Balancing of the mitotic exit network and cell wall integrity signaling governs the development and pathogenicity in Magnaporthe oryzae

The fungal cell wall plays an essential role in maintaining cell morphology, transmitting external signals, controlling cell growth, and even virulence. Relaxation and irreversible stretching of the cell wall are the prerequisites of cell division and development, but they also inevitably cause cell wall stress. Both Mitotic Exit Network (MEN) and Cell Wall Integrity (CWI) are signaling pathways that govern cell division and cell stress response, respectively, how these pathways cross talk to govern and coordinate cellular growth, development, and pathogenicity remains not fully understood. We have identified MoSep1, MoDbf2, and MoMob1 as the conserved components of MEN from the rice blast fungus Magnaporthe oryzae. We have found that blocking cell division results in abnormal CWI signaling. In addition, we discovered that MoSep1 targets MoMkk1, a conserved key MAP kinase of the CWI pathway, through protein phosphorylation that promotes CWI signaling. Moreover, we provided evidence demonstrating that MoSep1-dependent MoMkk1 phosphorylation is essential for balancing cell division with CWI that maintains the dynamic stability required for virulence of the blast fungus.


Introduction
The cell wall plays a critical role in maintaining cell morphology and progression throughout the cell cycle, and its composition depends on species. It varies in the cell types and developmental stages [1]. In plants, cellulose and pectin are abundant in the primary cell wall that influence the cell wall porosity and extensibility during the cell cycle [2], while lignin found in the secondary cell wall protects the cell from the external stress [3]. In filamentous fungi, α-(1,3) glucan and chitin of the cell wall form a hydrophobic scaffold surrounded by a hydrated matrix of β-glucans and capped by a layer of glycoproteins and α-1,3-glucan [4][5]. The cell wall protects against the osmotic shock or mechanical stresses [6][7], and may also induce the host's immune response [8]. Thus, the cell wall integrity is critical to its function, including pathogenesis [9].
The Cell Wall Integrity (CWI) pathway is one of the most pivotal signal transmission mechanisms in response to cell wall stresses in the budding yeast Saccharomyces cerevisiae and pathogenic fungi, such as the rice blast fungus Magnaporthe oryzae [1,10]. The yeast CWI pathway mainly consists of a conserved Mitogen-Activated Protein (MAP) kinase kinase kinase (Bck1), a redundant MAP kinase kinase (Mkk1 and Mkk2), and a MAP kinase (Mpk1/Slt2) [11][12][13]. The disruption of the MAP kinase pathway components results in various cell lysis [1,14]. In M. oryzae, the MAP kinase pathway is composed of MoMck1 (the Bck1 homolog), MoMkk1 (the Mkk1/2 homolog), and MoMps1 (the Mpk1/Slt2 homolog) that play an essential role in aerial hyphae morphogenesis, appressorium formation, and pathogenicity [10,[14][15]. In addition, recent studies have found that other kinases, such as autophagy-related MoAtg1, could also phosphorylate MoMkk1 in response to the endoplasmic reticulum (ER) stress [16]. This study suggested the importance and the multitude of regulatory mechanisms by CWI signaling.
M. oryzae produces a specialized infection structure called appressorium to initiate the infection [17][18]. Appressorium formation requires not only the recognition of physical signals such as hardness and hydrophobicity but also proper cell cycle [19][20][21]. The normal conidium contains three single-nucleated cells, and the cells then germinate to produce the germ tubes, each containing a newly produced nucleus. This new nucleus will enter the appressorium while the nuclei remaining in the conidium undergo degradation. The nucleus in the appressorium will migrate to the invading peg (nail) and replicate multiple rounds during the infectious hypha (IH) expansion [22][23][24][25][26][27]. Therefore, the infection-related process is tightly linked to the cell division in M. oryzae.
In yeast, the mitotic exit network (MEN) is required for normal mitosis during the budding process. The small GTPase Tem1 targets the spindle positioning checkpoint (SPOC) to regulate the MEN activity whereas Tem1 itself is regulated by the GTPase-activating protein (GAP) Bub2-Bfa1 [28]. During the late anaphase, Tem1 activates the protein kinase Cdc15, activating the downstream kinase Dbf2 associated with its activating subunit, Mob1 [29]. Previous studies in M. oryzae showed that the deletion of MoBub2 (the Bub2 homologs) affects the switch from apical to polar growth due to the shortened G1 phase [22]. Also, the expression of a temperature-sensitive allele of Cdc15 homolog resulted in increased septation and nuclear division in the germ tubes and the defects in appressorium formation and plant infection [23]. These studies suggested conserved mechanisms may present in MEN and its regulation between the yeast and the blast fungus.
During incessant cell division accompanying vegetative growth, spore germination, and infection-related development, cells may rely on the CWI pathway to relieve the stress associated with the cell wall's continuous stretching and contracting [10,30]. There have been recent attempts to connect cell division with CWI [9,[31][32]. We here report identifying a novel link between the cell cycle control and CWI signaling that underlies the growth and pathogenicity of the rice blast fungus.

