A new identified suppressor of Cdc7p/SepH kinase, PomA, regulates fungal asexual reproduction via affecting phosphorylation of MAPK-HogA

The septation initiation network (SIN), composed of a conserved SepH (Cdc7p) kinase cascade, plays an essential role in fungal cytokinesis/septation and conidiation for asexual reproduction, while the mitogen-activated protein kinase (MAPK) pathway depends on successive signaling cascade phosphorylation to sense and respond to stress and environmental factors. In this study, a SepH suppressor–PomA in the filamentous fungus A. nidulans is identified as a negative regulator of septation and conidiation such that the pomA mutant is able to cure defects of sepH1 in septation and conidiation and overexpression of pomA remarkably suppresses septation. Under the normal cultural condition, SepH positively regulates the phosphorylation of MAPK-HogA, while PomA reversely affects this process. In the absence of PbsB (MAPKK, a putative upstream member of HogA), PomA and SepH are unable to affect the phosphorylation level of HogA. Under the osmostress condition, the induced phosphorylated HogA is capable of bypassing the requirement of SepH, a key player for early events during cytokinesis but not for MobA/SidB, the last one in the core SIN protein kinase cascade, indicating the osmotic stimuli-induced septation is capable of bypassing requirement of SepH but unable to bypass the whole SIN requirement. Findings demonstrate that crosstalk exists between the SIN and MAPK pathways. PomA and SepH indirectly regulate HogA phosphorylation through affecting HogA-P upstream kinases.


Introduction
The filamentous fungus contains a mycelium of multinucleated cells that are partitioned by septa, and timely cytokinesis and septation is essential for hyphal growth and conidiation [1,2]. Conidia are the primary means of asexual reproduction and dispersal in the environment for a variety of fungi [3][4][5]. Generally, fungal cells respond to environmental stimuli through signaling pathways often involving kinase cascades that transmit external cellular signals to the nucleus [6][7][8]. Genetic analyses have reported that the model filamentous fungus Aspergillus nidulans and Schizosaccharomyces pombe have components of the septum initiation network (SIN), and septum formation requires the assembly of a septal band composed of a dynamic protein complex that is dependent upon conserved core SIN components including three protein kinases (SepH/Cdc7p, SidA/Sid1p and SidB/Sid2p), and their regulatory partners (SpgA/ Spg1p, Cdc14p and MobA/Mob1p), respectively [2,9,10]. Notably, downregulation of the serine/threonine protein kinase SepH cascade in A. nidulans would abolish septation and conidiation, its hyperactivation of it would induce the formation of multiple septa. Thus, SepH is a positive regulator of the SIN that triggers septation/conidiation in A. nidulans [9]. However, whether negative regulators exist for conidiation is elusive. In addition, individual cell signaling pathways can communicate with each other, thereby building a complex network to ultimately control gene expression and other cellular functions [11]. In all eukaryotic kingdoms, mitogen-activated protein kinases (MAPKs) play critical roles in cellular responses to environmental cues [12]. These MAPKs are activated by phosphorylation at highly conserved threonine and tyrosine residues [13] in response to specific inputs, leading to their accumulation in the nucleus and the activation of their downstream targets [14]. Three-tiered core signaling modules, including MAPK kinase kinases (MAPKKK), MAPK kinases (MAPKK) and MAPK, are included in the MAPK signaling pathway [12,15]. Cells respond to different external environment stresses, including pH, light signaling, and osmotic or oxidative stress, by phosphorylating the kinase cascade to activate the MAPK signaling pathway [11,14,16,17]. Thereafter, MAPK is phosphorylated and shuttles into nuclei upon stimuli to trigger gene expression. Thus, reversible protein phosphorylation plays a critical role in the regulation of virtually all eukaryotic biological processes, and the MAPK signaling pathway plays a crucial role in different organisms from mammals to fungi. In mammalian cells, the MAPK signaling pathway is critical for normal immune and inflammatory responses and contributes to cell cycle or cytoskeleton remodeling [18]. In model yeasts, the MAPK signaling pathway involves many biological processes, including cell wall integrity, cell separation and the regulation of mitotic spindle disassembly [15,19]. In most filamentous fungi, the high-osmolarity glycerol (HOG) gene hogA, which encodes a MAPK that sense osmotic and oxidative stress, is required for hyphal growth and conidiation under osmotic stress culture conditions [14,20]. Moreover, in the opportunistic fungal pathogen A. fumigatus, the high-osmolarity glycerol (HOG) MAPK signaling pathway is crucial for virulence and conidiation under stress conditions [12,21]. Therefore, both the SepH-SIN kinase cascade and the HOG-MAPK pathway are involved in cell growth and reproduction by regulating reversible protein phosphorylation. However, whether crosstalk between the SIN cascade and the HOG-MAPK pathway exists remains unknown.
