G protein α subunit suppresses sporangium formation through a serine/threonine protein kinase in Phytophthora sojae

Eukaryotic heterotrimeric guanine nucleotide-binding proteins consist of α, β, and γ subunits, which act as molecular switches to regulate a number of fundamental cellular processes. In the oomycete pathogen Phytophthora sojae, the sole G protein α subunit (Gα; encoded by PsGPA1) has been found to be involved in zoospore mobility and virulence, but how it functions remains unclear. In this study, we show that the Gα subunit PsGPA1 directly interacts with PsYPK1, a serine/threonine protein kinase that consists of an N-terminal region with unknown function and a C-terminal region with a conserved catalytic kinase domain. We generated knockout and knockout-complemented strains of PsYPK1 and found that deletion of PsYPK1 resulted in a pronounced reduction in the production of sporangia and oospores, in mycelial growth on nutrient poor medium, and in virulence. PsYPK1 exhibits a cytoplasmic-nuclear localization pattern that is essential for sporangium formation and virulence of P. sojae. Interestingly, PsGPA1 overexpression was found to prevent nuclear localization of PsYPK1 by exclusively binding to the N-terminal region of PsYPK1, therefore accounting for its negative role in sporangium formation. Our data demonstrate that PsGPA1 negatively regulates sporangium formation by repressing the nuclear localization of its downstream kinase PsYPK1.

Introduction the Gα subunit PiGPA1 has roles in swimming behavior and chemotaxis of zoospores, whereas Gβ and Gγ (PiGPB1 and PiGPG1, respectively) are critical for sporangia development and zoosporogenesis [23,25,26]. Also in P. sojae the single Gα subunit protein PsGPA1 is important for zoospore functioning because silencing of the encoding gene leads to defects in zoospore chemotaxis and encystment [27]. Attempts to find proteins interacting with PsGPA1 resulted in the identification of a histidine triad nucleotide-binding protein 1 (PsHint1) that is also involved in zoospore chemotaxis, and required for cyst germination and full virulence [28]. Apart from PsHint1, however, the downstream effectors or pathways mediated by Phytophthora G proteins and the signals recognized by the single G protein complex are largely unknown.
In this study, we identified a serine-threonine kinase in P. sojae, called PsYPK1, which directly interacts with the Gα protein PsGPA1. We found that PsGPA1 acts as a negative regulator of sporangium formation by inhibiting the nuclear localization of PsYPK1. Furthermore, PsYPK1 has Gα-independent pathways that are involved in mycelium growth and oospore formation. Our data support a model in which the G protein α subunit suppresses sporangium formation in P. sojae through a serine/threonine protein kinase.

The PsGPA1 protein interacts with the N-terminal region of PsYPK1
To identify PsGPA1-interacting proteins, we obtained a P. sojae transformant expressing 3×FLAG-tagged PsGPA1 and performed protein co-immunoprecipitation (Co-IP) assays on lysates of this transformant using FLAG antibodies as bait (S1A Fig). Mass spectroscopy (MS) analysis of the complex co-precipitating with PsGPA1 identified 65 candidates as potential interactors of PsGPA1 including two serine/threonine kinases (i.e. Ps350990 and Ps350499) for which the interaction with PsGPA1 was confirmed (S1 Table and  amino acids in size and apart from a conserved serine/threonine catalytic domain (amino acids 542-861) in the C-terminal part it has no known domains (Fig 1A). It belongs to the AGC family of serine/threonine kinases, one of the most evolutionarily conserved groups of protein kinases in eukaryotes [29]. Based on BLAST searches using the PsYPK1 protein sequence as query, homologs were identified in other Phytophthora spp. Sequence alignment with homologs from organisms in the various kingdoms revealed that the proteins have a conserved C-terminal catalytic domain preceded by a highly variable N-terminal region that is relatively large in Phytophthora YPK1 proteins (S3 Fig).
To verify the protein-protein interaction between PsGPA1 and PsYPK1 we used an in vitro glutathione S-transferase (GST) pull-down assay (Fig 1B). GST-PsGPA1 and 6×His-PsYPK1 were expressed separately in Escherichia coli, and then purified for use in a pull-down experiment. As shown in Fig 1B, 6×His-PsYPK1 bound specifically to GST-PsGPA1, but not to GST, suggesting that the two proteins interact directly. To confirm that this interaction also occurs in vivo, PsGPA1-FLAG was co-expressed with PsYPK1-GFP in P. sojae and lysates of the mycelium were used in Co-IP assays. Also these assays showed a specific interaction between the two proteins (Fig 1C). To determine which region of PsYPK1 is responsible for the PsGPA1-PsYPK1 interaction, deletion constructs of PsYPK1 were generated and tested in the interaction assays. The results revealed that deletion of the N-terminal region of PsYPK1 (N-ter, amino acids 2-303) abolished its interaction with PsGPA1, both in the in vitro (Fig 1B; His-PsYPK1 C-ter ) and in vivo assay (Fig 1C; PsYPK1 ΔN -GFP). In contrast, a construct lacking amino acids 543-861 in the C-terminal region retained its ability to interact with PsGPA1 ( Fig  1B; His-PsYPK1 N-ter ). These data suggest that PsYPK1 directly associates with PsGPA1 and that its N-terminal region is required and sufficient for the interaction with PsGPA1.