The balance between mitosis and CWI is vital for the vegetative and infectious growth of M. oryzae
To examine the link between cell division and CWI signaling, we co-introduced a histone H1-red fluorescent protein (RFP) and a MoMkk1-green fluorescent protein (GFP) gene fusion constructs into the wild-type strain Guy11. When stained with Calcofluor White (CFW), we found that hyphae grown in liquid culture exhibit a uniformly linear structure with one nucleus per cell. The septa are evenly distributed, in contrast to the infectious hyphae that are swollen ( Fig 1A). The hyphae grown in liquid culture contained about 4-6 nuclei per 100 μm hyphae, in contrast to [8][9][10][11] nuclei in the infectious hyphae ( Fig 1B). This indicated that the rate of cell division under vegetative and infectious growth was different. To investigate the role of CWI in the varied distribution of nuclei, we estimated MoMkk1 phosphorylation levels by Mn 2+ -Phos-tag SDS-PAGE. We found that the density of the phosphorylated-MoMkk1-GFP (P-MoMkk1) band is higher than that of MoMkk1-GFP (MoMkk1) in the presence of a phosphatase inhibitor, but not when the phosphatase was added. The P-MoMkk1 band from liquid-cultured hyphae was significantly weaker than that from the infectious hyphae ( Fig 1C).
To further explore the connection between CWI signaling and mitosis, we sprayed the conidial suspension of Guy11 treated by Nocodazole that inhibits mitosis onto 14 days old rice seedlings. The pathogenicity decreased significantly compared with the untreated conidial suspension (Fig 1D and 1E). We then added Nocodazole to the liquid culture and found that the normally even distribution of chitin is altered (Fig 1F). Nocodazole also caused abnormal hyphal branching in comparison to untreated hyphae with angled but forward branching ( Fig  1F). Consistent with the fluorescence observation results, the P-MoMkk1 level with Nocodazole treatment was much weaker than that without (Fig 1G). We introduced H1-RFP construct into ΔMomck1, ΔMomkk1, and ΔMomps1 mutants, respectively. Following CFW staining, we found that the hyphae contain multiple nuclei in each cell which shows that cell division is disordered when the cell wall integrity pathway is blocked (Fig 1H). These results indicated that there is a close relationship between CWI and cell division and that the dynamic balance between them is important for the normal vegetative growth of the fungus.

MoMkk1 interacts with the MEN kinase MoSep1
Since hyphal mitosis appears to be linked to MoMkk1 and its phosphorylation, we searched for MoMkk1 interacting proteins using both the yeast two-hybrid (Y2H) and co-immunoprecipitation (co-IP) approaches. We have identified a wide variety of potential interacting proteins, including cell cycle-associated proteins, mitochondrion proteins, carboxypeptidase, hyphae of (A). One hundred hyphae were counted for each type and the experiment was repeated three times. Error bars represent the standard deviations. Asterisks indicate statistical significance according to a Student's test (p<0.01). (C) Phosphorylation analysis of MoMkk1 in LH and IH. MoMkk1-GFP proteins were extracted in the presence of PMSF (PF), phosphatase (PE), as well as phosphatase inhibitors (PI), respectively, and then detected by the anti-GFP antibody. MoMkk1 phosphorylation was estimated by calculating the amount of phosphorylated-MoMkk1 (P-MoMkk1) compared to the total amount of MoMkk1 (the numbers underneath the blot). (D) Pathogenicity test in rice. The spore helicase, energy metabolism-related proteins, protein kinases, and proteins with unknown functions (S1 Table). Interestingly, a homolog of the yeast Ste/Ste11/Cdc15 protein kinase known to be associated with MEN functions [23] was identified and named as MoSep1 (MGG_04100) (S1A Fig). We again verified the interaction by Y2H and GST pull-down assays (Fig 2B and 2C). To demonstrate the interaction in vivo, we co-expressed MoMkk1-RFP and MoSep1-GFP fusion constructs into Guy11. We failed to detect the whole protein of MoSep1 in vivo by different methods. However, segmented MoSep1 proteins (STK and BACK) were able to purify and interact with MoMkk1 in vivo (Fig 2A-2D). MoSep1 STK is localized diffusely throughout the cell, MoSep1 BACK is localized to the spindle pole body (SPB) (S2 Fig). These localization patterns are similar to yeast Cdc15 [29].

