotiorum caused large lesions 48 h post-inoculation, while 5-10 mutant formed smaller ones ( Figure 1A,B). To test whether the reduced virulence was associated with hyphal development, we examined the growth rate of the mutant. As shown in Figure 1C,D, mycelial growth of the 5-10 mutant is comparable with WT. Accordingly, the defects in 5-10 seems specific to fungal virulence. leaves. ImageJ was used to quantify the lesion size. Statistical significance was analyzed using Student's t-test between WT and the 5-10 mutant. a and b represent different significance levels. (C) Colonial morphology and mycelial growth of WT and 5-10 strains after 12 h, 24 h, 36 h, 48 h and 7 days on potato dextrose agar (PDA) media. (D) Colony diameter of WT and 5-10. Strains were grown on PDA medium and colony diameters were measured every 12 h. All experiments were conducted at least three times with similar results. Error bars represent standard error (SE). Statistical significance was analyzed using Student's t-test between WT and the 5-10 mutant.

Less Oxalic Acid Is Observed in 5-10 Mutant
As oxalic acid (OA) is a key virulence factor in S. sclerotiorum [20], we examined whether the level of OA in 5-10 is altered. When OA was measured both qualitatively and quantitatively (Figure 2A,B), 5-10 indeed showed significantly reduced OA amounts. Thus, the reduced virulence observed in 5-10 is likely due to its reduced OA levels during mycelial growth.

Mutation in Sscle_10g079050 Is Responsible for the Mutant Phenotypes of 5-10
To identify the mutation responsible for the observed phenotypes of 5-10, the full genome of 5-10 was subjected to whole genome next-generation sequencing (NGS). The genomic sequence of an unrelated mutant pnk1 was used as negative control to exclude background mutations that occured in both 5-10 and pnk1 [26]. From whole genome sequence comparisons, a total of 77 SNPs (single nucleotide polymorphisms) and 689 INDELs (insertions and deletions) were captured in 5-10, including mutations in exonic, intronic, upstream, downstream, and intergenic regions. After manual elimination of synonymous, intronic, intergenic, and non-homozygous mutations, 2 exonic SNPs remain that led to nonsynonymous changes in the 5-10 genome. These two SNPs in Sscle_06g054440 (L1407S) and Sscle_10g079050 (L82S) became the prime candidate mutations for 5-10.
To determine which of the two mutations is responsible for the 5-10 phenotypes, transgenic complementation of Sscle_06g054440 and Sscle_10g079050 was individually conducted (Figures 3 and 4). Only when Sscle_10g079050 was transformed into 5-10, did the transgenic lines yield similar level of acidic substances and display almost identical virulence compared to WT ( Figure 3B-E). Transformation of the other candidate gene had no effect on virulence ( Figure 4A,B). Therefore, we conclude that the mutation in Sscle_10g079050 is likely responsible for the 5-10 phenotypes.  To investigate the function of Sscle_10g079050 (protein accession number, APA13135.1/ XP001584936.1), homology alignment and phylogenetic analysis were conducted within S. sclerotiorum ( Figure 5A). This gene appears to be a member of a large gene family, and the highest identity with its paralogous protein (Sscle_04g033530) was only 42.34%. Phylogenetic analysis among multiple species indicated that the transcripts of Sscle_10g079050 were conserved in fungi ( Figure 5B,C). Based on the annotation of its homologous proteins, Sscle_10g079050 was predicted to possess a function of amidase or acetamidase. The point mutation in 5-10 caused an amino acid to change from Leucine to Serine (L82S) and this Leu is conserved in fungi, especially in Leotiomycetes (Figure 5C), suggesting a critical role of this residue to its function. As Sscle_10g079050 encodes an enzyme catalyzing the hydrolysis of an unknown amide, it does not seem to be directly involved in OA biosynthesis or catabolism. It likely affects the OA level through indirect means.