Through UV mutagenesis, our previous studies identified antagonizing components of the SepH kinase cascade [22]. However, this complex protein cascade has not yet been dissected since the majority of mutants isolated by UV mutagenesis have no clear phenotype. Therefore, it is difficult to confirm these related mutations in the complementation assay by transforming a whole-genome library since a critical barrier in forward genetics is to verify gene functions with no detectable phenotypes. With the development of next-generation sequencing (NGS), it is possible to identify these mutations by comparing single nucleotide polymorphisms (SNPs) between mutants and wild type which all have the same parental genetic background. Based on these approaches, in this study, through backcross techniques combined with a new mutation comparison analysis of NGS, a putative protein kinase, PomA, a suppressor of SepH, was identified to be involved in septation and conidiation. Moreover, quantitative phosphoproteomics revealed that deletion of pomA causes activation of the HOG MAPK signaling pathway.

Screening and isolation of the sepH1 suppressors of septation and conidiation
Our previous study suggested three categories among the 116 independent UV-induced suppressor mutations of sepH1 based on colony phenotype [22]. To further dissect these suppressor-related mutations, we backcrossed these conidiation-restored mutants with the wild type strain R21, which has a yellow colony selection marker (yA1). Consequently, the isolated progenies from mutants S11 and S53 (suppressor of sepH1, No. 11 and No. 53, respectively) crossed with wild type R21 clearly displayed the predicted phenotypic separation as shown in S1 Fig. The ratio of the four different types of progeny colonies (n = 193 for S11 and n = 218 for S53) was approximately 1:1:1:1 for which genotypes were supposed to be wild type, sepH1 (referred to as sepH1-1 and sepH1-3, respectively), single mutations No. 11 and No. 53 (referred to as sin11 and sin53, respectively) and double mutations sepH1, sin11 or sin53 (S2 Fig). To further confirm whether sin11 and sin53 truly act as suppressors of sepH1 on conidiation and septation, we further crossed the isolated sin11 and sin53 mutants with sepH1. Consequently, four different colony phenotypes were observed again at 30˚C and 42˚C, suggesting that progenies may include four different genotypes: sin11 or sin53 and sepH1, wild type strains and the putative double mutants sepH sin11 and sepH sin53 (S2 Fig). These data suggest that sin11 and sin53 are extra-mutations from sepH1. Thus, we next selected only S11 and S53 to further examine the phenotypes. As shown in S3A Fig, the production of conidia and the development of conidiophores were rescued to some extent in mutants S11 and S53 compared to that of the sepH1 mutant with a completely abolished conidiophore structure under the same cultural restriction temperature (42˚C) (Fig 1A and 1B). Consistently, liquid culture experiments demonstrated that both S11 and S53 displayed calcofluor white (CFW)-stained septa in hyphal cells compared to sepH1 with no detectable sign of the septum in hyphal cells at 42˚C (Fig 1C  and S3B Fig), suggesting that defects of septation in sepH1 at 42˚C could be suppressed by mutation in S11 or S53. To further verify the phenotypes from isolated sepH1 progenies, hyphal cells of sepH1-1 and sepH1-3 were stained with CFW. No stained septa were observed in sepH1-1 or sepH1-3, while double mutants sepH sin11-2 and sepH sin53-2 displayed CFWstained septa in hyphal cells at 42˚C (S2C Fig), suggesting that mutations sin11 and sin53 truly suppress the defects of sepH1 on septation. In addition, strains sin11 had slightly reduced colony sizes compared to the wild type strain R21 at 42˚C (Fig 1D and S3D Fig). Quantification of septa, conidia and colony diameters for indicated strains as shown in S3 Fig. To further visualize the cellular phenotype caused by mutations sin11 and sin53, a red fluorescent protein (RFP) tag was tagged to the N-terminus of histone H2A, a nuclear protein marker, in the respective background of wild type, sin11 and sin53; the resulting strains were named ZXA05, ZXA06 and ZXA07. Distribution of septa stained by CFW and nuclei fused RFP with a nuclei marker histone H2A were displayed in related strains. Fluorescence microscopy revealed no detectable difference in the distribution of nuclei or septa, even for hyphal polarity growth in the isolated sin11 and sin53 mutants ( Fig 1E). Therefore, these data collectively imply that the mutation sin11 or sin53 may not cause the remarkable growth phenotype (only with slightly slow colony growth), but is able to suppress defects of sepH1 function for septation.