PsYPK1 is required for growth of P. sojae in nutrient poor conditions and for oospore production, sporangium formation and virulence
To investigate the biological function of PsYPK1 we generated PsYPK1 knockout mutants using CRISPR-mediated gene replacement [6] as well as complemented knockout mutants, and analyzed their phenotypes. We show the results obtained from one representative PsYPK1 knockout mutant, ΔPsYPK1, and two representative PsYPK1 complemented strains, ΔPsYPK1-C1 and ΔPsYPK1-C2. The recipient wild-type strain (P6497) and two empty vector lines (EV in which the PsYPK1 knockout was not successful and ΔPsYPK1-EV that was not successfully complemented with PsYPK1) were included as controls (S4A-S4D Fig).
When cultured on nutrient-rich V8 medium no significant differences in growth characteristics were observed (Fig 2A and 2B), whereas on nutrient-poor Plich medium a clear reduction in growth was observed for ΔPsYPK1 and ΔPsYPK1-EV (Fig 2A and 2C). When substituting glucose, the major nutrient source in Plich medium, by other monosaccharides we observed that in the presence of fructose ΔPsYPK1 and ΔPsYPK1-EV also showed a decreased growth rate while in the presence xylose, galactose and mannose the growth rate of the mutants was comparable to that of the wild-type strain P6497 (S5 Fig). With respect to sporangium formation P6497, EV, ΔPsYPK1-C1, and ΔPsYPK1-C2 all formed abundant sporangia, while ΔPsYPK1 and ΔPsYPK1-EV exhibited a significant reduction (97%) in The presence of His-tagged proteins was detected by western blot analysis using a His antibody. (C) Validation of the association between PsYPK1 and PsGPA1 in vivo. Co-immunoprecipitations (Co-IP) were performed in extracts of Phytophthora sojae mycelium expressing PsGPA1-FLAG with PsYPK1-GFP or PsYPK1 ΔN -GFP. The presence of FLAG-tagged proteins was detected by western blot analysis using a FLAG antibody. The bands detected with anti-GFP were quantified with the ODYSSEY infrared imaging system (application software version 2.1).
When P. sojae invades and colonizes soybean the transcription of PsYPK1 is highly upregulated suggesting a role for PsYPK1 during infection stages (S6A Fig). Soybean seedlings inoculated with ΔPsYPK1 or ΔPsYPK1-EV developed only small necrotic lesions at the site of inoculation, whereas P6497, EV, ΔPsYPK1-C1, and ΔPsYPK1-C2 all produced normal disease lesions (Fig 3A). Measurements of relative P. sojae biomass in infected soybean seedlings revealed hardly any pathogen biomass in ΔPsYPK1-or ΔPsYPK1-EV-infected tissues with levels less than 1% of the levels found in P6497-, EV-, ΔPsYPK1-C1-, or ΔPsYPK1-C2-infected tissues (Fig 3B).
To explore why the pathogenicity of the ΔPsYPK1 strain was impeded, epidermal cells from the seedling inoculation sites were excised at different time points and observed using microscopy. At 12 h post-inoculation (hpi), P6497, EV, ΔPsYPK1-C1, and ΔPsYPK1-C2 were able to penetrate into the epidermal cells and the invasive hyphae were able to expand into adjacent epidermal cells, whereas the hyphae of ΔPsYPK1 and ΔPsYPK1-EV were restricted to the regions surrounding the penetration site (Fig 3C). At 24 hpi, the infectious hyphae of the P6497, EV, and the complemented strains exhibited tapering tips and were much more abundant, which had extended into neighboring cells (Fig 3C). Approximately 76% of the infectious hypha produced by ΔPsYPK1 or ΔPsYPK1-EV showed abnormal growth in branches and apical swelling, which may contribute to the reduction in the number of successful P. sojae infections (Fig 3C and 3D). Taken together these results indicate that PsYPK1 has a role in growth in nutrient poor medium, oospore production, sporangium formation, and virulence.

