The bZIP Transcription Factor LtAP1 Modulates Oxidative Stress Tolerance and Virulence in the Peach Gummosis Fungus Lasiodiplodia theobromae

Lasiodiplodia theobromae is one of the primary causal agents in peach gummosis disease, leading to enormous losses in peach production. In our previous study, a redox-related gene, LtAP1, from the fungus was significantly upregulated in peach shoots throughout infection. Here, we characterized LtAP1, a basic leucine zipper transcription factor, during peach gummosis progression using the CRISPR-Cas9 system and homologous recombination. The results showed that LtAP1-deletion mutant had slower vegetative growth and increased sensitivity to several oxidative and nitrosative stress agents. LtAP1 was highly induced by exogenous oxidants treatment in the L. theobromae wild-type strain. In a pathogenicity test, the deletion mutant showed decreased virulence (reduced size of necrotic lesions, less gum release, and decreased pathogen biomass) on infected peach shoots compared to the wild-type strain. The mutant showed severely reduced transcription levels of genes related to glutaredoxin and thioredoxin in L. theobroame under oxidative stress or during infection, indicating an attenuated capacity for reactive oxygen species (ROS) detoxification. When shoots were treated with an NADPH oxidase inhibitor, the pathogenicity of the mutant was partially restored. Moreover, ROS production and plant defense response were strongly activated in peach shoots infected by the mutant. These results highlight the crucial role of LtAP1 in the oxidative stress response, and further that it acts as an important virulence factor through modulating the fungal ROS-detoxification system and the plant defense response.


INTRODUCTION
The necrotrophic fungus, Lasiodiplodia theobromae, is geographically widespread in the subtropical and tropical regions and is known to attack approximately 500 plant species, including crops and woody trees (Alves et al., 2008;Cipriano et al., 2015). This fungus has been regarded as a latent pathogen or an opportunistic pathogen leading to dieback, canker, or fruit rot diseases in many economically important woody crops (Slippers and Wingfield, 2007;Ali et al., 2019). In southern China, L. theobromae is also a causal agent of peach gummosis, one of the most devastating diseases of peach (Prunus persica), annually causing considerable quantity and quality losses (Beckman et al., 2003;Wang et al., 2011). A better understanding of the molecular mechanisms of the peach-L. theobromae interaction is necessary for effective control of peach gummosis.
To establish successful infections, pathogens need to overcome both preformed and induced host defenses (Qi et al., 2017). During pathogen attacks, one of the major and fastest plant defense responses is a rapid accumulation of reactive oxygen species (ROS) at the invasion site (Scheler et al., 2013). ROS, primarily superoxide ( O 2 -) and hydrogen peroxide (H 2 O 2 ), are produced by plasma membrane-localized NADPH oxidases, also known as respiratory burst homologs (RBOH), at the inoculation site (Suzuki et al., 2011). Due to the toxicity, ROS can cause oxidative stress and damage to biomolecules, such as DNA mutation, lipid peroxidation, and protein oxidation, eventually causing cell death of the pathogens (De Gara et al., 2003). Additionally, as a class of signaling molecules, ROS play crucial roles in plant-pathogen interactions. Plant-derived ROS act as signaling molecules to mediate various important responses of plant cells to fight against pathogen infection and enhance plant resistance by inducing plant defense-related gene expression and activating related enzyme activity (Torres and Dangl, 2005;Baxter et al., 2014). Our previous study demonstrated that the infection by L. theobromae caused a ROS burst, and transcripts of pathogenesis-related (PR) genes were markedly induced, potentially contributing to the restriction of disease development (Zhang et al., 2020).
To survive and colonize under harsh conditions, pathogens have developed ROS scavenging systems to efficiently reclaim excess ROS (Segal and Wilson, 2018). Scavenging enzymatic and non-enzymatic compounds, either preformed or induced, include superoxide dismutase (SOD), catalases (CAT), peroxidases (POD), glutaredoxins, and thioredoxins (Kawasaki et al., 1997;Lanfranco et al., 2005;Ma et al., 2018). The glutaredoxin system has glutathione, glutathione peroxidase (GPX), glutathione reductase (GLR), and NADPH. The thioredoxin machinery includes thioredoxin peroxidase [equal to thiol-specific antioxidant protein (TSA)], thioredoxin reductase (TRR), thioredoxins (TRX), and NADPH (Ma et al., 2018;Zhang et al., 2019). In fungal pathogens, transcription factor-mediated ROS detoxification through the regulation of antioxidant genes expression is vital in plant-pathogen interactions. One of the critical regulators mediating ROS detoxification is the Activating Protein 1 (AP1) class of basic leucine zipper (bZIP) family (Segal and Wilson, 2018). AP1 is a key transcriptional activator in response to oxidative stress in yeasts and filamentous fungi (Reverberi et al., 2008;Lin et al., 2018;Segal and Wilson, 2018). In our previous study, the LtAP1 gene was consistently and highly expressed in the infection stage of L. theobromae on peach shoots, implying that LtAP1 may play a crucial role in the pathogenicity of L. theobromae (Zhang et al., 2020).
Saccharomyces cerevisiae YAP1 serves as one of the most critical determinants of yeast to oxidative stress response, which is responsible for transcriptional activation of various ROS detoxification-related genes (Mendoza-Martínez et al., 2020). Subsequently, YAP1 homologs in several fungal pathogens were identified and characterized, and found to have conserved roles in oxidative stress response and tolerance, but differed in virulence. YAP1-mediated ROS detoxification has been identified as being an essential virulence determinant in the necrotrophic fungus Alternaria alternata (Lin et al., 2009), the hemibiotrophic rice blast fungus Magnaporthe oryzae (Guo et al., 2011), and the biotrophic maize pathogen Ustilago maydis (Molina and Kahmann, 2007). However, YAP1-assisted ROS detoxification is associated with avirulence in the animal pathogen Aspergillus fumigatus (Lessing et al., 2007), the necrotrophic plant pathogen Cochliobolus heterostrophus (Lev et al., 2005), or the hemibiotrophic plant pathogen Fusarium graminearum (Montibus et al., 2013). Although many studies have examined YAP1 homologs in other fungi, their function in L. theobromae during pathogenesis remains poorly understood, particularly for canker or gummosis disease in woody fruit trees. Understanding the role of the LtAP1 gene in L. theobromae may lead to new tools to develop novel, sustainable disease management strategies against peach gummosis.
In this study, transcription factor LtAP1 was isolated and functionally characterized through genetic transformation. We examined the effects of deletion of the LtAP1 gene on mycelial growth, sensitivity to oxidative and nitrosative stresses, and pathogenicity. This study shed some light on the function of the LtAP1 gene for ROS detoxification, virulence, and suppression of plant defense responses during L. theobromae and peach interaction, which could deepen our knowledge of the role of fungal YAP1s in plant diseases.