A putative protein kinase, PomA, as a suppressor of SepH, is involved in septation
To further identify which mutated gene contributes to the suppression of sepH1 in septation and conidiation, SNP comparison by NGS for the whole genomes of S11 and S53 was carried out. Since mutants S11 and S53 were derived from the same background strain, sepH1 (compared to the A. nidulans A4 genome database), their SNPs should be consistent except for those mutations induced by UV in sin11 and sin53. Total 1215 and 1085 SNPs were found, while 273 and 282 belonged to missense mutations occurring in the coding region in S11 and S53, respectively (Fig 2A and S1 Data File). Given that mutations that contributed to the suppression of sepH1 in S11 and S53 should differ since they were independently isolated from UV-induced mutants, the overlapping common SNPs were excluded. As a consequence, 15 and 24 specific SNPs remained in S11 and S53, respectively (Fig 2B). After a BLAST analysis of the whole genomic sequence assembly compared to the A. nidulans FGSC A4 (http://www. aspgd.org/) genome for these specific SNPs in S11 and S53, interestingly, multiple different SNPs were located in three genes (ID numbers: AN7678, AN7949 and AN10000), while AN7678 had a high-quality score. Hence, we next determined whether the AN7678 mutation contributed to the suppression of sepH1 on septation and conidiation. According to information at FungiDB (http://fungidb.org/fungidb/), the gene AN7678 has 4482 nucleotides and encodes a putative 1417-aa protein. From information obtained from a homology search by Blast-P in the S. pombe database, AN7678 shares 58% identity to that of Pom1, a DYRK family cell polarity protein kinase in S. pombe, which, according to SMART analysis (http://smart. embl-heidelberg.de/), harbors a conserved domain, S-TKc (Serine/Threonine protein Kinases, catalytic domain), ranging from 1003 to 1299 aa. Conserved domain comparison of the S-TKc domain between Pom1 in S. pombe and AN7678 indicated high identity (68%). Hence, these results suggest that AN7678 is a homolog of Pom1 in A. nidulans; thus, we then named AN7678 as PomA. According to SNP sequencing and alignment information, we found that S11 harbors a mutation in AN7678 in which leucine 1265 is replaced by serine, which is located in the S-TKc domain. Notably, the mutation occurring in S53 at codon 977 produces a stop codon before the S-TKc domain ( Fig 2C). Further cloning and sequencing of the pomA gene from mutants S11 and S53 consistently confirmed that the aforementioned mutations truly happened ( Fig 2D). To further demonstrate whether septum recovery and the suppression of sepH1 in S11 and S53 were a specific result of the PomA mutation, we then generated a site- directed mutation strain, PomA L1265S , that mimics the mutation in S11 (Leu to Ser at the position 1265) and a full-length pomA deletion strain in the background of sepH1 and its parental wild type strain. As shown in Fig 3A, there were no significant different colony phenotypes in the pomA deletion, PomA L1265S mutant compared to that their parental wild type. However, in double mutants sepH1, ΔpomA and sepH1 PomA L1265S , there were some rescued conidiation on agar-solid media and recovered septation in liquid media (Fig 3A and 3B, S3B Fig), suggesting both the pomA null and PomA L1265S site-directed mutants were able to suppress septation and conidiation defects induced by sepH1 (Fig 3A and 3B). Thus, these data repeatedly demonstrated that double mutations of sepH and pomA were capable of recovering septation and conidiation.