PsGPA1 is a negative regulator of sporangium formation in P. sojae
The high expression levels of the Gα subunit gene PsGPA1 during sporangium formation could point to a role for Gα in regulating sporangium formation (S6B Fig). To test this we generated PsGPA1-overexpressing lines using a constitutively active Ham34 promoter and analyzed the phenotype of one representative line, PsGPA1-OE, which showed a 2.5-fold increase in PsGPA1 transcript levels compared to the wild-type strain (P6497) and an empty vector control (PsGPA1-EV) (Fig 4A). In this analyses we included, a PsGPA1-silenced mutant (PsGPA1-M27) generated previously [27] in which the PsGPA1 transcript levels are strongly reduced (>80%) (Fig 4A). PsGPA1-OE exhibited a dramatic reduction (70%) in sporangia number, while the number of sporangia in PsGPA1-M27 was conversely increased (50%; Fig  4B and 4C). No obvious changes were found in oospore production and mycelium growth (Fig 4B and 4D-4F), indicating that PsYPK1 may have a PsGPA1-independent pathway that regulates vegetative growth and oospore production. Collectively, these results confirm that the Gα protein in P. sojae plays a negative role during sporangia development. Similar to silencing of PsGPA1, PsGPA1 overexpression also leads to reduced virulence (Fig 4B and 4G). This is a confirmation of findings of Hua et al. [27] who showed pleiotropic effects of PsGPA1silencing. Likely the PsGPA1-overexpression and silencing causes aberrations in the various pathways regulated by Gα, with several of these governing virulence functions. Moreover, PsGPA1 mutants only showed obvious reduction in virulence on unwounded plants upon inoculation with zoospores but not on wounded plants or upon inoculation with hyphae while

Nuclear localization of PsYPK1 is determined by the N-terminal region and facilitates sporangium formation
Given that the N-terminal region of PsYPK1 is essential for its interaction with PsGPA1 ( Fig  1B and 1C), and that all Phytophthora YPK1 homologs have an extended N-terminus (S3 Fig) we sought to investigate the role of the N-terminal region. To this end constructs were made in which GFP is fused to either the N-terminal region of PsYPK1 (PsYPK1 N-ter -GFP) or to a N-terminal truncated version of PsYPK1 in which the amino acids 2-303 were deleted (PsYPK1 ΔN -GFP) and these were transformed into P. sojae. Hyphae of the resulting transformants and of transformants expressing the full-length PsYPK1 fused to GFP (PsYPK1-GFP) were examined by fluorescence microscopy (Fig 5 and S7 Fig). In the PsYPK1-GFP and PsYPK1 N-ter -GFP transformants the GFP fluorescence was primarily localized in the cytoplasm, while in the PsYPK1 ΔN -GFP transformants there was hardly any fluorescence in the cytoplasm. In contrast, the nuclei, visualized by DAPI staining, showed a very strong fluorescence indicating that the truncated PsYPK1 ΔN -GFP protein accumulates in the nucleus (Fig 5  and S7 Fig). Because of the interaction between PsGPA1 and N-terminal region of PsYPK1, the question was raised where PsGPA1 is localized. Fluorescence microscopy of PsGPA1-GFP To assess whether the altered localization pattern of PsYPK1 was related to its function, we generated P. sojae tranformants in which the N-terminal region of PsYPK1 is deleted and used one of the lines, named PsYPK1 ΔN , in infection assays. This revealed that PsYPK1 ΔN has a decreased virulence on soybean (Fig 6A, 6B and 6D), albeit that this decrease is far less than that of the full-length knockout (Fig 3). Notably, sporangium production in PsYPK1 ΔN was dramatically increased (Fig 6A, 6C and 6D) as opposed to the significant decrease in sporangia production in the full-length knockout (Fig 2). These results indicate that the catalytic domain of PsYPK1 is sufficient for sporangium formation. For further confirmation of the essential role of the kinase domain, we generated a mutant in which the kinase domain of PsYPK1 is deleted (PsYPK1 ΔC ). Similar to the full length PsYPK1-knockout (ΔPsYPK1), PsYPK1 ΔC produced less sporangia and less oospore formation, and showed reduced virulence (S8A- S8C Fig).

PsGPA1 prevents nuclear localization of PsYPK1 and this leads to suppression of sporangia formation
Considering that the N-terminal region of PsYPK1 is required for interaction with PsGPA1, the nuclear localization of PsYPK1 ΔN might reflect the default steady-state distribution of PsYPK1. This led us to investigate to what extent PsGPA1 affects the localization of PsYPK1. To this end, we examined the localization of PsYPK1-GFP in PsGPA1 overexpression and silenced strains. In the PsGPA1 overexpression strain the fluorescence is more or less equally distributed with a slightly stronger signal close to the nucleus but not in the nucleus. Conversely, in the PsGPA1-silenced mutant the fluorescence in the nucleus is stronger than in the cytoplasm (Fig 5A and 5B and S7D Fig) suggesting that PsGPA1 prevents nuclear localization of PsYPK1.
This then raised the question whether the decreased level of PsYPK1 in the nucleus instigated by the presence of PsGPA1 is biologically relevant. After all, the PsYPK1 knock-out strain and PsGPA1 overexpressing strain have similar phenotypes, i.e. decreased sporangium formation and reduced virulence on P. sojae (Figs 3 and 4). To investigate this, we generated transformants co-expressing PsGPA1 and PsYPK1 and analyzed their phenotypes. To eliminate the influence of the native PsYPK1 in the wild-type, we used a PsYPK1 knockout strain as recipient strain and included a complemented PsYPK1 knockout strain as control. Consistent with the localization change described above, complementation with PsYPK1 was found to rescue the virulence defects and the sporangium production. In contrast, co-expression of PsGPA1 and PsYPK1 resulted in a significant decrease in sporangia numbers and reduced virulence on soybean (Fig 6A-6D). To further confirm that this phenotype is dependent on the interaction between PsGPA1 and PsYPK1, we also co-expressed PsGPA1 and PsYPK1 ΔN in the PsYPK1 knockout strain. PsYPK1 ΔN can no longer interact with PsGPA1 but has retained its catalytic domain that is sufficient for sporangia formation. In the PsYPK1 ΔN complemented strain, PsGPA1 overexpression did not affect sporangia production and did not change the virulence (Fig 6A-6D). Taken together, these findings point, at least in part, to the negative role of PsGPA1 in sporangium formation and support the conclusion that PsGPA1 prevents localization of PsYPK1 in the nucleus by binding to its N-terminal portion, thus regulating sporangium formation.

