Systematic characterization of the ubiquitin-specific proteases in Magnaporthe oryzae

Ubiquitin-specific protease (UBP) family is the largest group of deubiquitinases, which plays important roles in eukaryotic organisms. Comprehensive analysis of UBP genes has not been conducted in the plant pathogenic fungi. In this study, 11 putative UBP genes were identified and characterized in the rice blast fungus Magnaporthe oryzae. Expression profile analysis showed that UBP3, UBP6, UBP12 and UBP14 were highly expressed in different tissues of M. oryzae. In all ubp mutants, especially Δubp3, Δubp12 and previously reported Δubp14, the ubiquitination levels were evidently elevated, which is consistent with their molecular roles in de-ubiquitination. The Δubp1, Δubp3, Δubp4, Δubp8 and Δubp14 mutants were reduced in colony growth. Most of the ubp mutants were severely reduced in conidia production capacity, indicating important roles of the UBPs in conidia formation. Except for Δubp2 and Δubp16, all of the other mutants were decreased in virulence to host plants and defective in invasive growth. These ubp mutants also induced massive ROS accumulation in host cells. We also found that the UBPs may function as both positive and negative regulators in stress response and nutrient utilization of M. oryzae. Collectively, UBPs are important for development, stress response, nutrient utilization and infection of M. oryzae.


Background
Ubiquitination is one of the most important posttranslational modifications, which plays central roles in diverse cellular pathways such as cell-cycle progression, signal transduction, DNA repair and endocytosis and apoptosis (Hershko and Ciechanover 1998;Pickart and Eddins 2004;Dikic et al. 2009). The ubiquitin molecules are catalyzed by a cascade of three enzymes: the ubiquitin activating enzyme (E1), the ubiquitin conjugating enzyme (E2) and the ubiquitin ligase (E3) (Pickart and Eddins 2004). The fate of ubiquitinated substrate proteins mainly depends on the number of conjugated ubiquitin and the type of linkage. Proteins linked by Lys48 bearing four or more ubiquitin chains are mostly recognized by the 26S proteasome complex system for degradation. Attachment of ubiquitin to cellular proteins as different linkage types are involved in distinct cellular processes (Pickart and Eddins 2004).
As a reversible event, ubiquitins can be removed from proteins, which is also important for regulation of different cellular processes. Cellular ubiquitin monomers are kept in homeostasis by processing from ubiquitin precursors or deubiquitinating from protein substrates (Amerik and Hochstrasser 2004;Nijman et al. 2005). These processes are regulated by the de-ubiquitinating enzymes (DUBs), which can specifically remove ubiquitin moiety. The cleavage of ubiquitin from proteins by DUBs can affect the activity and fate of substrates (Wilkinson 1997;Amerik and Hochstrasser 2004;Crosas et al. 2006). Among several types of DUBs, the ubiquitin-specific proteases (UBPs) are highly conserved and the largest subfamily among eukaryotic organisms (Wilkinson 1997;Amerik and Hochstrasser 2004;Nijman et al. 2005), which commonly possess a triad of catalytic residues in conserved cysteine and histidine boxes (Hu et al. 2002). Ubiquitin C-terminal hydrolases (UCHs) contain a similar triad of catalytic residues in two conserved cysteine and histidine boxes (Johnston et al. 1999;Amerik and Hochstrasser 2004). Other DUBs include the ovarian tumor proteases (OTUs), Machado-Joseph disease protein domain proteases (MJDs), etc. (Balakirev et al. 2003;Burnett et al. 2003;Scheel et al. 2003;Nanao et al. 2004).
There are 16 UBP genes in Saccharomyces cerevisae (Wilkinson 1997) and 27 UBP genes in Arabidopsis thaliana (Liu et al. 2008). Studies on S. cerevisae revealed UBPs are involved in nutrient utilization, energy metabolism, sexual reproduction, and stress responses (Kahana 2001;Enyenihi and Saunders 2003;Dudley et al. 2005;Auesukaree et al. 2009). In plants, UBPs have been proved to play roles in cell proliferation, root hair elongation, mitochondria morphogenesis, pollen development and transmission, and abscisic acid (ABA)-mediated resistance to salt and drought stress, etc. (Doelling et al. 2001;Liu et al. 2008;Du et al. 2014;Xu et al. 2016;Zhao et al. 2016). However, few studies have addressed systematic analysis on the roles of all members of the UBP gene family, although roles of single UBP gene was reported (Liu et al. 2008;Wu et al. 2019).
Magnaporthe oryzae is a fungus threatening rice production worldwide, and also a model fungal plant pathogen (Wilson and Talbot 2009;Yan and Talbot 2016). This fungus forms a specialized infection structure appressorium on the host surface, and penetrates into host cells (de Jong et al. 1997;Dixon et al. 1999;Thines et al. 2000). In the host cells, the fungus forms invasive hyphae to establish biotrophic colonization and eventually converts into necrotrophic growth for spreading (Kankanala et al. 2007). Previous study has found that both M. oryzae UBP14 and UBP4 play key roles in growth, conidiation, stress response, and carbon sources utilization, and meanwhile, are required for virulence of the pathogen Que et al. 2019). In order to systematically determine functions of UBPs, we identified and characterized roles of all M. oryzae UBP genes in development, stress response, carbon source utilization, as well as the virulence.
To determine the phylogenetic relationship of the UBPs among different species, a Neighbor-joining phylogenetic tree was constructed based on the alignments of fulllength UBP protein sequences from S. cerevisiae, Aspergillus nidulans, Neurospora crassa, Fusarium graminearum and M. oryzae (Fig. 1b). In the phylogenetic tree, all sequences were classified into 11 groups. The phylogenetic dendrogram shows that, for most cases, M. oryzae UBP proteins are evenly distributed among these clades, implying that genes in the clade may have evolved from a copy of their common fungal ancestor (Fig. 1b).
Expression patterns of UBP genes at different developmental stages of M. oryzae The gene expression pattern can provide vital clues for gene's function. To characterize the expression patterns of UBP genes, we analyzed their transcription levels in different tissues, including mycelia, conidia, appressoria (at 3 hours post inoculation [hpi] or 12 hpi), and invasive hyphae (at 18, 24 or 42 hpi) according to an unpublished RNAseq data. As shown in Fig. 2, UBP3, UBP6, UBP12 and UBP14 showed high expression levels in nearly all tested tissues, while UBP1, UBP13 and UBP15 showed low expression levels in these tissues. In particular, four genes (UBP3, UBP6, UBP12 and UBP14) in mycelia, seven genes (UBP3, UBP4, UBP6, UBP8, UBP12, UBP14 and UBP16) in conidia, most of these genes in appressoria, and five genes (UBP3, UBP6, UBP12, UBP14 and UBP16) in invasive hyphae showed high expression levels. Taken together, M. oryzae UBP genes exhibited different expression profiles in different tissues, indicating diverse biological functions of UBP genes in developmental and infection processes.