In addition, to provide further insights into the functions of PomA, a conditional alcA(p):: GFP-pomA strain was constructed by homologous integration in which the pomA gene was under the control of different carbon source de-repression/repression by glycerol/glucose and induction by threonine. GFP-PomA showed a similar pattern of localization with the highly accumulated at the tips of hyphal cells and the analogous spindle pole body near the nucleus when cultured either in de-repression glycerol or induction-threonine media (Fig 3C), suggesting that PomA may be involved in cell division or hyphal cell polarity growth. However, the ΔpomA and PomA L1265S mutants showed almost similar colony phenotypes compared to their parental wild type strain regardless of culture conditions (in rich media YAG or supplemented with 1 M NaCl or KCl) (

Quantitative phosphoproteomics revealed that the deletion of pomA causes a highly activated HOG pathway
Considering that homologs of PomA and SepH in yeasts (Pom1 and Cdc7p) belong to the protein kinases that execute major functions by phosphorylating their related substrates, we then carried out tandem mass tag (TMT)-based quantitative phosphoproteomics to analyze the overall phosphorylation levels in the septa-restored strain sepH1 ΔpomA (ZXA10) and the septa-abolished control strain sepH1. A total of 8409 phosphorylation sites involved in 2530 proteins were identified, in which 6766 phosphorylation sites mapped to 2188 proteins had convincible quantitative detection ( Fig 4A and S2 Data File). Using a cut-off was considered to be statistically significant (p-value < 0.05) for 1.2-fold changes in which fold changes has a ratio-compression effect result from isobaric mass tagging techniques for the phosphorylated proteins [23]. Normalized quantitative proteomics revealed 389 (293) increased and 354 (233) decreased phosphorylation modification sites (proteins) in sepH1 ΔpomA compared to sepH1 ( Fig 4B). According to GO categories, these phosphorylation modification site-containing proteins are enriched in cellular and metabolic processes ( Fig 4C). Moreover, using protein pathway annotation by the Kyoto Encyclopedia of Genes and Genomes (KEGG) database, for which p-values are less than 0.05, these phosphorylation-increased proteins were mainly mapped to four KEGG signaling pathways ( Fig 4D). Among them, the HOG MAPK signaling pathway gained the highest score, suggesting that the loss of pomA enhances phosphorylation levels in the MAPK pathway. To further verify these phosphorylation-changed proteins-  involved in pathways, we re-calculated data in phosphoproteomics based on increased value of cutoff to 1.3-fold. As shown in S5 Fig, these phosphorylation-increased proteins still are mapped to the MAPK signaling pathway, suggesting this pathway may involve in defected pomA-induced septation. The quantitative phosphoproteomics was normalized by the proteomics of total proteins while expression of HogA was too low to be considered based on the measurement probably due to strains were cultured under the normal cultural condition. Based on fission yeast S. pombe MAPK signaling pathway information, further visual analysis revealed six homologous components in the MAPK signaling pathway (labeled as "red" in Fig  4E) with increased phosphorylation levels in the sepH1 ΔpomA double mutant compared to the sepH1 single mutant, indicating that a lack of PomA causes increased phosphorylation of MAPK. Therefore, quantitative phosphoproteomics data imply that PomA works as a negative regulator to regulate the phosphorylation level of MAPK.