The PsYPK1 knockout and PsGPA1 overexpression strains are sensitive to rapamycin
In S. cerevisiae, YPK1 is directly phosphorylated by the target of rapamycin (TOR) complex-2 (TORC2) and this TOR signaling is required for many plasma membrane-and cell wall-associated events [30]. To determine whether PsGPA1 can affect the phosphorylation state of PsYPK1, we expressed PsYPK1 in wild-type, PsGPA1 overexpression or silenced strains. The phosphorylation state of PsYPK1 was tested by Pro-Q diamond phosphoprotein gel staining which can specifically reveals phosphorylated protein. In this study we included, the phosphorylated band of PsYPK1 show no changes in PsGPA1 overexpression or silenced strains compared with wild-type strain, while an obvious reduction was found in PsYPK1 incubating with phosphatase (S8D Fig), suggesting that PsGPA1 does not affect the phosphorylation state of PsYPK1.
To further insight to whether PsYPK1 is associated with TOR signaling in P. sojae, a rapamycin based growth assay was conducted. The PsGPA1 silenced mutant (PsGPA1-M27) showed the same growth rate as wild-type (P6497) with limited sensitivity to rapamycin, while a much higher growth inhibition was observed in both the ΔPsYPK1 and PsGPA1-OE strains (S9 Fig). In addition, growth assays of the same strains were performed upon exposure to oxidative, osmotic or cell wall stresses but none of these stresses had a significant effect (S10 Fig).
Although these results point to involvement of PsYPK1 in TOR signaling, there is no evidence for requirement of PsYPK1 to cope with stress conditions.

PsGPA1 and PsYPK1 co-regulate transcription of genes associated with sporangia formation in P. sojae
It has been demonstrated in S. cerevisiae, that the protein kinase Ypk1 plays an indirect role in transcriptional regulation [31,32]. Considering that PsYPK1 directly interacts with PsGPA1 and that they both affect sporangia formation, we examined whether PsGPA1 and PsYPK1 are engaged in regulating genes whose expression is associated with sporangia formation in P. sojae. We utilized RNA-Seq to analyze the genome-wide expression during sporangia formation in the PsYPK1 knockout mutant, the PsGPA1 overexpression strain, and the wild-type P6497 strain. Two independent biological replicates were analyzed by RNA-Seq and the strong correlation of gene expression levels between the replicates indicates that the data are reliable (S11A Fig). Of the more than 27 million transcript reads approximately 80% could be mapped onto the P. sojae genome. Compared to the wild-type, 963 and 559 differentially expressed genes (DEGs) were identified in ΔPsYPK1 and PsGPA1-OE, respectively (Fig 7A and S3  Table) with 249 DEGs overlapping between the two strains. The majority of these overlapping DEGs exhibited a good concordance with a correlation coefficient (R) of 0.59 (Fig 7B) and it is conceivable that these DEGs have a role in sporangium formation.
To further validate the reliability of the RNA-Seq data, 10 genes were selected for qRT-PCR analyses. This included genes whose expression in ΔPsYPK1 and PsGPA1-OE was either upregulated (Ps137349), downregulated (Ps140839, Ps129943, Ps128547, Ps132858, Ps144052, Ps132427, Ps135642, Ps135666), or unchanged (Ps108222) in comparison to wild-type. For all 10 genes the expression profiles obtained by qRT-PCR showed similar patterns of up-and downregulation or no changes (Fig 7C), thus supporting the reliability of the RNA-Seq data.
By GO annotation the 249 overlapping DEGs genes were, amongst others, classified as transcription factors (6%), metabolism-related proteins (8%), transporters (10%), kinase proteins (8%), and oxidoreductase-related proteins (8%) and E3 ubiquitin-protein ligases (5%) (Fig  7D). Most of these were significantly downregulated with the exception of the E3 ubiquitin- protein ligases that mainly showed a drastically elevated expression (Fig 7D and S3 Table). To ascertain significant pathways among these DEGs, KEGG analyses were performed. When compared with the entire genome background two pathways (glycolysis and gluconeogenesis pathways, and starch and sucrose metabolism pathways) were found to be enriched in overlapping DEGs (S12A Fig), suggesting these primary pathways might have a role in sporangium formation.
To determine whether these DEGs function in sporangia generation, we focused on those known to participate in asexual reproduction in Phytophthora species. Notably, PsMYB1, a Myb transcription factor, was clearly downregulated in both the RNA-seq and qRT-PCR analyses (Fig 7C), and this is consistent with previous findings showing that PsMYB1 plays an essential role in sporangial development in P. sojae [33]. Also in P. infestans Myb transcription factors have been implicated in sporulation. Xiang & Judelson (2014) reported induction of Myb gene expression during sporulation and found that silencing of Myb2R3 results in suppression of sporulation [34]. A previous study revealed that PsCDC14, a gene encoding cdc14 phosphatase, a cell cycle regulator that regulates nuclear behavior at an early stage of sporulation and is required for sporulation in P. infestans and P. sojae [35,36]. However, PsCDC14 expression was also not impacted by deleting PsYPK1 or overexpressing PsGPA1 (Fig 7C). These gene expression profiles point to an unconventional G protein signaling route in which GPA1 acts in association with PsYPK1 to mediate the regulation of a Myb transcription factor. The results demonstrate that PsGPA1 and PsYPK1 are both active in the pathway regulating the expression of genes predicted to be required for sporangium formation.