MoSep1, MoDbf2, and MoMob1 are involved in vegetative growth, conidiation, and virulence of M. oryzae
In yeast, the MEN kinase pathway consists of Cdc15 and Dbf2-Mob1 [33]. Since MoSep1 is the homolog of Cdc15, we extended the search and identified MoDbf2 (MGG_02757) and MoMob1 (MGG_03151) as the Dbf2 and Mob1 homologs of M. oryzae (S1B and S1C Fig). The mutant strains ΔModbf2, ΔMomob1, and ΔMosep1 were respectively generated (S3A- S3D  Fig). All mutants showed apparent attenuated growth after incubation at 28˚C for 7 days in the dark (Fig 3-3C) and the defect in conidiation producing abnormal and approximately 33% conidia in comparison to Guy11 (Fig 3C). These results indicated that MEN plays a vital role in the growth and conidiation of the fungus.
In addition, the ΔModbf2, ΔMomob1, and ΔMosep1 mutant strains were sprayed onto 14 days old rice seedlings (cv. CO39), along with Guy11 and the respective complemented strains (ΔMosep1/MoSEP1, ΔModbf2/MoDBF2, ΔMomob1/MoMOB1). The mutants caused fewer and smaller lesions by producing more type 1-3 (70%) and few type 4-5 (5%) lesions in comparison to~25% type 4-5 lesions in Guy11 (Fig 4A and 4B). Similar results were observed in assays using the rice sheath injection assay (Fig 4C and 4D). Moreover, an excised leaf sheath assay was performed to detect the mutants' infectious hyphae extension ability [34]. At 48 hours post-inoculation (hpi), all mutants' invasive hyphae were restricted to a single cell, in contrast to Guy 11 whose invasive hyphae were free extending (Fig 4E and 4F). These results suggested that MoSep1, MoDbf2, and MoMob1 are all important for the pathogenicity of M. oryzae.
We then tested the appressorial turgor pressure using the incipient cytorrhysis [35] to examine the cause of virulence attenuation. At 1 M glycerol, the appressorium collapse rate was~37% in the ΔMosep1, ΔModbf2, and ΔMomob1 mutants compared with 11% in Guy11. With the increase of the external glycerol concentration, the proportion of collapsed appressoria remained higher in the mutants than in Guy11, indicating that mutant appressoria were defective in maintaining turgor pressure (S4A Fig). Since the turgor pressure is largely dependent on the breakdown of glycogen and the transfer of triacylglycerols [36], we estimated the cellular distribution of glycogen and lipid bodied using I 2 and KI staining [36]. We found that suspension of Guy11 was divided into two parts, one was treated with Nocodazole, and another without. Two-week-old rice seedlings were inoculated with these conidial suspensions treated with or without Nocodazole, and photographed at 7 days post-inoculation (dpi). (E) Diseased leaf area analysis of (D). Data were presented as a bar chart showing the percentage of lesion areas. Error bars represented the standard deviations from three independent experiments. Asterisks indicate statistical significance according to a Student's test (p<0.01). (F) Wild-type hyphae treated with or without Nocodazole were stained with CFW and observed by epifluorescence microscopy. Bar, 10 μm. (G) Phosphorylation analysis of MoMkk1 in Guy11 treated with or without Nocodazole (the steps are similar to C). (H) Hyphae of Guy11 and the ΔMomck1, ΔMomkk1, ΔMomps1 mutants expressing the H1-RFP construct were stained with CFW and examined by epifluorescence microscopy. Bar, 10 μm. https://doi.org/10.1371/journal.ppat.1009080.g001

PLOS PATHOGENS
MoSep1-dependent MoMkk1 phosphorylation mediates the balance between MEN and CWI in M.oryzae

PLOS PATHOGENS
MoSep1-dependent MoMkk1 phosphorylation mediates the balance between MEN and CWI in M.oryzae glycogen breakdown was delayed by 10 h in the three mutants, compared to 6 h in Guy11 (S4B Fig). We also performed a Nile red staining assay to examine lipid body distribution [37] and found that lipid degradation was significantly postponed in the mutants than in Guy11. Degradation slowed down with numerous lipid bodies remaining in more than 55% of mature appressoria 24 h following germination in the mutants. In comparison, lipid body transfer into appressoria was completed 8 h after germination in >90% of appressoria by Guy11 (S4C Fig).

MoSep1, MoDbf2, and MoMob1 are involved in MEN
To test the conserved roles of these three proteins in MEN of M. oryzae according to Saunders et al. [23], we introduced the histone H1-RFP gene fusion construct into these three mutants, analyzed by immunoblot (IB) with anti-His and anti-GST antibodies. GST, glutathione transferase. His, histidine. (D) Co-IP assay for the interaction between MoMkk1 and MoSep1. Plasmids of MoMkk1-RFP and MoSep1 STK -GFP, MoMkk1-RFP, and MoSep1 BACK -GFP were co-expressed in wild-type Guy11, respectively, and proteins were detected using anti-RFP and anti-GFP antibodies. Lysed hyphal proteins were allowed to bind to RFP beads at 4˚C for 4 h and then analyzed by IB with appropriate antibodies.