SepH and PomA coordinately regulate the phosphorylation of HogA
To further determine whether phosphorylation levels in the MAPK pathway could be affected by SepH or PomA, we next verified the phosphorylation status of the MAPK HogA (yeast S. cerevisiae Hog1 homolog in Aspergillus), the final-step kinase in this pathway that is phosphorylated in response to external environment stresses in the activated MAPK cascade. Western blot analysis revealed a clear band of the predicted size in the A. nidulans wild type strain probed with the antibody against phospho-p38 MAPK (HogA homolog in mammals) (S6 Fig). Upon deleting the hogA gene or three key amino acids for predicted phosphonate sites at HogA (Fig 5B), this predicted HogA-P band was completely disappeared, indicating that this antibody might be able to recognize HogA-P (S6 Fig). Therefore, we used this antibody to further analyze the phosphorylation status of HogA induced by the high-osmolarity stimulus, NaCl (1 M). The protein band intensity was remarkably increased in the NaCl-induced sample compared to the control (Fig 5A), suggesting that the phosphorylation level of HogA is truly increased after treatment with NaCl for 10 min, further implying that this antibody is able to recognize HogA phosphorylation. Thus, we used this antibody to determine whether there were any altered phosphorylation levels in the aforementioned pomA-and sepH1-related mutants. Compared to the parental wild type strain, sepH1 caused reduced expression of HogA-P while deletion of pomA or site-directed PomA L1265S increased HogA-P expression (Fig 5A, 5D and 5F). Notably, in the background of sepH1, lack of PomA recovered HogA-P expression to a level similar to that of wild type (Fig 5F). These data suggest that SepH and PomA coordinately regulate the phosphorylation of HogA, likely playing a key role in septation and conidiation. Moreover, to further investigate whether the upstream kinase of HogA is required for phosphorylation of HogA regulated by SepH and PomA, we deleted PbsB (MAPKK, a putative upstream member of HogA) in the background of wild type (TN02A7), ΔpomA and sepH1, respectively. As shown in S7 Fig, all ΔpbsB related strains showed no detectable band of HogA-P in the Western blotting compared to its parental wild type with a predicted HogA-P band, suggesting PbsB is required for phosphorylation of HogA. In the absence of PbsB, PomA and SepH were unable to affect the phosphorylation level of HogA which suggest that PomA and SepH indirectly regulate HogA phosphorylation may via affecting HogA-P upstream kinases. Meanwhile, it also suggests that the antibody could recognize HogA-P. We next tested whether the increased expression of HogA-P induced by osmotic stress could bypass the requirement of SepH. Hyphal cells of the sepH1 mutant cultured in liquid MM medium with NaCl or KCl were stained with CFW, which clearly showed that the sepH1 mutant was able to produce septa culturing at 42˚C (Fig 5C), suggesting that osmotic stress bypasses the requirement of sepH1. This result also suggests that septation defects in sepH1 may be due to the reduced expression of HogA-P. These phenomena also coincided with the expression level of HogA-P in the sepH1 mutant, suggesting that the phosphorylated activation of MAPK may play important roles in septation and conidiation. In contrast, the deletion or turn off of pomA expression in a conditional strain enhanced the expression of HogA-P (Fig 5D and 5E), demonstrating that PomA works as a negative regulator for the phosphorylation of HogA. We next deleted hogA gene or three key amino acids for phosphonate sites at HogA in the background of sepH1 to obtain a double-mutation strain hogA, sepH1 for which osmotic stresses were no longer able to induce septation again, further demonstrating HogA is required for bypassing sepH1 to respond stress stimuli for septation ( Fig 5C). Moreover, to gain insights in the actual cross-talk point between the HOG and the SIN, we carried out a strains-crossing technique to make a conditional stain alc (p):: GFP-mobA, sepH1 in which MobA is a partner of SidB, a downstream kinase of SepH in the SIN pathway. In this strain, MobA was conditional expressed and labelled by GFP. As shown in S8A and S8D Fig, turn-off of mobA was no longer to induce septation by the osmotic stress stimulus although NaCl addition still could induce enhanced expression of HogA-P. Comparably, deletion of a mobA paternal, sidB, also was unable to induce septation (S8B Fig). These data collectively suggest that complex MobA/SidB, downstream targets of SIN-SepH, is required for septation even under the osmotic-stress condition. Moreover, in the background of sepH1, we found that GFP-MobA from observation for localization, could not translocate to the septum site instead still localized at the site of the spindle pole body no matter under the normal or the osmotic stress culture condition (S8E Fig). In addition, deletion of pomA was unable to rescue septation when turn-off expression of MobA as shown in S8C Fig. These data imply that PomA is a suppressor of SepH but not for MobA/SidB and SepH is required for the MobA translocation to septa. Thus, the osmotic stimuli-induced septation is capable of bypassing requirement of SepH but unable to bypass the whole SIN requirement.