Discussion
Although G-protein signaling is ubiquitous in eukaryotes it is not uniform among organisms. Several basic components of G-protein signaling are highly conserved and this includes the Gα subunit of the heterotrimeric G-protein. Yet, there is a large variation in the number of different Gα subunits per species and also their downstream effectors. As a consequence, the pathways regulated by each of these individual Gα subunits vary. Phytophthora species are exceptional in this regard. They possess only a single Gα protein, and this by itself plays a central role in the life cycle and in virulence of these pathogens [26,27]. Earlier studies have demonstrated that the Gα protein in Phytophthora controls zoospore motility and chemotaxis [26,27] and that the encoding gene has a relatively high expression level in sporulating hyphae [27]. In the present study, we discovered the suppressive role of the Gα protein during the sporangium stage in P. sojae.
In both fungi and plants, the Gα protein can act as a negative regulator in several processes. In budding yeast Gα protein Gpa1 directly binds to the mating-specific MAPK Fus3 to downregulate the mating signal [37,38]. Aspergillus fumigatus has two Gα subunits GpaA and GpaB, that transduce signals leading to inhibition or stimulation of conidiation, respectively, while activating vegetative growth through modulation of the activity of a protein kinase A, i.e. PKAC1 [39,40]. Arabidopsis GPA1 interacts with phospholipase Da1 (PLDa1) and inhibits its activity by coupling with guanosine diphosphate [41], and soybean Gα proteins are negative regulators of nodule formation [42]. The inherent nature of the Gα protein includes multiple modes of regulation in which Gα can transduce the signal to negatively or positively regulate downstream pathways thus allowing for an extremely high degree of plasticity in signal responses mediated by G proteins. In our current study we show that the Gα protein in P. sojae acts as a negative regulator of sporangium development but to accomplish this, it exploits the serine/threonine kinase PsYPK1. We also show that this kinase physically interacts with PsGPA1 and thus can be considered as an immediate downstream effector of PsGPA1. Furthermore, we found that PsYPK1 contributed to sporangium formation, whereas PsGPA1 suppresses it.
PsYPK1 is a member of the AGC (PKA-, PKG-and PKC-related) family of serine/threonine kinases and homologous to S. cerevisiae YPK1. ScYPK1 and ScYPK2 are functionally redundant paralogs that represent a distinct class of fungal AGC kinases [43,44]. They both have an extended N-terminal region that has no detectable similarity to any characterized sequence motif or structural domain. Similarly, we found that PsYPK1 and its orthologs in other Phytophthora spp. have an extended N-terminal region that precedes the conserved Cterminal catalytic domain. Compared to YPK orthologs in other organisms the Phytophthora YPKs have the longest N-terminal region (S3 Fig). Despite the lack of a known domain the Nterminal region plays an important role. When deleted the localization of PsYPK1 changes with a higher abundance in the nucleus and the number of sporangia increases pointing to a negative role of the N-terminal region during in sporangium formation (Figs 5 and 6).
Such a regulatory effect has also been demonstrated for the N-terminal region of YPK1 in S. cerevisiae. For example, overexpression of a truncated version of YPK1 lacking the entire Nterminal region YPK1 (YPK1 ΔN ) was found to be toxic, whereas overexpression of the fulllength YPK1 or a kinase-dead YPK1 ΔN was not toxic. Moreover, consistent with our findings on PsYPK1, Roelants et al. showed that the N-terminal region of YPK1 is a determinant of its subcellular localization [45]. They found that the full-length YPK1 is exclusively localized to the cytosol, whereas the catalytic domain of YPK1 (YPK1 ΔN ) localizes predominantly to the nucleus [45]. The relevance of the N-terminal region is further supported by the observation that a mutated form of YPK1 with a single substitution mutation in the N-terminal region was able to rescue the nonviability of a tor2 ts mutant, suggesting that the alteration of the N-terminal domain alleviates the need for TORC2-dependent phosphorylation of YPK1 [46]. As yet, however, regulators of YPK1/2 that act via direct interaction with the N-terminal region have not been identified. Our study in P. sojae is the first to identify a protein, namely a Gα subunit, which interacts with the N-terminal region of YPK1 ortholog and that act as direct regulator of this YPK1.