PLOS PATHOGENS
MoSep1-dependent MoMkk1 phosphorylation mediates the balance between MEN and CWI in M.oryzae respectively. Following CFW staining, we found that the hyphae have fewer and unevenly distributed septa (Fig 5A and 5C). They often contain multiple nuclei in each hyphal compartment and the number of nuclei per 100 μm hyphae was about 2-3 in the mutants compared to more of the leaf area). Error bars represent the standard deviations from three independent experiments. Asterisks indicate statistical significance according to a Student's test (p<0.01). (C) Pathogenicity was tested by rice injection assay with the conidia concentration of 5 × 10 4 spores/ml and photographed at 7 dpi. The arrow points to the injection site. (D) Diseased leaf area analysis of (C). The data were presented as a bar chart showing the percentage of lesion areas. Error bars represent the standard deviations from three independent experiments. Asterisks indicate statistical significance according to a Student's test (p<0.01). (E, F) Statistical analysis of the infectious hyphal type (type 1, no penetration; type 2, with penetration peg; type 3, with a single invasive hypha; type 4, with extensive hyphal growth) on rice leaf sheaths. Rice leaf sheaths were inoculated with conidial suspensions and examined at 36 h post-inoculation (hpi). One hundred infectious hyphae were counted for each strain and the experiment was repeated three times. Error bars represent the standard deviations. https://doi.org/10.1371/journal.ppat.1009080.g004

PLOS PATHOGENS
MoSep1-dependent MoMkk1 phosphorylation mediates the balance between MEN and CWI in M.oryzae 4-6 nuclei in Guy11 (Fig 5A, 5D and 5E). These results indicated that these three mutants are defective in mitotic exit and the nuclei do not stop to divide.
Our previous results showed that all three mutants produce fewer and abnormal conidia. More than 76% conidia of these three mutants were two-celled instead of typical three-celled, and more than 75% had only two nuclei ( Fig 3C and S5 Fig). Only one septum was observed in nearly all of the mutant conidia, in contrast to Guy11 whose conidia contain three nuclei with one nucleus per compartment. These findings suggested that MEN play significant functions in mitosis and cytokinesis of the conidia. To further explore this function, we assayed the appressorium formation ability on hydrophobic surfaces [38]. In Guy11, one nucleus moves back into the conidium following a single round of mitosis, and the other nucleus moves into the developing appressorium 6 h following germination. However, in the mutants, appressoria remain unformed, and the nuclei remained in the germinal tubes (S5 Fig). Infectious hyphae growth in rice sheath assays showed that the number of nuclei per 100 μm infectious hyphae was about 3-4 in the mutants compared to 8-11 nuclei in Guy11 (Fig 5B and 5F). All these results suggested that MoSep1, MoDbf2, and MoMob1 are required for mitotic exit and proper cytokinesis.

Disruption of MEN results in defects in CWI
Since mitosis requires normal CWI activities including MoMkk1 phosphorylation [10], we wondered whether MoSep1, MoDbf2, or MoMob1 are involved in CWI signaling. We compared the impact of cell wall-degrading enzymes on mycelial cell walls and found that the hyphae of the ΔMosep1, ΔModbf2, and ΔMomob1 mutants released more protoplasts than Guy11, suggesting more sensitive (Fig 6A and 6B). The mutant strains were also more susceptible to cell wall stress induced by CFW and Congo red (CR) (Fig 6C).
We used CFW staining [10,39] to test the cell wall properties of the three mutants and found that chitin distribution is uneven and not concentrated on the growing apices (Fig 6D  and 6E). To detect whether MoSep1, MoDbf2, and MoMob1 are involved in chitin biosynthesis the regulation of chitin biosynthesis, we tested the expression levels of seven chitin synthase (CHS) genes and found six were significantly reduced (S6A Fig). We also estimated the chitin contents of ΔMosep1, ΔModbf2, and ΔMomob1 mutants to be approximately 30% lower than that of Guy11 (S6B Fig). These results collectively suggested that MoSep1, MoDbf2, and MoMob1 are required for CWI functions and regulation.