AnkA, a yeast S. pombe wee1 homolog in A. nidulans has a similar suppression function with PomA against SepH
In fission yeast S. pombe, Pom1 regulates the cell cycle by modulating the activity of a Pom1/ Cdr2/Wee1 geometry network in which Pom1 negatively regulates Cdr2, a previously described Wee1 inhibitor, suggesting that Pom1 and Wee1 play a similar role in the cell cycle [24][25][26][27]. To address whether ankA has suppression function to sepH1 in A. nidulans, we deleted ankA, a wee1 homolog in A. nidulans [28,29], in the background of sepH1, which was named ZXA17 (Fig 6A). Consequently, ΔankA sepH1 exhibited rescued septation accompanied with more robust colony growth and abundant conidia production than that of the sepH1 mutant ( Fig 6A) and rescued septation (Fig 6B and S3B Fig) under the condition of a restrictive temperature (42˚C), suggesting that ankA is able to suppress the sepH1 defects in septation and conidiation, consistent with that of pomA. Therefore, these data suggest that A. nidulans may also have a PomA/Cdr2homolog/AnkA (Wee1) geometry network that is similar to that of S. pombe Pom1/Cdr2/Wee1. Taken together, PomA suppresses the defects caused by the loss of function of SepH, which may regulate MAPK activity via the PomA/AnkA network to suppress the defects of SepH (Fig 6C).

Discussion
In this study, by SNPs comparison and the backcross techniques between the mutant and the same parental background wild-type strain, we isolated a new suppressor of SepH (an SIN cascade component), PomA. Our principal conclusions are as follows. First, a putative protein kinase, PomA, in A. nidulans localized with a spindle pole body-like pattern and the tips of hyphal cells, as a suppressor of SepH, is involved in septation and conidiation such that lack of PomA cures defects of septation and conidiation in sepH1 while overexpression of PomA remarkably abolishes septation and condiation. Second, SepH positively regulates the phosphorylation of HogA, while PomA reversely affects this process. Third, the phosphorylation of HogA induced by osmotic stress bypasses the requirement of SepH to fulfill septation and conidiation. We demonstrated that crosstalk exists between the SIN and MAPK pathways such that PomA and SepH indirectly regulate HogA phosphorylation through affecting HogA-P upstream kinases.
Several lines of evidence indicate that SepH, a homolog of serine-threonine kinase Cdc7p in fission yeast, is a main positive regulator in the SIN in A. nidulans [9,30,31]. The sepH1 mutant with completely abolished septation and fluffy colony phenotypes was previously isolated by screening in temperature-sensitive cytokinesis mutants [9,22,31]. Pom1, a dualspecificity tyrosine phosphorylation-regulated kinase in S. pombe [32], inhibits the SAD-like kinase Cdr2 and Mid1/anillin to negatively control cell division plane positioning and restricts the cell middle cortical cytokinetic ring precursor nodes through an unknown mechanism [32][33][34][35]. The Pom1 mutants divide at a significantly shorter size [32] than wild type cells and some mutants display an abnormal growth axis, resulting in angled and branched cells [36]. These findings suggest that Pom1 in S. pombe is required for the timing of cytokinesis and for normal polarized cell growth [32,34,[36][37][38][39]. However, the functions of the Pom1 homolog PomA in A. nidulans have not yet been reported. Our findings suggest that the loss of pomA causes no detectable difference in hyphal cell polarized growth or in septation compared with the wild type strain, even though PomA contains the SPB and exhibits cell tip localization patterns that are similar to those of Pom1 (Fig 3C-3F). This phenomenon may also suggest other alternative candidates to bypass the requirement for polarity growth and for septation in the absence of PomA. In contrast, the overexpression of pomA caused significantly reduced colony growth accompanied with abolished septation/conidiation (Fig 3D and 3G), suggesting that PomA is a negative regulator of this process and also implying PomA is a potential efficient inhibitor for conidiation. Our findings demonstrated that osmotic stressors NaCl or KCl were able to induce septation and conidiation regardless of the absence or presence of the SIN component SepH (Fig 5C). These data indicate that when fungi sense environmental stresses, they bypass the originally required SIN protein cascade-SepH to fit the environmental niches by highly expressing phosphorylated HogA. These findings also suggest that phosphorylated HogA plays important roles in septation and conidiation. Accordingly, loss of the SIN component SepH caused reduced expression of HogA-P (Fig 5A and 5F) with completely