5-10 and Sscle_10g079050 Knockout Mutants Exhibit Similar Hypovirulence
To further confirm that the mutation in Sscle_10g079050 is responsible for the hypovirulence in 5-10, targeted gene replacement by homologous recombination was performed in WT background. The Sscle_10g079050 transformants were selected on PDA with 50 µg/mL of hygromycin, and two independent transformants, KO1 and KO2, were obtained ( Figure 6A). PCR analysis showed that the WT copy of Sscle_10g079050 was not fully absent in the knockout strains, suggesting that the mutants generated likely represent partial loss-of-function alleles ( Figure 6B,C). The virulence in KO1 and KO2 nevertheless were reduced compared with WT on both lettuce and Arabidopsis ( Figure 6D-G). Therefore, we conclude that the mutation in Sscle_10g079050 is responsible for the hypovirulence in 5-10. As knocking out Sscle_10g079050 resulted in attenuated virulence, the encoded amidase serves as a positive regulator of virulence. When its gene expression pattern was analyzed from publicly available RNA-Seq datasets [27], we found that it is up-regulated after 24 h of infection ( Figure 6H), which supports its critical role in virulence.

Discussion
Historically, virulence-related genes in S. sclerotiorum were generally identified based on either homology with well-characterized genes from other pathogens, or genes encoding secreted proteins with effector functions. However, these approaches have limitations and biases. In this study, we identified an amidase-encoding gene (Sscle_10g079050) using an effective and reliable pipeline which combined forward genetic screening with NGS. Mutants of this gene contained less OA and showed reduced virulence. It is the first evidence that an amidase might contribute to the homeostasis of OA in fungi.
Amidases are versatile enzymes owing to their broad substrate specificity [28]. They are widespread in all organisms including bacteria, fungi, plants, and animals [29,30]. According to their primary structure, amidases are divided into three families, namely amidase signature (AS) family, nitrilase superfamily, and FmdA_AmdA family [31]. Ss-cle_10g079050 belongs to the AS family, which carry a highly conserved stretch of approximately 130 amino acids known as the AS sequence [32]. The crystal structure of analyzed AS enzymes share an α/β sandwich folded core domain, containing a Gly/Ser-rich motif and Ser-cis Ser-Lys catalytic triad [33][34][35]. After sequence alignment with Pam (PDB ID: 1M21), a structural analyzed AS enzyme determined in Stenotrophomonas maltophilia, the catalytic triad of Sscle_10g079050 was predicted as Ser 202 -Ser 226 -Lys 127 (Supplementary Figure S1A). According to the 3D structure model, the residue Leu 82 was located sterically far from that core region but spatially close to Lys 127 (Supplementary Figure S1B). We speculate that the substitution of Leu 82 might affect the catalytic function in an indirect way, leading to loss of the enzymatic function.
In filamentous fungi, OA is derived from either the hydrolysis of oxaloacetate or the oxidation of glyoxylate or glycolaldehyde [20], and OA biogenesis in Ascomycota species relies predominately on oxaloacetate acetylhydrolase (OAH)-mediated oxaloacetate hydrolysis [12]. As a metabolic intermediate of both the tricarboxylic acid (TCA) and glyoxylate cycle, malate or pyruvate might be an oxaloacetate precursor in the OA biosynthesis [36][37][38][39]. In addition, many enzymes involved in lipid metabolism are required for virulence, such as isocitrate lyase in Magnaporthe grisea, Leptosphaeria maculans, and Colletotrichum lagenarium [40,41], and malate synthase in Stagonospora nodorum and S. sclerotiorum [42,43]. It is well known that isocitrate lyase and malate synthase are key enzymes in glyoxylate cycle, but whether they affect pathogenic processes in an OA-dependent manner remains unclear. Amidases diverged widely with regard to substrate specificity and function. A large spectrum of amide substrates, including aliphatic amides, aromatic amides, and α-hydroxyamides were found to be catalyzed by AS enzymes [44][45][46]. Oneway Sscle_10g079050 affects OA accumulation might be through hydrolyzing suitable amide-containing substrates during OA biosynthesis.
Amidases in the AS family usually exhibit both hydrolytic and acyl transfer activities. The intracellular transport of acetyl coenzymeA (acetyl-CoA) is a valuable component of the glyoxylate cycle, and a peroxisomal carnitine acetyl transferase (Pth2) identified in Magnaporthe grisea was found to facilitate acetyl-CoA transport specificity, which was required for both β-oxidation of fatty acid and pathogenicity [47,48]. With this in mind, it is possible that Sscle_10g079050 could also play a role in glyoxylate pathway by its acyl transfer activity. Moreover, OA level is also indirectly affected by the redox status of S. sclerotiorum [10]. Thus, Sscle_10g079050 may affect OA accumulation by participating in maintenance of redox environment. The amidase may even affect OA-independent processes involved in virulence.
Sscle_10g079050 is predicted to be a cytoplasmic enzyme, and it contains a N-terminal cytoplasmic tail (1 to 463aa) and a C-terminal single transmembrane segment, implying that it might anchor to membrane and cannot be accessible to other organelles. Since malate and citrate biosynthesized by the TCA cycle or the glyoxylate cycle can diffuse across the membrane barrier to be exported from organelles and finally be transformed into oxaloacetate, Sscle_10g079050 may co-locate with malate or citrate in the cytosol pool, and play a role in oxaloacetate or OA metabolism processes. In the future, biochemical and chemical studies will be essential for identifying the exact substrate of this amidase to illuminate its precise mechanism of affecting OA levels and pathogenic processes.