We demonstrated that PsGPA1 physically interact with PsYPK1 and that both are involved in sporangium formation. Knockout mutants (ΔPsYPK1) showed strongly reduced sporangia production, a phenotype that could be rescued by complementation with PsYPK1. In contrast, targeted deletion of the N-terminal region of PsYPK1 (PsYPK1 ΔN ) or PsGPA1 overexpression resulted in increased sporangia production. The observation that PsGPA1 overexpression abolished complementation of ΔPsYPK1 by PsYPK1 but not by PsYPK1 ΔN (Fig 6) is a strong indication that PsYPK1 regulates sporangium formation in P. sojae through PsGPA1-dependent and independent mechanisms. To verify this hypothesis we compared the genome wide expression during sporangia formation in the ΔPsYPK1 strain and the PsGPA1 overexpressing strain and found that indeed, there is a large overlap in genes that are up-or downregulated in both strains when compared to the wild-type strain. The noteworthy downregulated genes seem to be involved in a range of functions and included genes or gene families that were previously identified as being associated with sporulation. One example is a Myb transcription factor that was shown to be required for sporangial development in P. sojae [33]. Our data suggest that this Myb transcription factor functions downstream of the kinase PsYPK1. Also many dehydrogenases appeared to be downregulated based on the RNA-Seq analysis and this is in line with the report of Kim & Judelson who found an higher expression of dehydrogenase encoding genes in sporangia than in nonsporulating hyphae [47]. Moreover, the dramatically decreased expression of carbohydrate-active enzymes, including glycosyl hydrolase, cellulose, glucosidase, and carbohydrate-binding protein in both the ΔPsYPK1 and PsGPA1 overexpressing strains, is consistent with the finding that TORC2-Ypk1 action was connected to NADPH requirement and the pentose phosphate pathway [48].
Apart from the overlap in up-and downregulated we found many genes that are differentially expressed between the PsGPA1 overexpressing strain and ΔPsYPK1 mutant. In addition, comparing to overlapping DEGs, a larger number of DEGs from ΔPsYPK1 mutant enriched in metabolic pathways (S12B Fig). This is not unexpected as it aligns with the idea that PsYPK1 functions in both PsGPA1-dependent and independent pathways and with the incomplete matching of phenotypes of the two strains (Fig 7D and S4B Fig). For example, in the ΔPsYPK1 strain, many more carbohydrate-active enzymes were found to be downregulated, which may explain the increased severity of defects observed in this strain. Furthermore, from the virulence phenotypes it is also obvious that PsGPA1 and PsYPK1 affect the virulence of the pathogen at different levels. For PsGPA1 mutants, the reduction in virulence is mainly caused by a defect in chemotaxis and zoospores are no longer specifically attracted by the host plant soybean [27,28]. It thus seems that PsGPA1 has a major role in an early infection stage with zoospores as infectious propagules (S6C and S6D Fig). The reduction of virulence in PsYPK1 mutants is mainly due to defects in the growth of invasive hyphae (Fig 3C and 3D). Therefore, the PsYPK1 mutant exhibits reduced virulence under all infection conditions (S6C  and S6D Fig).
In conclusion, we propose a new model addressing the role of the G protein α subunit in the asexual development of P. sojae. In this model Gα acts as a negative regulator by binding to the N-terminal region of PsYPK1, an AGC type serine/threonine kinase that is essential for sporangium production. This binding prevents localization of PsYPK1 in the nucleus, thereby repressing expression of its target genes which are likely essential for sporangium formation.