MoSep1 phosphorylates MoMkk1 and MoSep1-dependent MoMkk1 phosphorylation sites are vital for CWI
Given that the MEN pathway is linked to the CWI pathway and MoSep1 interacts with MoMkk1, we were curious about whether the MoSep1-MoMkk1 interaction mediates the link. We then tested but failed to establish any interactions of MoDbf2 and MoMob1 with MoMck1, MoMkk1, or MoMps1 through Y2H (Fig 7A). To further test whether MoSep1 phosphorylates MoMkk1 directly, Mn 2+ -Phos-tag SDS-PAGE was again performed. Results showed that the P-MoMkk1 band is weaker in the ΔMosep1 strain than in Guy11 but higher than the ΔMomck1 mutant as MoMck1 normally phosphorylates MoMkk1 (Fig 7B). The further in vitro phosphorylation assay using protein gel-staining fluorescence dye supported that MoSep1 (His-MoSep1) could phosphorylate MoMkk1 (GST-MoMkk1) (Fig 7C).
We further determined the specific phosphorylation sites on MoMkk1 by MoSep1. MoMkk1-GFP proteins purified from the ΔMosep1 mutant and Guy11 were subject to mass spectrometry (MS) analysis that identified S19, T24, S125, S136, T139, T207 as the candidate phosphorylated sites (Fig 7D and S7 Fig). To validate this result, the phosphomimic mutation
To further verify if this MoSep1-dependent phosphorylation is involved in regulating the CWI pathway, we again estimated protoplast release from these strains. ΔMosep1 hyphae released more protoplasts than ΔMosep1/MoMKK1 6D , and ΔMomkk1/ MoMKK1 6A hyphae released more protoplasts than Guy11 (S9A and S9B Fig). Consistently, ΔMomkk1/
We further tested the phenotypes of constitutively unphosphorylated strains. We found that the transformants of MoMkk1 S19A -GFP, MoMkk1 S125A -GFP, and MoMkk1 T207A -GFP undergo progressive hyphal autolysis compared with the abundant aerial hyphae of the wild type, they are similar to the ΔMomkk1 mutant ( Fig 9A). The conidia of MoMkk1 S136A -GFP and MoMkk1 T207A -GFP were severely reduced, similar to the ΔMomkk1 mutant ( Fig 9B). Further pathogenicity tests using mycelial fragments to inoculate detached wounded rice leaves showed that the virulence of these mutants varies (Fig 9C and 9D), with inactivation of these six phosphorylation sites concurrently showed the most serious defects. The phosphomimic mutation of MoMkk1 6D can completely restore the defect in growth, conidiation, and virulence of ΔMomkk1 mutant as a control (S10 Fig). These results indicated that the MoSep1-dependent MoMkk1 phosphorylation is essential for the development and pathogenicity of M. oryzae.

MoSep1-dependent MoMkk1 phosphorylation is important for MEN in M. oryzae
As the constitutive activation of MoMkk1 phosphorylated sites suppresses the defects of the ΔMosep1 mutant, we were interested in testing whether such phosphorylation sites are also involved in the regulation of the MEN pathway. The H1-RFP construct was transformed into  (A) Growth of ΔMosep1/MoSEP1, ΔMosep1, ΔMosep1/MoMKK1 6D (ΔMosep1/MoMKK1 S19D, T24D, S125D, S136D, T139D, T207D ), ΔMosep1/ MoMKK1 S19D , ΔMosep1/MoMKK1 T24D , ΔMosep1/MoMKK1 S125D , ΔMosep1/MoMKK1 S136D , ΔMosep1/MoMKK1 T139D , ΔMosep1/ MoMKK1 T207D mutant strains on CM medium. (B) Statistical analysis of conidia production on SDC medium cultured at 28˚C for 7 days in the dark followed by 3 days of continuous illumination under fluorescent light. Error bars represent the standard deviations from three independent experiments. Asterisks indicate statistical significance according to a Student's test (p<0.01). (C) Pathogenicity analysis using rice spraying assays and photographed at 7 dpi. (D) Diseased leaf area analysis of (C). Data were presented as a bar chart showing the percentage of lesion areas. Error bars represented the standard deviations from three independent experiments. Asterisks denote statistical significance according to a Student's test (p<0.01). (E) Statistical analysis of the infectious hyphal type (type 1, no penetration; type 2, with penetration peg; type 3, with a single invasive hypha; type 4, with extensive hyphal growth) on rice leaf sheaths. Rice leaf sheaths were inoculated with conidial suspensions and examined at 36 h post-inoculation (hpi). One hundred infectious hyphae were counted for each strain and the experiment was repeated three times. Error bars represented the standard deviations. https://doi.org/10.1371/journal.ppat.1009080.g008

PLOS PATHOGENS
MoSep1-dependent MoMkk1 phosphorylation mediates the balance between MEN and CWI in M.oryzae the ΔMosep1/MoMKK1 6D and ΔMomkk1/MoMKK1 6A strains. When stained with CFW, additional septa were observed in the ΔMosep1/MoMKK1 6D strain than in ΔMosep1 mutant and fewer septa in the ΔMomkk1/MoMKK1 6A strain than Guy11 (Fig 10A-10C), suggesting that the activation of the MoSep1-dependent phosphorylation sites is important for the formation

PLOS PATHOGENS
MoSep1-dependent MoMkk1 phosphorylation mediates the balance between MEN and CWI in M.oryzae of septa. In addition, more than one nucleus per hyphae cell was observed in the ΔMomkk1/ MoMKK1 6A strain in comparison to one nucleus per cell of the normal hyphae (Fig 10A-10C) shows that the inactivated mutations MoMkk1 6A cannot restore the detect of ΔMomkk1. The abnormal conidia of the ΔMosep1/MoMKK1 6D strain were fewer than those in the ΔMosep1 mutant (Fig 10D, 10F and 10G). The infectious hyphae extension ability was more robust in the ΔMosep1/MoMKK1 6D strain than in the ΔMosep1 mutant, and 6-7 nuclei per 100 μm of infectious hyphae were more similar to the wild type strain (Fig 10E and 10H). These results further indicated the importance of MoSep1-dependent MoMkk1 activation in the cell cycle and mitosis of M. oryzae.