Fungal Strains, Culture Conditions, and Chemical Treatments
Lasiodiplodia theobromae pathogenic strain JMB122, obtained from a peach tree with gummosis in Hubei Province, China (Wang et al., 2011), was used as a recipient host for transformation experiments. Both JMB122 and its derivatives were cultured on PDA medium (200 g L −1 potato, 20 g L −1 dextrose, and 15 g L −1 agar) in a growth chamber at 28°C for 36 h under a 12 h-light/ 12 h-dark cycle to assess growth and colony characteristics.
To test stress treatments, the wild type (WT) and genetic transformants of JMB122 were cultured on PDA plates (diameter 9 cm) containing various chemical reagents. The integrity of cell walls and cell membranes was examined on PDA supplemented with calcofluor white (CFW; 0.05 mg ml −1 ), Congo red (2.5 mg ml −1 ), or sodium dodecyl sulfate (SDS; 0.02%). For oxidative stress, PDA was amended with H 2 O 2 (1 or 2.5 mm), tert-butyl-hydroperoxide (TBHP; 0.5 mm), cumene H 2 O 2 (0.68 mm), or menadione (0.1 mm). For nitrosative stress, PDA was amended with sodium nitroferricyanide dihydrate (SNP; 5 mm). PDA was supplemented with glucose (1 M) or KCl (1 M) for osmotic stress. PDA without amendments was used as control. Mycelial plugs (5 mm diameter) were removed from the edge of 2-day-old colonies of each isolate and placed hyphal side down into the center of PDA plates. After 36 h, the colony diameter was measured using a digital caliper, with four measurements from each plate. The growth inhibition rate (%) was calculated using the following formula: (diameter of untreated colony grown on PDA -diameter of colony grown on PDA with inhibitor treatment)/ diameter of untreated colony grown on PDA × 100%.
As for the NADPH oxidase inhibitor diphenylene iodonium (DPI), the L. theobromae-inoculated shoots were treated with 5 ml of DPI [dissolved in dimethyl sulfoxide (DMSO) and then diluted with water] at a concentration of 0.4 μm at 12 and 24 h after inoculation. Some inoculated peach shoots were mock treated with 0.04% DMSO. All the assays were independently performed in triplicate.