P. sojae strains and culture conditions
The genome-sequenced P. sojae strain P6497 (Race 2), generously provided by Dr. Brett Tyler (Department of Botany and Plant Pathology, Oregon State University, Corvallis, OR, USA), served as the wild-type strain, and all transgenic lines in this study were routinely grown on 10% V8 agar medium at 25˚C in the dark (Erwin and Ribeiro, 1996). The asexual life stages, such as vegetative hyphae, sporulating hyphae, zoospores, cysts, and germinated cysts, were collected as previously described [27]. The growth rates of the transformants were calculated using cultures growing on Plich agar medium. Plich medium is minimal medium with poor nutritional content. It contains per liter 0.5 g KH 2 PO 4 , 0.25 g MgSO 4 �7H 2 O, 1 g asparagine, 1 mg thiamine, 0.5 g yeast extract, 10 mg b-sitosterol, 25 g glucose, and 15 g agar (26,48,49). Sporulating hyphae were prepared by repeatedly washing 2-day-old mycelia grown in 10% (v/ v) V8 broth with sterile distilled water (SDW), followed by incubation in the dark at 25˚C for 4-8 h, until sporangia developed. Zoospores were produced as described above.

Co-immunoprecipitation (Co-IP) for selecting PsGPA1 interacting proteins
Constructs carrying 3×FLAG-tagged PsGPA1 and 3×FLAG-tag (control) in the vector pTOR were transformed into P. sojae. Stable P. sojae transformants were selected for total protein isolation and total protein lysates were analysed by western blotting using anti-FLAG antibodies (Abmart Inc., Shanghai, China). This revealed a band of 42 kDa, representing the PsGPA1-3×FLAG fusion protein that was present in the transformants, but not in the control (S1A Fig). Protein lysate from sporulating hyphae of PsGPA1-3×FLAG expressing transformants was mixed with anti-FLAG M2 beads. The bead-bound proteins were then eluted and analyzed by MS.

Targeted gene deletion and complementation
Gene deletion mutants were generated using the CRISPR-mediated gene replacement strategy [6]. The NPTII gene ligated with two 1.0-kb fragments flanking the target gene was used as donor DNA in homology-directed repair (HDR) (S4B Fig). We transformed P. sojae using PEG-mediated protoplast transformation [27]. Putative transformants were selected by growth on 10% (v/v) V8 medium containing 50 μg/mL geneticin and screened by PCR using primer combinations as shown in S2B Fig. Primer set F1/R1 (S2 Table) located within the PsYPK1 ORF was used to screen for deletion of PsYPK1 in the genome of resistant transformants. Primer sets F2/R2 and F3/R3 (S2 Table) were used to detect homologous recombination events (S4B and S4C Fig). HDR events were analysed by Sanger sequencing to confirm that the PsYPK1 ORF was cleanly replaced (S4D Fig) and qRT-PCR was used to demonstrate absence of PsYPK1 (S4B and S4E Fig).
For complementation, the knockout mutant was transformed using NPTII as a selection marker and with a construct containing the entire gene coding region with mutated sgRNA sites inserted into the pTOR vector.

RNA extraction and gene expression analysis
Total RNAs from distinct stages of the life cycle of P. sojae (vegetative hyphae, sporulating hyphae, zoospores, cysts, germinated cysts and stages of infection) were extracted as described previously [27,49]. RNA integrity was tested via agarose gel electrophoresis. Prior to cDNA synthesis, all RNA samples were DNase I-treated using a DNase kit, following the manufacturer's protocol. First strand cDNA was synthesized from 1-5 μg of total RNA using oligo (dT) primers with an M-MLV reverse transcriptase kit (Takara Bio, Inc.), following the manufacturer's protocol. Semi-RT-PCR was performed to amplify the actin A gene (ACT-RTF/R) and PsYPK1 (PsYPK1-QRT-F/R). SYBR green qRT-PCR assays were used to measure PsYPK1 expression with the aid of the primer pair PsYPK1-QRT-F/R (S2 Table). All reactions were performed on an ABI 7500 Fast Real-Time PCR System (Applied Biosystems, Inc., Foster City, CA, USA). The results were analyzed using ABI 7500 sequence detection software. Relative gene expression levels were determined using actin A (primers ACT-RTF/R S2 Table) levels as internal controls. Means and standard deviations were calculated using data from three replicates.
RNA for RNA-seq analyses was isolated from mycelium that was cultured on 10% V8 broth at 25˚C in the dark for 2 days. The mycelium was rinsed twice with sterile distilled water and flooded with sterile distilled water. After 6 h, mycelium was collected for RNA extraction.

Growth, oospore and sporangium production
To determine growth rate, all strains were grown on V8 and Plich medium at 25˚C in the dark. Colony diameters on Plich medium were measured during 7 days and on V8 medium during 4 days and the average growth rates were calculated (mm/day).
To monitor and quantify oospore production, strains were grown on lima bean agar (LBA) medium at 25˚C in the dark. After 10 days the cultures were examined by microscopy. Three blocks of 1 cm 2 were cut from the medium and in each block three random fields at 40×magnification were selected for counting the number of oospores.
To quantify sporangia production 10% V8 broth was inoculated with three round mycelial disks (7 mm in diameter) cut from a culture, and incubated for 48 h at 25˚C in the dark. The mycelia were then rinsed twice with sterile distilled water, then flooded with sterile distilled water and left for 8 h to stimulate sporangia formation. This was then gently mixed in a blender to obtain a homogenous mixture. Subsequently, three random samples of 100 μl were taken and by microscopic examination at 40×magnification the number of sporangia in each sample was counted. All assays were repeated at least 3 times.

Virulence assays
The soybean cultivar Hefeng 47, which is susceptible to P. sojae strain P6497, was grown in plastic pots containing vermiculite at 25˚C for 4 days in the dark. Zoospores were retrieved as described previously [27] and diluted to a concentration of 100 zoospores/10 μl. Etiolated seedlings were inoculated by pipetting 10 μl of the zoospore suspension on the hypocotyls, and maintained in a climate-controlled room at 25˚C and 80% relative humidity in the dark.
Manifestations of pathogenicity were evaluated at 3 dpi, and photographs were taken. To allow microscopic observation of the penetration of, and infectious hyphal expansion within soybean tissue, infected epidermal cells were collected at 12 and 24 hpi and soaked in Trypan Blue. After destaining in chloral hydrate and water, the infected epidermis cells were examined under a light microscope. Each strain was tested using at least two different preparations of zoospores, and five plants. All assays were repeated at least three times.