Single deletions of the UBP genes increase ubiquitination levels in M. oryzae
To characterize the functions of UBP genes in M. oryzae, single gene deletion mutant for each UBP gene was generated by using a split-PCR strategy (Goswami 2012). At least two independent null mutants with identical phenotypes were obtained for each UBP gene, and all deletion mutants were confirmed by polymerase chain reaction (PCR) and gene expression level detection (Additional file 1: Figure S1 and Additional file 2: Figure  S2).
In order to test whether deletion of single UBP gene affects total ubiquitination level in M. oryzae, all of the single ubp mutants (except for Δubp14) were tested for represented. The color (from green to red) indicates gene expression intensity from low to high. HY: Mycelial hyphae; CO: Conidia; 3H AP: appressoria at 3 hpi; 12H AP: appressoria at 12 hpi; 18 IH: invasive hyphae at 18 hpi; 24 IH: invasive hyphae at 24 hpi; 42 IH: invasive hyphae at 42 hpi changes of the ubiquitination level (Fig. 3). As expected, all of the mutants displayed increased ubiquitination levels. Especially, the Δubp3 and Δubp12 mutants, as well as our previously reported Δubp14 , each of them showed significant increase of the ubiquitination level compared to the wild type. The Δubp4, Δubp8, Δubp13, Δubp15 and Δubp16 mutants showed a moderate increase in ubiquitination level. These data reflected that all of the UBP genes are indeed involved in deubiquitinating process in M. oryzae.