RNA Extraction, cDNA Synthesis, and Gene Expression
RNA extraction, cDNA synthesis, and gene expression were conducted following Zhang et al. (2020). The two genes, translation elongation factor 2 (PpTEF2; Gao et al., 2016;Zhang et al., 2020) and tubulin (LtTUB; Zhang et al., 2020), were used as internal standards to normalize gene transcripts of L. theobromae and peach, respectively. The primers used for quantitative real-time PCR (qRT-PCR) are detailed in Supplementary Table S1. The relative expression was calculated using the comparative 2 −ΔΔCT method (Livak and Schmittgen, 2001) and expressed as means ± SD. The experiments were conducted with three independent biological replicates, each with four technical replicates.

Gene Cloning and Identification
For cloning and identification of LtAP1, the strain JMB122 was cultured on PDA plates for 36 h, and then the hyphae were collected for genomic DNA extraction following Wang et al. (2011). The putative LtAP1 protein sequences were obtained using orthologs of AP1 protein sequences of S. cerevisiae (Kuge and Jones, 1994) and M. oryzae (Guo et al., 2011) as BLASTP queries against the L. theobromae genome assembly (Félix et al., 2019), and one putative LtAP1-encoding gene was obtained from the genome assembly of L. theobromae. To confirm the presence of LtAP1 in JMB122, the full length of LtAP1 was amplified by PCR with primers FD120/FD121 (Supplementary Table S1). Open reading frames (ORF) and exon/intron positions in LtAP1 were determined by comparison with LtAP1 genomic DNA and cDNA sequences.
The predicted LtAP1 protein sequences from JMB122 were used to find orthologs in GenBank. The protein sequences of LtAP1 and its orthologs from various fungal species were aligned using Clustal X 1.81 (Thompson et al., 1997), and then, a phylogenetic tree was constructed using the neighborjoining method with 1,000 bootstrap replications in MEGA 6.0 software (Tamura et al., 2013).

Targeted Gene Disruption
The LtAP1 knockout transformants were obtained using homologous recombination and the CRISPR/Cas9 approach (Ma et al., 2018;Zhang et al., 2020). The upstream (1,688 bp) and downstream (1,722 bp) fragments of the LtAP1 gene of strain JMB122 and a fragment of the hygromycin B resistance phosphotransferase gene (HPH, 1,423 bp) cassette in the pBHt2 vector were amplified separately. As illustrated in Supplementary Figure S1, a 5′ fragment of LtAP1 (1,722 bp) amplified with primers 1F/1R was fused with an HY/g (917 bp) fragment amplified with primers 2F/2R to generate a construct 5′LtAP1::HY/g; meanwhile, a 3′ fragment of LtAP1 (1,722 bp) amplified with primers 4F/4R was fused with a h/YG (966 bp) fragment amplified with primers 3F/3R to produce a construct h/YG::3′LtAP1. As shown in Supplementary Figure S2, the pmCas9 empty vector was digested with Esp3I FastDigest (Thermo scientific, United States). A 20 bp fragment ahead of NGG in the exon region of LtAP1 was selected for singleguide RNA (sgRNA) design, and its specificity was tested against the L. theobromae genome assembly. The sgRNA sequence was synthesized using primers adapted with sticky ends at the 5′ end (Supplementary Table S1), then inserted into the digested pmCas9 vector by T4 DNA ligase (Thermo Scientific, United States). The inserts in plasmids were then confirmed by sequencing.

Genetic Complementation
The complementation strains were obtained using homologous recombination (Ma et al., 2018). As displayed in Supplementary Figure S1, the full-length ORF of LtAP1 carrying its native promoter region (1,500 bp genomic sequence upstream of the ATG start codon) but without stop codon was amplified with primers 8F/8R from genomic DNA of strain JMB122 and used for genetic complementation of ΔLtap1. The amplified PCR fragment was fused with a neomycin resistance gene (NEO) cassette under the control of the Aspergillus nidulans trpC promoter and terminator, conferring resistance to G418 from plasmid pCETNS. The LtAP1::NEO construct was transformed into protoplasts prepared from the mutant ΔLtap1-8. The resultant transformants were recovered from the medium amended with 100 mg ml −1 G418 (Sigma, United States) and screened by PCR with primers 5F/5R.