Discussion
The cell wall is essential for maintaining cell morphology and functioning as a barrier against extracellular stress [10,40]. Balancing the rigidity or integrity while allowing growth-related expansion or plasticity remains the most important and sought after research projects [41][42][43][44]. This study found that the cell cycle-related kinase MoSep1 phosphorylates the cell wall integrity (CWI) pathway component MoMkk1 to link the mitotic exit network (MEN) with CWI signaling. This connection provides a means to balance the rigidity with the cell wall's remodeling to allow cell division and growth. Significantly, MoSep1-MoMkk1 mediates the MEN and CWI crosstalk necessary for the virulence of M. oryzae (Fig 11). Furthermore, we found that the deletion of MoDBF2 and MoMOB1 that are downstream of MoSep1 also resulted in CWI defects. We speculated that there might be other connections between the cell wall integrity and cell division. There is still much to explore between these two essential pathways.
The cell cycle is a fundamental biological process that underlies the growth and development [45] and the entry, maintenance, and exit are all tightly regulated processes [46][47].Cdc15, a MoSep1 homolog, a vital kinase of the mitosis exit network that integrate Tem1 G-protein cycle and Cdc5 synthesis/degradation cycle to ensure mitosis exit after normal genome division [29]. Meanwhile, CWI is the key pathway regulating the adaptive response to cell wall biogenesis, morphogenesis, and stress. During cell elongation, the CWI detection system will adjust to the cell wall's stress, providing the possibility that the maintenance of cell wall integrity is associated with the cell cycle [30,48]. Many previous studies have found contacts between cell wall synthesis and the cell cycle. For example, the plant cell wall thickness could influence polarity growth during cell division [31]. Newer technologies, such as a sub-resolution method to monitor cell wall thickness dynamics in living cells, were also used to detect cell wall dynamics during growth and division, revealing the connection between mitosis and cell wall remodeling [30]. Yeast cell wall damage promotes a checkpoint-like mechanism inhibiting DNA replication via Mck1-dependent Cdc6 degradation [49]. Here, we found that the deletion of MEN components also severely hindered cell wall function and decreased CWI activities (marked by MoMkk1 phosphorylation). This is the first time for a relationship between CWI and mitosis to be revealed in fungi. We further demonstrated that the link is established in a phosphorylation manner through the specific interaction between the MEN kinase MoSep1 and the CWI component MoMkk1. Specific MoSep1-dependent MoMkk1 phosphorylation sites were also identified and these roles in the development and virulence of M. oryzae revealed, proving that the crosstalk between MEN and CWI is important for the rice blast fungus.
During the growth and infection of M. oryzae, it is challenging how to relieve the stress originated from cell elongation. Studies have shown that some MEN components required for the G1/S phase progression are essential for maintaining proper appressorium development and pathogenicity of M. oryzae [22][23]. At the same time, the CWI pathway is also indispensable for the morphogenesis of aerial hyphae and fungal infection since the deletion of any CWI

PLOS PATHOGENS
MoSep1-dependent MoMkk1 phosphorylation mediates the balance between MEN and CWI in M.oryzae MAPK cascade components resulted in autolysis and attenuated virulence [10]. We found that it is difficult to detect the phosphorylation of MoMkk1 in the absence of MoMck1 (Fig 7B). Considering the importance of this conserved MAPK cascade, we wondered whether other kinases necessitate MoMck1 in order to phosphorylate MoMkk1 or that MoMkk1 phosphorylation is hard to detect in vivo. We have purified MoSep1 and MoMkk1 proteins in vitro, and found that MoSep1 can phosphorylate MoMkk1 without the presence of MoMck1. A previous study revealed that MoAtg1-mediated MoMkk1 phosphorylation could not be detected in the ΔMomck1 mutant in vivo, but the detection improves when cells are under dithiothreitol (DTT)-induced stress. We reasoned that MoMkk1 phosphorylation by various kinases, including MoMck1, MoAtg1, or MoSep1, is an independent and specific event. Based on that these independent biological processes are respectively important, we further revealed that the link between the MEN and CWI pathways is equally vital. The MoSep1-MoMkk1 interaction ensures the proper response to the stress caused by cell elongation that underlies the fungal growth, development, and virulence.
We previously found that MoMkk1 also mediates crosstalk between the CWI pathway and other pathways, including the unfolded protein response (UPR) pathway, the high osmolality

PLOS PATHOGENS
MoSep1-dependent MoMkk1 phosphorylation mediates the balance between MEN and CWI in M.oryzae glycerol (HOG) MAPK pathway, and autophagy [10]. We demonstrated that the autophagyrelated kinase MoAtg1 phosphorylates MoMkk1 under ER stress [16]. By identifying the cell cycle-related kinase MoSep1 that phosphorylates MoMkk1 to enhance the CWI activity, we revealed the complex and multitude of signaling pathway crosstalk surrounding CWI. The concurrent inactive phosphorylation sites mutation of MoSep1-and MoMck1-dependent mutant strains abolishing MoMps1 phosphorylation allowed us to conclude that MoSep1 and MoMck1 significantly contribute MoMkk1 phosphorylation under normal conditions. Such a conclusion implicates that the function of CWI in response to cell cycle-related stress plays an important part in the growth and pathogenicity of the blast fungus. Given such an important role, further investigation of MEN-CWI crosstalk is highly warranted.