Characterization of vegetative growth of the ubp mutants
All of the ubp mutants were examined for phenotypic changes in colony growth, mycelium formation, conidium formation, appressorium formation, and invasive growth. As summarized in Table 1, the deletion mutants of each UBP gene displayed distinctive effects on colony growth, conidiation, appressorium formation and infection-related morphogenesis. Among 11 mutants, Δubp1, Δubp3, Δubp4, Δubp8 and Δubp14 mutants displayed respectively 22.0%, 9.4%, 6.1%, 12.7% and 25.3% reductions of colony diameter on oatmeal tomato agar (OTA) plates (Fig. 4a, b and Table 1). Correspondingly, significant reductions of mycelial cell length were observed in these mutants (Fig. 4c, d). These results indicated that UBP1, UBP3, UBP4, UBP8 and UBP14 are involved in the regulation of vegetative growth in M. oryzae.

Characterization of conidium production of the ubp mutants
Conidia play important roles in the disease cycle of the rice blast fungus. Most UBP genes were found here to be involved in conidiation of M. oryzae. The Δubp1, Δubp3, Δubp4, Δubp6, Δubp8, Δubp12, Δubp13, Δubp14 and Δubp16 mutants were markedly reduced in conidiation compared with the wild type P131 strain ( Fig. 5a and Table 1). For the Δubp8 and Δubp14 mutants, conidia were very rare with only few visible spores following examination of more than ten 10-day-old culture plates. However, the Δubp2 and Δubp15 mutants showed normal conidium production ( Fig. 5a and Table  1). Accordingly, conidiophore formation of the Δubp1, Δubp3, Δubp4, Δubp6, Δubp8, Δubp12, Δubp13, Δubp14 and Δubp16 mutants were severely reduced (Fig. 5b). These results showed that most members of the UBP genes regulate conidiation in M. oryzae.
Except for Δubp8, no other ubp mutants were blocked in conidium germination, and most of them did not show significant changes in appressorium formation. The Δubp8 mutant displayed a lower percentage of appressorium formation compared with the wild-type strain and other ubp mutants (Table 1).

Virulence and infection process of the ubp mutants
To further test whether these M. oryzae UBP genes are involved in virulence, conidia of each of the ubp mutants were sprayed onto susceptible rice seedlings. As shown in Fig. 6a, very few lesions were formed on the rice Fig. 3 Anti-ubiquitin immunoblot analysis of ubp mutants. A monoclonal antibody against ubiquitin was used (1:5000). The coomassie brilliant blue (CCB) stain was used for protein loading seedlings inoculated with Δubp1, Δubp3, Δubp4, Δubp6, Δubp8 and Δubp14 mutants. In contrast, rice seedlings sprayed with Δubp2 and Δubp16 mutants developed numerous typical rice blast lesions similar to that inoculated with the wild-type M. oryzae strain (Fig. 6a). When inoculating the mycelial agar plugs onto the wounded rice leaves, all of the virulence-decreased mutants produced evidently restricted spreading of the lesions (Fig.  6b). This result indicated that UBP1, UBP3, UBP4, UBP6, UBP8 and UBP14 play key roles in virulence to host. Since the Δubp1, Δubp3, Δubp4, Δubp6, Δubp8 and Δubp14 mutants exhibited normal appressoria formation, we examined the ability of the appressoria in the penetration of the barley epidermal cells. At 24 hpi, only Δubp14 showed severe reduction of penetration, while no evident changes were found for the other ubp mutants, with majority of the appressoria in these mutants forming invasive hyphae (Fig. 6c). However, at both 24 hpi and 30 hpi, the formation of infection hyphaein the Δubp1, Δubp3, Δubp4, Δubp6, Δubp8 and Δubp14 mutants was significantly retarded, compared with the wild-type strain (Fig. 6c). These results indicate that UBP1, UBP3, UBP4, UBP6, UBP8 and UBP14 regulate virulence of M. oryzae through affecting infection hyphal growth.  Fig. 4 Vegetative growth of the ubp mutants. a The wild-type strain and the ubp mutants were inoculated on OTA plate and cultured at 28°C for 5 days and then photographed. b Colony diameter. Significant differences compared with the wild type are indicated by asterisks (P < 0.05). c Calcofluor White staining of hyphal tips shows the distance of septa. White arrows indicate the cell septa. Bar, 20 μm. d Average cell length of the hyphal tips. Significant differences compared with the wild type are indicated by asterisks (P < 0.05) The ubp mutants led to accumulation of host reactive oxygen species (ROS) Because some ubp mutants were blocked in host invasive growth, we explored whether each of these UBP gene deletion mutants can result in host ROS accumulation. ROS accumulation in the barley epidermis cells infected by each strain at 30 hpi was detected by staining with 3,3′-diaminobenzidine (DAB). The result demonstrated that, except for Δubp2 and Δubp16, the barley epidermis cells infected by all other ubp mutants can be easily stained by DAB, which were detected with abundant reddish-brown precipitates in the infected host cells (Fig. 7a, b). At the same time, host cells infected by the wild-type strain were not well stained by DAB (Fig. 7a, b). These results showed that UBP genes are important for M. oryzae to respond to host cellular oxidative stress, and therefore required for invasive growth.