Virulence Assay
The virulence assay was conducted as previously described (Zhang et al., 2020). The lesion sizes were recorded 5 days post-inoculation (dpi). Green bark tissues within 0.5-1.0 cm of a wound site were sampled, and immediately placed in liquid nitrogen and stored at −80°C until further analysis. Relative amounts of fungal DNA represented by cycle threshold of L. theobromae internal transcribed spacer 1 (LtITS1) were compared to peach-derived elongation factor 1α (PpEF1α, reference gene) using the comparative cycle threshold (2 −ΔΔCT ) method (Svetaz et al., 2017). The primers are shown in Supplementary Table S1. Each treatment was tested on 15 peach shoots, and the virulence assay was independently repeated three times.

Measurement of Superoxide Anion and Hydrogen Peroxide
Absorbance was measured on a spectrophotometer (UV-2450, Shimadzu, Japan

Statistical Analysis
Data were subjected to ANOVA at p < 0.05. The student's t-test was used to test for significant differences of two-sample treatments at p < 0.05 or p < 0.01. Duncan's multiple range test (p < 0.05) was used to separate means when there were more than three treatments, and a significant difference was found in the ANOVA.

Cloning and Identification of LtAP1, a YAP1 Homolog in Lasiodiplodia theobromae
The LtAP1 genomic DNA and cDNA sequences were obtained using primer set FD120/FD121 with the genomic DNA and cDNA of strain JMB122 as templates. The results showed that the LtAP1 gene contained a 1,945 bp coding sequence with a 47 bp intron. The LtAP1 gene (deposited in GenBank with accession number MN933613.1) was predicted to encode a 612 amino acid-long protein that displayed 46 and 43% overall identity with ScYAP1 and MoAP1, respectively. Multiple sequence alignment revealed that AP1s had widely conserved domains: an N-terminal bZIP DNA-binding domain and a nuclear export signal (NES) embedded in a C-terminal cysteine-rich domain (c-CRD; Supplementary Figure S3). Figure S4) demonstrated that AP1-like proteins were evolutionarily conserved among filamentous fungi and separated from the ScAP1 clade. The LtAP1 amino acid sequence had 56% identity with the AP1 homolog in Alternaria alternata (Supplementary Figure S4).

Generation of LtAP1 Deletion and Complementation Strains
To investigate the biological function of LtAP1, we knocked out the gene. The mutants were confirmed by PCR. The primers 6F/6R and 7F/7R amplified two DNA fragments of 2,747 and 2,714 bp, respectively, from genomic DNA of the obtained ΔLtap1 transformants, while no fragment was obtained from the WT strain, indicating that the LtAP1 gene was successfully deleted and replaced by the HYG gene in the ΔLtap1 transformants (Supplementary Figure S5). Furthermore, the authenticity of transformants was screened by PCR with primers 5F/5R, and no fragment was amplified, indicating that these transformants were positive deletion mutants. We obtained seven positive transformants, and two (ΔLtap1-8 and -10) were analyzed further.
A 1,178 fragment was amplified from the genomic DNA of complemented strains using primer set 5F/5R, while no fragment was obtained from the knockout transformants, indicating that the WT allele could be re-introduced into the ΔLtap1 transformants to generate complemented strains (Supplementary Figure S5). We obtained six strains, and strain ΔLtap1/AP1 was used in further analyses. The Role of LtAP1 in Mycelial Growth The mycelial growth rate of ΔLtap1 mutant lines was reduced by 30% compared to the WT strain ( Figure 1C). As well, the ΔLtap1 mutant showed apparent defects in radial growth and aerial hyphal (Figures 1A,B). In contrast, both phenotypes were recovered in the ΔLtap1/AP1 strain. The result indicated that the loss of LtAP1 impaired the vegetative growth of L. theobromae.