Fungal Strains and Culture Condition
Wild-type strain S. sclerotiorum 1980 and mutant were maintained on standard PDA (potato dextrose agar, Bio-Way Technology, Shanghai, China) medium at 20 • C and stored on PDA slants at 4 • C or as sclerotia. Transformants (deletion mutants and complemented strains) were grown on PDA containing 50 µg/mL hygromycin B (Roche).

Plant Infection Assay
Virulence of the strains was tested on detached leaves of lettuce and Arabidopsis thaliana (ecotype Col-0). Mycelial agar plugs 2 mm in diameter were inoculated on detached leaves of A. thaliana (5 mm in diameter on lettuce) placed on moistened paper towels in Petri dishes (90 mm diameter), with five repeats. Photos were taken at 48 h post-inoculation.
The pathogenicity test was repeated three times with similar results. Lesion areas were measured by ImageJ.

Screening for Mutants with Deficiency in Virulence
Ascospores were collected from apothecia of wild-type (WT) strain S. sclerotiorum 1980. The UV mutagenesis was conducted on ascospores as previously described [26]. The mutagenized population was initially screened using lettuce leaves obtained from supermarkets. The putative mutants were then confirmed with multiple repeats on both lettuce and Arabidopsis.

Genomic DNA Extraction and NGS
Fresh fungal mycelia were cultured in PDB for one week. Surface floating mycelia were harvested and ground to fine powder in liquid nitrogen. The genomic DNA extraction was conducted by cetyltrimethylammonium bromide (CTAB) method from the mutant and WT strains [49]. The crude extract was further purified for NGS with a commercial service (Novogene, Bioinformatics Technology Co., Ltd., Beijing, China). DNA degradation and contamination were evaluated on a 1% agarose gel. Paired-end library was built by Novogene, using Illumina sequencer NovaSeq 6000, and clean reads were aligned to the S. sclerotiorum genome (ASM185786v1).

Candidate Genes Identification
The sequence reads from NGS were mapped to the reference genome of WT strain S. sclerotiorum 1980. Mutations were identified by SAMtools with default parameters [50]. Annotation of the mutations was performed with germline short variant discovery (SNPs + IN-DELs) based on GATK best practices [51]. False mutations in repetitive sequences were manually removed as previously described [52].

Sanger Sequencing of Sscle_10g079050 in 5-10
Sscle_10g079050 gene fragment was amplified with SeqF and SeqR primers and genomic DNA of 5-10 was isolated by CTAB method and used as template. A 660-bp PCR product was then sequenced using universal primer by Sangon Biotech in China.