Members of UBP genes are involved in stress response in M. oryzae
We evaluated the effect of UBP genes disruption on stress tolerance of M. oryzae. The wild-type and ubp mutant stains were inoculated onto the CM plates each containing different stresses, including the cell wall stresses ( and Δubp14 mutants were more sensitive to osmotic stress; and the Δubp3, Δubp12 and Δubp14 mutants were more sensitive to oxidative stress (Fig. 8). Interestingly, we also noticed some mutants were more resistant to different stresses. For example, the Δubp2 and Δubp4 mutants were more resistant to cell wall stresses; the Δubp1, Δubp2 and Δubp4 mutants were more resistant Means and standard errors were calculated from three independent experiments (n > 100). Significant differences compared with the wild type are indicated by asterisks (P < 0.01). b Conidiophore development was observed under light microscopy to oxidative stress (Fig. 8). These data indicated that UBP genes positively or negatively regulate response to different stresses in M. oryzae.

Nutrient utilization was positively or negatively regulated by different UBP genes
We also evaluated whether the disruption of UBP genes can affect carbon source utilization of M. oryzae. The wild-type and ubp mutant strains were cultured on MM plates supplemented with each of glucose, NaAc, ethanol and glycerol as the sole carbon source. After 5 days culture at 28°C, colony diameters of the Δubp4 and Δubp14 mutants were significantly reduced when grown on all of these conditions compared with those grown on the CM plates (Fig. 9). These demonstrated that the carbon sources utilization ability of the Δubp4 and Δubp14 mutants were significantly reduced, suggesting UBP4 and UBP14 are positive regulators of carbon source utilization. In contrast, the Δubp1 was significantly increased in growth in the NaAc and ethanol conditions, and the Δubp8 mutant was increased in growth in all tested conditions (Fig. 9), suggesting UBP1 and UBP8 are negative regulators of carbon source utilization. Therefore, nutrient utilization of M. oryzae can be either positively or negatively regulated by different UBP genes.