Effect of LtAP1 on Response to Different Exogenous Stresses
To evaluate whether LtAP1 can mediate adaptation to exogenous stress, we inoculated mycelial plugs of different genotypes on PDA plates containing cell wall damaging agents (Congo red or CFW), osmotic stress agents (KCl, sorbitol, or glucose), and a cell membrane damaging agent (SDS). After 36 h, the mycelial growth in ΔLtap1 mutants was significantly reduced in Congo red-, KCl-, sorbitol-, glucose-, and CFW-treated plates, while the diameter of ΔLtap1 mutants was significantly increased in sorbitol-and glucose-amended PDA plates compared to the WT (Figure 2). No significant difference was observed for sensitivity to SDS between ΔLtap1 mutants and the WT (Figure 2). In all cases, the mycelial morphology and colony diameter of ΔLtap1/AP1 under exogenous stress treatments were restored to the WT level (Figure 2).
When exposed to H 2 O 2 , cumene H 2 O 2 , TBHP, and menadione treatments, the ΔLtap1 mutants were much more sensitive to 2.5 mm H 2 O 2 , 0.68 mm cumene H 2 O 2 , 0.5 mm TBHP, and 0.1 mm menadione than the WT (Figures 3A,B). The ΔLtap1 mutants showed a substantial growth reduction compared to the WT at 36 hpi, with more than 90% reduction in H 2 O 2 and TBHP treatments, and approximately 60% reduction in cumene H 2 O 2 and menadione treatments ( Figure 3B). In the ΔLtap1/AP1 strain, the stress resistance of strain JMB122 was rescued to the WT level (Figure 3). Further, we tested the transcriptional change of LtAP1 in WT to exogenous oxidants H 2 O 2 and TBHP treatment. When compared with untreated mycelia at the initial time point, exposure to 2.5 mm H 2 O 2 increased the transcripts of LtAP1 quickly at 15 min, peaking at 45 min, followed by a sharp reduction to the end of monitoring (120 min; Figure 3C). Similarly, the expression of LtAP1 was upregulated rapidly but peaked at 30 min under 0.5 mm TBHP treatment ( Figure 3D).
Additionally, we also tested the involvement of LtAP1 in nitrosative stress tolerance. The mycelial growth of the ΔLtap1 strain was significantly reduced in the SNP treatment compared with the WT (Figure 4). Moreover, the growth inhibition of the ΔLtap1 mutant was higher in the treatments with SNP and H 2 O 2 together than in the single treatments with SNP or H 2 O 2 (Figure 4).

Pathogenicity of the LtAP1 Mutant Strain on Peach Shoots
Pathogenicity assays on detached current-year peach shoots revealed that the ΔLtap1 strains caused small brown necrotic lesions and invisible gum release at the site of fungal inoculation, when compared with the WT at 5 dpi, the last observation time ( Figure 5A). The ΔLtap1/AP1 induced necrotic lesions at a rate and magnitude comparable to the WT ( Figure 5A). Quantitative analysis demonstrated that the size of lesions induced by the ΔLtap1 mutants was about 43% of that caused by the WT (Figure 5B). Furthermore, the relative fungal biomass (as revealed by qPCR) in infected peach shoots of the ΔLtap1 mutants was significantly less than that of the WT (Figure 5C). The lesion sizes and fungal biomass in the ΔLtap1/AP1 straininoculated peach shoots were rescued to WT levels (Figures 5B,C). were significantly increased, respectively, with 1.1-and 2.6-fold higher levels in the ΔLtap1 mutant-inoculated shoots than the controls (Figures 6A,B). Furthermore, we tested whether LtAP1 was involved in the regulation of ROS production during infection. The transcripts of core ROS production-related genes, PpRBOHs, were examined. Our data showed that the expression levels of both PpRBOHD and PpRBOHF were significantly higher in shoots inoculated with ΔLtap1 mutant than the WT (Figures 6C,D).

Effects of Prevention of ROS Generation on Pathogenicity of the LtAP1 Mutants
To elucidate the role of LtAP1-modulating oxidative stress tolerance in fungal pathogenicity, an NADPH oxidase inhibitor, DPI, was used in virulence testing of the ΔLtap1 mutants. We observed that the ΔLtap1 mutants induced much larger brown necrotic lesions and more visible gum release in the DPI-treated shoots than the mock-treated ones at 5 dpi (Figures 7A,B). Moreover, in the DPI-treated shoots, the size of lesions induced by the ΔLtap1 mutants was only about 79% of that caused by the WT (Figure 7B).

Role of LtAP1 in the Expression of ROS Detoxification-Related Genes in L. theobromae
To identify genes regulated by LtAP1, the transcripts of genes related to antioxidants (glutaredoxin and thioredoxin) were analyzed in the WT and the ΔLtap1 mutant exposed to 2.5 mm H 2 O 2 or distilled water for 1 h. Relative transcripts of the core genes of both the glutaredoxin system (LtGPX3 and LtGLR1) and the thioredoxin system (LtTRX2, LtTSA1, and LtTRR1) were significantly lower in the ΔLtap1 mutant in the absence of H 2 O 2 ( Figure 8A). Under H 2 O 2 treatment, the expression of all tested genes was consistently and significantly further decreased to 68 to 100% in the ΔLtap1 mutant, as compared to the WT (Figure 8B).
To further elucidate the possible mechanism behind the impairment of oxidative stress response and pathogenicity in the ΔLtap1 mutant, transcript levels of genes in the glutaredoxin and thioredoxin systems were assayed for ΔLtap1 or WT infected tissues. Indeed, the inactivation of LtAP1 led to significant reductions of all tested genes expression, ranging from 49 to 90% in the shoots inoculated with ΔLtap1 relative to the WT at 5 dpi ( Figure 8C).