Target Gene Knockout and Transgene Complementation
The split marker method was used to generate Sscle_10g079050 gene replacement constructs [53]. In brief, 1,102-bp upstream and 655-bp downstream flanking fragments of Sscle_10g079050 were amplified with the primer pairs 5-10-10g KO 1F/2R and 5-10-10g KO 3F/4R, respectively, and were fused with two overlapped hph fragments amplified with the primers HYG-F/HY-R and YG-F/HYG-R from pUCH18 plasmid (shared by D. Jiang from Huazhong Agricultural University in China) [54]. All primers used for PCR are listed in Supplementary Table S1. Gene flanking fragments were fused with two hph fragments, respectively, by homologous recombination methods. The amplified PCR products were cotransformed into protoplasts of WT as described [54]. Transformants were selected on fresh PDA plates supplemented with hygromycin B. The individual germinating transformants were transferred onto new PDA plates with hygromycin B. Fresh mycelia were used to perform colony PCR with a pair of primers, 5-10-10g KO 7F/8R (Supplementary Table S1), to confirm the insertion of the hygromycin gene. Positive transformants were inoculated onto new PDA plates with hygromycin B by hyphal tip transfer.
Two additional rounds of the same colony PCR and selection were conducted to accumulate more nuclei with hygromycin insertions. Then, the transformants were cultured in PDB for 2 days. The fresh mycelia mat was used to produce protoplasts as mentioned above. The resulting protoplasts were diluted with PDB supplemented with hygromycin B to 10 3 protoplasts per milliliter. A 1-mL suspension was evenly spread on PDA plates supplemented with hygromycin B and dried. The germinating colonies were transferred onto individual PDA plates with hygromycin B. Transformant DNA was extracted using the CTAB method, which were later used as template to conduct PCR with four pairs of primers as shown in Figure 6A. The transformants with the target gene being homogeneously knocked out were chosen for further experiments.
For transgenic complementation, a knock-in method was performed to accomplish the in-situ complementation by candidate gene homologous recombination. As described in the knockout protocol, a 3307-bp fragment including upstream and genomic DNA and a 655-bp downstream flanking fragment were PCR-amplified. The resulting two fragments were fused with hpg fragments, respectively, and the amplified PCR products were co-transformed into protoplasts of 5-10 as described above. SNP primers (Supplementary Table S1) designed on the mutation site were used to check the insertion of Sscle_10g079050 gene.

OA Analysis
Mycelial plugs (5 mm diameter) with growing hyphal tips were inoculated on PDA medium containing 100 µg/mL bromophenol blue to qualitatively measure oxalic acid. Three mycelia-colonized plugs were also inoculated in PDB medium containing 100 µg/mL bromophenol blue for UV absorbance test after 48 h oscillation with three repeats. OA contents were detected using Oxalic Acid Content Detection Kit following the manufacturer's instructions (Solarbio Life Science, Beijing, China).

Phylogenetic Analyses
Homologous proteins were searched within S. sclerotiorum and among different species. The phylogenetic tree of selected homologous proteins was generated using MEGA7.0 with the maximum likelihood (ML) algorithm. Bootstrap analysis with 1000 replications was performed to assess group support.
Supplementary Materials: The supporting information can be downloaded at: https://www.mdpi. com/article/10.3390/ijms231911207/s1. Author Contributions: W.L., J.L. and C.Y. carried out the experiments. W.L., J.L., K.A. and X.L. wrote the manuscript with inputs from all authors. X.L. conceived the original idea and S.X. supervised the project. All authors contributed to the article and approved the submitted version. All authors have read and agreed to the published version of the manuscript. Acknowledgments: The authors cordially thank Daohong Jiang for plasmid of pUCH18 and Yan Xu for mutagenized population of S. sclerotiorum.

Conflicts of Interest:
The authors declare no conflict of interest.