Discussion
As one of the largest families of DUBs, the eukaryoticspecific UBP family protein plays important roles in fungal growth and development. Detailed characteristics and functions of UBP genes have been reported in S. cerevisae (Amerik et al. 2000), but not been systematically described in plant pathogenic fungi. Here, we identified and characterized 11 putative UBP genes in the rice blast fungus M. oryzae using a genome-wide analysis. The expression profiles of the UBP genes at different developmental stages of M. oryzae have been evaluated. According to the gene expression patterns, different UBP proteins may have diverse functions. The UBP3, UBP6, UBP12 and UBP14 were highly expressed at different stages, suggesting they might play important roles at these stages. Moreover, some genes showed similar expression patterns in different tissues, suggesting that they might have retained redundant functions in regulating the same biological functions.
In fungi, functions of the UBP genes have been revealed in yeast and other fungi including M. oryzae  Percentage of DAB-stained barley epidermis cells infected by the wild-type and ubp mutant strains. Means and standard errors were calculated from three independent replicates. Significant differences were indicated by asterisks (P < 0.05) Fig. 8 Statistical analysis of growth reduction rates of colony growth of the ubp mutants under different stresses. The growth reduction was calculated through dividing growth diameter of each strain grown under normal condition by the disparity between one grown under normal condition and stress conditions. Means and standard errors were calculated from three independent replicates. Significant differences were indicated by asterisks (P < 0.05) (Amerik et al. 2000;Wang et al. 2018;Que et al. 2019). Sixteen UBP genes are found in S. cerevisae to be involved in nutrient utilization, energy metabolism, sexual reproduction, and stress responses (Kahana 2001;Enyenihi and Saunders 2003;Dudley et al. 2005;Auesukaree et al. 2009). For example, DOA4 is required for carbon utilization, UBP10 is important for nitrogen utilization (Enyenihi and Saunders 2003), UBP15 plays important roles in growth and stress response (Yoshikawa et al. 2011), and UBP14 plays roles in proteasome-mediated proteolysis (Hadari et al. 1992;Amerik et al. 1997;Eisele et al. 2006). Disruption of several S. cerevisiae UBP genes, including DOA4, UBP6, UBP8, UBP10 and UBP14, led to increasement of the ubiquitin levels (Amerik et al. 2000), similar to most of the UBP gene disruption mutants of M. oryzae. The UBP genes were also studied in other fungi. For example, in Cryptococcus neoformans, UBP5 is likely the major deubiquitinating enzyme-coding gene for stress responses, whose disruption mutant is severely reduced in virulence, capsule formation, melanization, sporulation, and is more sensitive to different stresses (Fang et al. 2012). Our previous study found that deletion of M. oryzae UBP14 resulted in loss of virulence, as well as pleiotropic phenotypic defects such as decrease in colony growth, reduction in conidium and appressorium formation, increase in sensitivity to stresses, and reduction in carbon sources utilization . Similarly, another study found that the ΔMoubp4 mutants also lose their pathogenicity, and are blocked in penetration and invasive growth. The conidia produced by the ΔMoubp4 mutants are defective in appressorium formation (Que et al. 2019). In this study, we systemically analyzed functions of the UBP genes in M. oryzae, and found some of the UBP genes also play important roles in nutrient utilization, stress responses, mycelium growth, conidium formation, and infection.
In many fungal pathogens, including M. oryzae, conidiogenesis is a key step in the infection process. Our result showed the Δubp1, Δubp3, Δubp4, Δubp6, Δubp8, Δubp12, Δubp13, Δubp14 and Δubp16 were significantly reduced in conidiation, suggesting that UBP gene family plays key roles in conidium formation. We also found that the UBP genes may function as both positive and negative regulators in stress responses and nutrient utilization of M. oryzae. For example, UBP1, UBP3, UBP8, UBP12 and UBP14 could be positive regulators in different stress responses of M. oryzae, UBP2 and UBP4 were negative regulators in cell wall stress response, and UBP1, UBP2 and UBP4 were negative regulators in oxidative stress response (Fig. 8). Similar situation was also found in nutrient utilization. UBP4 and UBP14 were positive regulators of nutrient utilization, while UBP1 and UBP8 were negative regulators. These data suggest that de-ubiquitination system function in both positive and negative manners in different cellular processes.

Conclusions
In this study, we reported that 11 members of the UBP family are involved in the growth, development, stress responses, nutrient utilization, and pathogenicity of M. oryzae. Given the importance of these proteins, further investigation of the UBP genes and their respective regulatory networks are required, and they may also serve as putative fungicide targets for disease control.