Effect of LtAP1 Deletion on Plant Defense Response
ROS often act as signaling molecules to activate defense-related genes, such as pathogenesis-related (PR) genes, to enhance plant defense response (Camejo et al., 2016). To further assess whether PR genes were activated by the ΔLtap1 mutant inoculation, transcripts of several PR genes, including PpPR1a, PpPR8, PpPR10-1, PpPR10-4, PpDFN1 (Defensin 1, PR12 family), and PpLTP1 (Lipid-transfer protein 1, PR14 family), were examined at 5 dpi in the peach shoots inoculated with the ΔLtap1 mutant or WT. The transcripts of all tested PR genes Transcript abundance of ROS production-related genes PpRBOHD and PpRBOHF in infected peach shoots at 5 dpi. Relative transcript levels of genes compared with that of the control using reference gene PpTEF2 for normalization. All data are means ± SD of three biological replicates. Asterisks indicate the significant difference between two genotypes for the same parameter comparison at p < 0.01.
were significantly higher in shoots inoculated with ΔLtap1 than WT (Figures 9A-F). Notably, the transcripts of PpPR10-4 and PpLTP1 were 2.7-and 4.0-fold higher, respectively, in tissues inoculated with the ΔLtap1 than the control (Figures 9D,F).
In addition, the transcripts of plant defense-related gene PpPAL1 (Phenylalanine ammonia lyase 1) were also significantly induced and were 2.0-fold higher in the ΔLtap1-inoculated shoots than those with the control (Figure 9G). The transcripts of PpICS1 (isochorismate synthase 1) and PpNPR1 (nonexpressor of pathogenesis-related gene 1), which were required for SA biosynthesis and signal transduction, were also significantly upregulated in the peach shoots inoculated with the ΔLtap1 mutant than the WT (Figures 9H-I).