Interaction assay in vivo and in vitro
For the in vivo assays, FLAG-tagged PsGPA1 and GFP-tagged PsYPK1 were stably coexpressed in P. sojae. Mycelia of co-expressing transformants were harvested and suspended in lysis buffer. The samples were centrifuged at 4˚C for 10 min at 12,000 × g to remove debris and the protein lysates (supernatants) were transferred to a new tube. These were incubated at 4˚C for 3-4 h with 25 μl of GFP-Trap_A beads (Chromotek, Planegg-Martinsried, Germany). The beads were then collected by centrifugation at 2500 × g and washed three times in 1 mL of washing buffer (per manufacturer's recommendations). Bound proteins were boiled for 5 min and the presence of FLAG proteins was detected by western blotting using anti-FLAG M2-Peroxidase (HRP) antibody (#A8592; Sigma-Aldrich).
For the in vitro assays, the coding sequences of PsYPK1 and PsGPA1 were inserted into a pET28a vector (containing the His tag) and pGEX-4T-2 vector (containing the GST tag), respectively (GE Healthcare Life Science). The plasmids (pET28a empty vector, His-PsYPK1, GST empty vector, and GST-PsGPA1), were transformed to E. coli strain BL21 (DE3), protein expression was induced with 0.1 mM IPTG and after 12 hours of growth protein lysates were obtained. The pull-down assay was performed using the ProFound pull-down GST proteinprotein interaction kit (Pierce) according to the manufacturer's instructions. The soluble total GST-fusion proteins and His proteins were incubated with 25 μL glutathione agarose beads (Invitrogen) for 8 h at 4˚C. Then, the beads were washed three times, and the presence of His proteins was detected by western blotting using anti-His antibody (Abmart, Inc.).

Pro-Q analysis of protein phosphorylation
The Pro-Q Diamond Phosphoprotein Gel Stain (Invitrogen) selectively stains phosphoproteins in polyacrylamide gels. Total proteins were isolated from stable P. sojae transformants expressing PsYPK1-GFP-tagged proteins and incubated with anti-GFP affinity beads. Proteins bound to resins were eluted after a series of washing steps, as described in the manufacturer's instructions. The purified proteins were separated by SDS-PAGE and detected by a standard Pro-Q Diamond Phosphoprotein Gel Stain system (Invitrogen, P33300). The stained gels were visualized on a blue light trans-illuminator using a 532-nm laser and a 560-nm longpass as the excitation source and emission filter, respectively (Typhoon Trio + , Amersham Biosciences).
After obtaining results with Pro-Q Diamond Phosphoprotein Gel Stain, the total protein in the gel was stained with SYPRO Ruby Protein Gel Stain to ascertain the relative phosphorylation state of the proteins (Invitrogen, S12001).

RNA-seq analysis
The resulting high-quality reads which had been filtered by fastQC were aligned to the complete genome assemblies of Phytophthora sojae (v1.1 for isolate P6497; jgi.doe.gov). A total of two mismatches and gaps per read were allowed, and only when both of a pair of reads were successfully mapped, the data was included in the analyses. Transcript abundance was indicated as Fragments Per Kilobase of exon model per Million mapped reads (FPKM). EdgeR software (www.bioconductor.org/packages/release/bioc/html/edgeR.html) was used to perform differential expression analysis between duplicate samples for all genes. To identify differentially expressed genes (DEGs), read counts for each gene model were obtained using featureCounts software (subread.sourceforge.net), then the log 2 fold change (log 2 FC) value and adjusted P value were calculated using DESeq2 software (www.bioconductor.org/ packages/release/bioc/html/DESeq2.html), and genes with an adjusted P value < 0.05 and an absolute |log 2 FC| � 1 were considered differentially expressed. (C) Analysis of genomic DNA from all the strains mentioned above using the full-length primer of PsYPK1. (D) PsGPA1 cannot affect the phosphorylation state of PsYPK1. PsYPK1 protein was expressed in wild-type, PsGPA1 overexpression or silenced strains for the phosphorylation assay, or incubating with phosphatase as a control (marked as PsYPK1-GFP, PsGPA1-FLAG/PsYPK1-GFP and PsGPA1-M27/PsYPK1-GFP, PsYPK1-GFP/ phosphatase, respectively) and the gel image was visualized after Pro-Q Diamond phosphoprotein gel staining. In all cases, the upper panel represents Pro-Q Diamond-stained phosphoprotein gel and the two bottom panels show the same protein samples using western blot analysis to display the PsYPK1-GFP or PsGPA1-FLAG protein. Numbers represent relative signal intensities. Western blot analysis of PsYPK1-GFP was used as a loading control. The rich factor, number of genes, and q value were used to measure the degree of KEGG enrichment. When a pathway with greater rich factor, larger number of genes and, q value is less than 0.05, the enrichment is more significant. (TIF) S1