Strains and culture conditions
The wild-type strain P131 and mutant strains of M. oryzae were grown at 28°C on Oatmeal Tomato Agar (OTA) plates. Mycelia cultured in liquid CM at 28°C were used for extracting genomic DNA, RNA, protein, and isolating protoplasts. Colony growth and conidiation were performed as described previously (Chen et al. Fig. 9 Statistical analysis of growth reduction rates of colony growth on different carbon sources compared to growth on CM plates. The wildtype and ubp mutant stains cultured on CM or minimal medium agar containing 1% glucose, 50 mM sodium acetate, 50 mM ethanol, or 50 mM glycerol for 120 h at 28°C. Means and standard errors were calculated from three independent replicates. Significant differences were indicated by asterisks (P < 0.01) 2014). Conidia harvested from 7-day-old OTA cultures were used for testing virulence and observing infection process. Stress sensitivity of different strains were tested at 5 days after incubation on CM plates containing 0.1 mg/mL CFW, 0.2 mg/mL CR, 0.005% SDS, 0.5 M NaCl, or 10 mM H 2 O 2 . Colony diameters were recorded to calculate the growth reduction rates. Carbon source utilization was tested by recording colony diameters of different strains incubated on MM agar plates containing 1% glucose, 5 mM sodium acetate, 2% ethanol, or 2% glycerol as the sole carbon source. Colony diameters were recorded to calculate the growth reduction rates.

Gene deletion transformation
A split-PCR mediated gene deletion strategy was used to delete the UBP genes (Goswami 2012). Gene replacement construct was generated by respectively fusing 1.5kb upstream and 1.5-kb downstream flanking sequences with part of the hygromycin segment, and then transformed into protoplasts of the wild type ) (Additional file 3: Table S1). The deletion transformants were selected by using 250 μg/mL hygromycin B (Roche, USA), and verified by PCR-mediated methods ) (Additional file 3: Table S1).

Virulence test and infection process observation
Conidia suspensions (5 × 10 4 conidia/mL, in ddH 2 O containing 0.025% Tween 20) of different strains were sprayed onto one-month-old rice seedlings (Oryza sativa cv. LTH). The inoculated rice seedlings were incubated at 28°C with full humidity, and the disease lesions were observed and photographed at 5 days post inoculation [dpi]. For infection process observation, the conidia suspensions (1 × 10 5 conidia/mL) were dropped onto the barley leaves, and the leaves were then put into a dark chamber with full humidity at 28°C. Infection processes were observed under a microscope (Nikon Ni90, Japan) at different time points (24 hpi and 30 hpi).

Expression level analysis of UBP genes
To examine UPB gene transcript levels at different tissues of M. oryzae, samples of mycelia, appressoria and infection hyphae were harvested as previously described (Hendy et al. 2019), and were used to extract total RNA for the RNAseq experiment, from which the expression levels of UBP genes were determined by FPKM (Fragments Per Kilobase of transcript per Million mapped reads) values.

Phenotypic characterization of the ubp mutants
Mycelial growth of the different strains was measured on OTA plates after 120 h of incubation at 28°C. Conidiation were calculated from conidia produced on OTA plates as previously reported .
Appressoria formation was observed by dropping conidia suspension (1 × 10 5 conidia/mL) onto a hydrophobic coverslip, then incubating in a moistened chamber at 28°C. Appressoria formation percentage was calculated under a microscopy with at least 100 conidia per replicate and three replicates per experiment.

Staining assays
For CFW staining assay, mycelia were harvested from CM medium and stained with 10 μg/mL CFW (Sigma-Aldrich, USA) for 10 min in the dark. Before observation under a fluorescence microscope (Nikon Ni90 microscope, Japan), the samples were rinsed twice with PBS buffer. For DAB staining assay, the barley leaves were inoculated with different strains by dropping with conidia suspension (1 × 10 5 conidia/mL), then leave samples were stained with DAB solution (1 mg/mL, pH 3.8, Sigma-Aldrich, USA) at 36 hpi. After staining for 8 h, the leave samples were de-stained with an ethanol/acetic acid solution (ethanol/acetic acid; 94:4) for 1 h and observed under an epifluorescence microscope (Nikon Ni90 microscope, Japan).

Ubiquitination level detection
Total proteins of different strains were extracted as previously reported , which were subsequently assayed by Western blot analysis with an antiubiquitin as the primary antibody (1:5000, Sigma, USA) and an anti-mouse horseradish peroxidase as the secondary antibody (1:10000, Sigma, USA). Results were visualized with the ECL detection system (Amersham Biosciences, Piscataway, NJ).

Statistical analyses
All the data analysis was performed using SPSS (Statistical Program for Social Sciences) Version 10.0. species; SDS: Sodium dodecyl sulfate; UBP: Ubiquitin-specific protease; UCH: Ubiquitin C-terminal hydrolase