DISCUSSION
The peach gummosis pathogen, L. theobromae, is a destructive threat to peach harvests (Wang et al., 2011), and infection events at the molecular level need deeper investigation. In a previous study, L. theobromae infection caused an oxidative burst in peach shoots and promoted expression of LtAP1 and other genes associated with the ROS scavenging system (Zhang et al., 2020). Subsequently, we attempted to uncover how plant infection is regulated by an oxidative stress regulator, YAP1, in L. theobromae. Eukaryotic microorganisms have stress-protective functions against a variety of adverse conditions. We first compared the growth performance of the ΔLtap1 mutant and the WT grown on media supplemented with different exogenous chemicals to mimic environmental stresses. Our results suggest that in L. theobromae, LtAP1 was involved in response to various stresses. Deletion of LtAP1 led to decreased sensitivity to osmotic and cell wall inhibitors, indicating that LtAP1 negatively regulated the sensitivity to osmotic pressure and the maintenance of cell wall integrity in L. theobromae (Figure 2). Likewise, in C. gloeosporioides, ΔCgap1 mutants had higher resistance to sorbitol than the WT (Li et al., 2017). However, in F. graminearum, the Fgap1-deficiency mutant exhibited increased sensitivity to sorbitol or NaCl-induced stresses (Montibus et al., 2013). In addition, our oxidative stress tests indicated that ΔLtap1 mutants were hypersensitive to H 2 O 2 , cumene H 2 O 2 , and TBHP, as well as menadione (Figures 3A,B). These results suggest that LtAP1 plays a vital role in the regulation of fungal response to oxidative stress. This is consistent with studies on M. oryzae (Guo et al., 2011), F. graminearum (Montibus et al., 2013), A. alternata (Lin et al., 2009), and C. gloeosporioides (Sun et al., 2016), where the mycelial growth of the respective mutant was severely reduced by oxidative stress compared to their respective WT. Moreover, the expression levels of LtAP1 were significantly upregulated under the oxidant treatments (Figures 3C,D). A similar finding was observed in C. gloeosporioides (Sun et al., 2016) and Monilinia fructicola (Yu et al., 2017). The results suggest that the fungal YAP1s transcription factors are highly conserved for oxidative stress response in different species.
Interestingly, the ΔLtap1 mutant showed a significant reduction of mycelial growth with the SNP treatment. The growth suppression of the ΔLtap1 mutant after treatment with SNP plus H 2 O 2 was higher than of either SNP or H 2 O 2 alone (Figure 4), which showed an additive effect between SNP and H 2 O 2 . ROS can react with nitric oxide and generate toxic reactive nitrogen species (RNS), such as peroxynitrite (Marroquin-Guzman et al., 2017). Hence, LtAP1 is likely an essential player in oxidative and nitrosative stress adaptation.
Pathogenicity tests revealed that the ΔLtap1 mutant induced smaller necrotic lesions, less gum release, and decreased pathogen biomass than WT (Figure 5), suggesting that LtAP1 is essential for growth and virulence of the necrotrophic fungus L. theobromae on peach shoots. Similarly, in the biotrophic U. maydis and necrotrophic A. alternata, deletion of AP1 failed to incite necrotic lesions (Molina and Kahmann, 2007;Lin et al., 2009). In the hemibiotrophic pathogen M. oryzae, Moap1 is essential to the growth of invasive hyphae for successful infection (Guo et al., 2011). In hemibiotrophic C. gloeosporioides, ΔCgap1 mutant showed severely attenuated virulence on poplar leaves (Sun et al., 2016) and could not induce lesions on mango fruits (Li et al., 2017). However, in necrotrophic B. cinerea (Temme and Tudzynski, 2009) and F. graminearum (Montibus et al., 2013), the deletion of AP1 did not show noticeable effects on pathogenicity, indicating that YAP1 homologs are not necessary for virulence in all pathogenic fungi studied. This might be because fungal virulence associated with YAP1 differs in the types of associations established between specific fungi and plant hosts. A better and deeper understanding of the mechanisms of pathogen virulence associated with YAP1 homologs is needed.
It is well known that a major mechanism of plant defense is the production of ROS against pathogens attack. Therefore, fungal pathogens need robust strategies for ROS scavenging, which involves YAP1 homologs (Segal and Wilson, 2018). In the L. theobromae-infected peach shoots, the expression of LtAP1, the glutaredoxin system genes (LtGPX3 and LtGLR1), and the thioredoxin system genes (LtTRX2, LtTSA1, and LtTRR1) was markedly upregulated, which was perhaps to scavenge ROS derived from the host (Zhang et al., 2020). In this study, we found higher ROS accumulation at the inoculation site with the mutant than with the WT (Figures 6A,B). Moreover, the ΔLtap1 mutants were hypersensitive to exogenous oxidative stress (Figures 3A,B). This suggested that LtAP1-modulating oxidative stress tolerance might play a crucial role in fungal pathogenicity. To further investigate the link between LtAP1 modulation of oxidative stress tolerance and fungal pathogenicity, we used an NADPH oxidase inhibitor, DPI, to prevent ROS generation. The results clearly showed that the DPI treatment increased necrotic lesion size and enhanced gum release in the shoots inoculated with the ΔLtap1 mutants as compared to the mock control, suggesting that the pathogenicity of ΔLtap1 mutants was partially restored (Figures 7A,B). Overall, LtAP1 modulation of oxidative stress tolerance, at least in part, contributed to the pathogenicity of L. theobromae. Similarly, the ΔAaAP1 mutant of necrotrophic A. alternata was hypersensitive to oxidants, and its pathogenicity was rescued by the NADPH oxidase inhibitor treatment (Lin et al., 2009). In biotrophic U. maydis, H 2 O 2 was markedly accumulated at sites inoculated with the Umap1 mutant, and inhibition of the plant NADPH oxidase decreased ROS accumulation and restored the virulence of the mutant, suggesting that Umap1 acts in neutralizing the ROS generated by the maize NADPH oxidase (Molina and Kahmann, 2007).