Strains and culture conditions
The M. oryzae Guy11 strain was used as the wild type (WT) in this study. All strains were cultured on complete medium (CM) agar plates in the dark at 28˚C, unless indicated otherwise [10]. Fungal mycelia were harvested from liquid CM for 1 day or 2 days at 28˚C for microscopic observation, DNA, RNA, and protein extraction.

Epifluorescence microscopy and nucleus staining and CFW staining
M. oryzae cells, including hyphae, conidia, and appressorium expressing fluorescent fusion proteins were incubated under appropriate conditions. The constructs, including H1-RFP, MoSep1 STK -GFP, MoSep1 BACK -GFP, and other phosphorylation mutations were transformed into ΔMomck1, ΔMomkk1, ΔMomps1, ΔMosep1, ΔModbf2, and ΔMomob1 mutants or the wild type Guy11 strain. Epifluorescence microscopy was performed using a Zeiss LSM710 microscope. To visualize nuclei, DAPI was used at a concentration of 1 mg/ml in room temperature in darkness for 5 min. For CFW staining, mycelia were stained with 10 mg/ml for 5 min in the dark.

Co-immunoprecipitation (Co-IP) assay
The MoMkk1-GFP fusion construct was transferred into Guy11. Total proteins were extracted and then incubated with anti-GFP beads. Proteins bounded to the beads were eluted and analyzed by MS. To confirm in vivo interactions involving MoMkk1-MoSep1 STK , and MoMkk1--MoSep1 BACK , the MoMkk1-RFP, MoSep1 STK -GFP, and MoSep1 BACK -GFP fusion constructs were constructed. Total proteins were isolated from transformants and incubated with anti-RFP beads. Proteins bounds to the beads were eluted and analyzed by Western blot with anti-GFP and anti-RFP antibodies.

Yeast-two-hybrid (Y2H) assay
Full-length cDNA MoMCK1, MoMKK1, and MoMPS1 was cloned into pGADT7 and cDNA of MoSEP1 (MoSEP1 STK and MoSEP1 BACK ), MoDBF2, and MoMOB1 genes was cloned into pGBKT7, respectively. The resulting constructs were first confirmed by DNA sequencing and then transformed in pairs into the yeast strain AH109 as previously described [16].

PLOS PATHOGENS
MoSep1-dependent MoMkk1 phosphorylation mediates the balance between MEN and CWI in M.oryzae supernatants were then mixed with 100 μl glutathione Sepharose beads incubated at 4˚C for 2 h. The recombinant GST-MoMkk1 bound to glutathione Sepharose beads was incubated with E. coli cell lysate containing His, His-MoSep1. His-MoSep1 STK , His-MoSep1 BACK at 4˚C for another 4 h. Finally, the beads were washed with buffer (50 mM Tris, pH 8.0, 50 mM NaCl, 1 mM PMSF, 1% Triton X-100) 5 times and eluted from the beads. Eluted proteins were analyzed by immunoblot (IB) with anti-His and anti-GST antibodies [50].

Phosphorylation analysis
The MoMkk1-GFP fusion construct was transferred into Guy11 and ΔMosep1 mutant. The total protein extracted from mycelium was resolved on 8% SDS-PAGE prepared with 50 μm acrylamide-dependent Phos-tag ligand and 100 μm MnCl 2 as described [16]. Gel electrophoresis was performed with a constant voltage of 80 V for 6-8 h. Gels were soaked in transfer buffer with 5-10 mM EDTA for 15 min three times before transferring, and followed by transfer buffer without EDTA for another 10 min two times. Protein transfer from the Mn 2+ -phos-tag acrylamide gel to the PVDF membrane was performed for~48 h (depend on different proteins) at 80 V at 4˚C, and then the membrane was analyzed by Western blotting using the anti-GFP antibody.

Gene deletion and complementation
The gene deletion mutants were generated through the standard one-step gene replacement strategy [51]. To generate the MoSEP1 gene replacement construct, 0.9-kb upstream and 1.1-kb downstream flanking sequences were amplified from M. oryzae genomic DNA. The resulting PCR products were digested with restriction endonucleases and ligated with the hygromycin resistance cassette (HPH) released from pCX62. The flanking sequences and the HPH cassette was amplified and transformed into Guy11 protoplasts. Putative mutants were verified by PCR. The complement fragment, which contains the entire MoSEP1 gene coding region and its native promoter region, was amplified by PCR with primers and inserted into pYF11 to complement the mutant strain. The MoDBF2 and MoMOB1 gene deletion mutants were obtained using a similar strategy.