In filamentous fungi, YAP1 homologs are major regulators of the antioxidant response, but YAP1 homologs involve a wide array of processes by regulating genes involved in ROS scavenging (Mendoza-Martínez et al., 2020). The expression of such genes, such as the core glutaredoxin system genes (LtGPX3 and LtGLR1) and thioredoxin system members (LtTRX2, LtTSA1, and LtTRR1), was dramatically downregulated in the ΔLtap1 mutant treated with water or H 2 O 2 (Figures 8A,B), indicating that LtAP1 acts as a major regulator in the antioxidant (C) RNA samples were collected from the border of L. theobromae-colonized peach shoots at 5 dpi. The transcript levels were normalized with LtTUB and are displayed in relation to the transcript levels in the corresponding WT samples (which was therefore set to one). The values are means ± SD of three biological replicates. Asterisks indicate significant differences for genes between two genotypes, with *p < 0.01 and **p < 0.05. system. Similarly, the transcription factor AaAP1 could activate glutaredoxin (AaGPX3 and AaGLR1) and thioredoxin systems (AaTSA1 and AaTRR1) to cope with oxidative stress (Yang et al., 2016;Ma et al., 2018). The thioredoxin MoTrx2 was found to be a target of the transcription factor MoAP1 in M. oryzae, and ΔMotrx2 mutant displayed higher ROS levels and lower POD and laccase activities . However, in F. graminearum, the expression of three CAT-and two Cu/ ZnSOD-encoding genes was downregulated in the Fgap1 mutant (Montibus et al., 2013). Likewise, EfAP1 in Epichloe festucae was required for expression levels of the CAT gene (Cartwright and Scott, 2013). As a whole, it suggests that YAP1 homologs could regulate/target different antioxidant system-related genes to overcome oxidative stress in different fungi.
During infection, the transcript levels of genes in glutaredoxin and thioredoxin systems were significantly downregulated in the ΔLtap1 mutant ( Figure 8C). Concomitantly, the contents of O 2 -and H 2 O 2 and transcripts of PpRBOHs were markedly higher in shoots inoculated with the ΔLtap1 mutant, demonstrating a reduced ability of the ΔLtap1 mutant to scavenge overproduced ROS during the interaction (Figure 6). It is speculated that LtAP1 is likely to modulate glutaredoxin and thioredoxin systems to scavenge host-derived ROS. Similarly, AaAP1 could also modulate glutaredoxin and thioredoxin systems to cope with oxidative stress (Lin et al., 2009;Yang et al., 2016;Ma et al., 2018).
ROS serving as the primary signaling molecule during pathogens attack can activate an array of defense responses, such as induction of defense-related genes (Qi et al., 2017;Segal and Wilson, 2018). We observed significantly higher expression levels of PR genes, such as PpPR1a, PpPR8, PpPR10-1, PpPR10-4, PpDFN1, and PpLTP1, in the ΔLtap1 mutant-infected shoots in comparison with the WT treated (Figures 9A-F). Furthermore, the transcripts of SA biosynthesis and signaling-related genes (PpPAL1, PpICS1, and PpNPR1) were also significantly upregulated after ΔLtap1 mutant inoculation than the WT (Figures 9G-I). The accumulation of PR proteins and SA-mediated plant defense response might assist in limiting disease development, which was reflected by the reduced lesion size, gum release, and fungal biomass at inoculation sites with the ΔLtap1 mutant (Figure 5). Similarly, the thioredoxin MoTrx2 regulated by MoAP1 played an essential role in the ROS scavenging during host invasion and in the suppression of the rice defense response, in which the transcript levels of plant defense genes were markedly higher in rice cells infected with the ΔMotrx2 mutant than the control . The rice cells inoculated with the ΔModes1 mutant exhibited strong defense responses accompanied by the accumulation of ROS and PR genes transcript in neighboring tissues, indicating that DES1 is required to suppress the host basal defenses (Guo et al., 2011). Taken together, we propose that the restricted expansion of the ΔLtap1 mutant in peach shoots is partly caused by the defect in active suppression of peach defense response.
In summary, we cloned and characterized the LtAP1 gene, which encodes a homolog of yeast YAP1. Our experiments demonstrated that LtAP1 was valuable for mycelial growth, stress response, and pathogenicity. We found that LtAP1 was a key regulator of oxidative stress response, acting in activating fungal glutaredoxin and thioredoxin systems, and suppressing plant defense responses during infection. The prevention of ROS production could partially restore pathogenicity of ΔLtap1 mutant. LtAP1 plays a central role in adjusting ROS homeostasis between fungal pathogen and plant host and is necessary for full virulence of L. theobromae. This study advances our understanding of the link between oxidative stress response, ROS detoxification, and virulence in L. theobromae. Given the critical roles of LtAP1 in L. theobromae-induced peach gummosis, it would be urgent to identify its potential targets in the downstream network, which would be helpful for future disease management. Expression pattern of SA biosynthetic (PpPAL1 and PpICS1) and signaling (PpNPR1) genes. Relative transcript levels of genes compared with that of the control using reference gene PpTEF2 for normalization. Values are means ± SD of three biological replicates. Asterisks indicate a significant difference between two genotypes for genes at p < 0.01.