Mass spectrometric analysis
To identify MoMkk1 interacting proteins. Total proteins were extracted from Guy11/ MoMKK1-GFP transformants. Approximately 60 μl of anti-GFP beads was added into 2 ml diluted total protein samples to capture MoMkk1 interacting proteins. After incubated at 4˚C overnight, the beads were washed 3 times with 700 μl PBS and proteins were eluted with 90 μl elution buffer (0.2 M glycine, pH 2.5). The eluent was immediately neutralized with 10 μl neutralization buffer (1.5 M Tris, pH 9.0). The eluted protein samples were subject to mass spectrometry analysis (Beijing Protein Innovation).
To identify phosphorylation sites of targeted proteins, total proteins were extracted from Guy11/MoMKK1-GFP and ΔMosep1/MoMKK1-GFP transformants. Approximately 30 μl of anti-GFP beads was added into 1 ml diluted total protein samples. Following similar purification steps as above, the eluted proteins were neutralized and separated on 10% SDS-PAGE gel. The gel bands corresponding to the targeted proteins were excised from the gel and subject to mass spectrometry analysis as above.

Plant infection assays
Conidia were adjusted to 5 × 10 4 spores/ml in a 0.2% (w/v) gelation solution and 5 ml was sprayed on 2-week-old rice seedlings (Oryza stative cv. CO39). Inoculated plants were kept in

PLOS PATHOGENS
MoSep1-dependent MoMkk1 phosphorylation mediates the balance between MEN and CWI in M.oryzae a chamber at 25˚C under 90% humidity for the first 24 h and followed by a light/dark cycle for 4-7 days. For the injection assay, conidia (5 × 10 4 spores/ml) were injected into the 4-weekold rice sheath using a 1ml syringe. After 30~48 h incubation at 25˚C, the infected sheath inner epidermis of were torn for observation. For those mutants that could not form conidia, mycelia cultured in liquid CM for 2 days were harvested and fragmented prior to inoculation inoculated on wounded rice leaves and kept in the same conditions as described above.

Protoplast release assay
Mycelia cultured in liquid CM were filtered and collected with a filter cloth. They were then washed and resuspended in a liquid containing 10 mg/ml lysing enzyme. Protoplast release was observed in 0.5 h, 1 h, and 1.5 h with a light microscope. This experiment was repeated three times.

Chitin content
Mycelial samples were freeze-dried, and 5 mg of the freeze-dried mycelia were treated in 1 ml 6% KOH at 80˚C for 1.5 h. The treated samples were centrifuged (14,000 g, 10 min) and the precipitates were washed 3 times with PBS. Precipitates were then suspended in 0.5 ml of Mcllvaine's buffer (pH 6). An aliquot of 100 μl (13 units) of Streptomyces plicatus chitinase was added, incubated at 37˚C for 16 h, and then mixed with 100 μl of 0.27 M sodium borate (pH 9) before boiling for 10 min. 1 ml of 10 folds diluted Ehrlich's reagent was then added and further incubated at 37˚C for 20 min. A 1 ml sample was then transferred to a 2.5-ml plastic cuvette and absorbance at 585 nm was recorded.

In vitro phosphorylation analysis
The His-MoSep1, GST-MoMkk1, GST-MoMkk1 S19A , GST-MoMkk1 T24A , GST-MoMkk1 S125A , GST-MoMkk1 S136A , GST-MoMkk1 T139A , GST-MoMkk1 T207A , GST-MoMkk1 6A were expressed in Escherichia coli BL21-CodonPlus (DE3) cells and purified. A rapid Fluorescence Detection in Tube (FDIT) method using Pro-Q Diamond Phosphorylation Gel Stain was used to analyze protein phosphorylation in vitro. First, 2 mg MoMkk1 (or the unphosphorylated mutations) was mixed with MoSep1 in a kinase reaction buffer (100 mM PBS, 1 mM ascorbic acid, pH 7.5, 10 mM MgCl 2 ), with 50 mM ATP at 25˚C for 1 h. 10 folds of cold acetone was then added to terminate the reaction. Casein was homogenized and suspended in Mili-Q water at a concentration of 0.2 mg/ml to stain the proteins. Briefly, 100 ml of Pro-Q Diamond was mixed with 10 ml of casein and the mixture was kept in dark 1 h. 10 folds of cold acetone was then added and the mixture was allowed to incubate overnight at -20˚C. Proteins were precipitated by centrifugation at 14,000 rpm for 1 h at 4˚C. Discard the supernatant. The pellet (proteins) was washed twice using 500 μl cold acetone. The pellet was dissolved in 200 μl of Mili-Q water and transferred to a 96 well plate. Fluorescence signal at 590 nm (exited at 530 nm) was measured using a Cytation3 microplate reader.