abolished septation/conidiation, while the deletion of PomA rescued these defects induced by defects of sepH as well as suppressing decreased HogA-P expression in the sepH1 pomA double mutant (Fig 5F). However, we still could not exclude another possibility for a full-length SepH protein may be also required for the osmotic stimuli-induced septation since sepH1 is a temperaturesensitive cytokinesis mutant not a null mutant. Our new findings also suggest that the HogA phosphorylation was able to bypass requirement of SepH but not for MobA/SidB since deletion of pomA was unable to rescue septation when expression of MobA was turn-off even under the osmotic stress cultural condition. SepH, a key player for early events during cytokinesis while SidB is the last kinase in the core SIN protein kinase cascade. These findings demonstrate that MobA and SidB are required for septation and may also imply HogA-P might be an upstream member of MobA/SidB for septation as shown in the working model of Fig 6. Taken together, SepH and PomA were able to retroregulate the phosphorylation of HogA for septation and conidiation under non-osmotic stress stimulus conditions. PomA and SepH indirectly affect the expression of HogA. SepH, a main kinase of the SIN, regulates septation/ cytokinesis by phosphorylating downstream targets [2]. However, PomA is also a putative kinase and, similar to other serine/threonine protein kinases, the catalytic domain (S-TKc) is included. Thus, we hypothesize that PomA suppresses the defects of SepH indirectly by phosphorylating other proteins. In the absence of PbsB (MAPKK, a putative upstream member of HogA), HogA phosphorylation is completely abolished such that deletion of PomA or SepH is unable to rescue or change the phosphorylation level of HogA. Thus, we conclude that PomA and SepH indirectly regulate HogA phosphorylation through affecting HogA-P upstream kinases.

Strains, media, and culture conditions
All A. nidulans strains used in this study are summarized in S1 Table. For culture media, in general, A. nidulans strains were grown on rich media YAG or YUU (YAG supplemented with 5 mM uridine and 10 mM uracil) containing 2% glucose, 0.5% yeast extract, and 1 ml/liter 1,000× trace elements, minimal media PGR or PGRUU (PGR supplemented with 5 mM uridine and 10 mM uracil) or PGRT (PGR supplemented with 100 mM threonine) containing 50 ml/liter salt, 1% glycerol, 0.5 mg/liter pyridoxine, 2.5 mg/liter riboflavin and 1 ml/liter 1,000× trace elements and PDR (PGR replaces the carbon source glycerol with glucose) or PDRUU (PDR supplemented with 5 mM uridine and 10 mM uracil) [40,41]. For strains crossing of A. nidulans, most of procedure was described as reference [42]. In brief, two parent strains are inoculated for about 5 mm-apart on rich medium YUU for 2 days at 37˚C and then digging agars with mixed grown mycelia were transferred to the selective medium incubated for 15 days at room temperature. After that, a matured cleistothecia was picked up and punctured to a 1 ml sterilized H 2 O. Lastly, 10 μl suspension of ascospores was cultured to the rich medium YUU at 37˚C and all progenies were scored for phenotypic quantification. Growth conditions, DNA transformation procedures and induction conditions for alcA(p)-driven expression were performed as described previously [43,44].

Construct design and protein tagging
All primers used in this study are shown in S2 Table. To generate the pomA deletion cassette, the fusion PCR method was used as previously described. Briefly, approximately 1 kb of the upstream and downstream flanking sequences of the pomA gene were amplified using primers pomA-P1/P3 and pomA-P4/P6, respectively, and using the genomic DNA (gDNA) of TN02A7 as a template. As a selectable nutritional marker for fungal transformation, A. fumigatus pyrG was amplified from the plasmid pXDRFP4 using primers pyrG-F/R. Next, the three aforementioned PCR products were used as templates for amplification using primer pairs pomA-P2/P5 to generate the cassette and was then transformed into the recipient strain TN02A7 [45]. A similar strategy was used to construct the ankA, hogA and sidB deletion mutant. For the generation of sepH and pomA or ankA double mutants, sepH1 was crossed with ΔpomA or ΔankA, respectively.
To generate the pbsB deletion strain, 5'-and 3'-flanking region fragments of pbsB were amplified by using the genomic DNA of TN02A7 as a template and pbsB-p1/p2 and pbsB-p3/ p4 as primers, respectively. As a selectable nutritional marker for fungal transformation, the A. nidulans riboB gene was amplified from the genomic DNA of the parental wild type (R21) using primers Ribo-F/R and then cloned into pEASY-Blunt-zero vector (TransGen Biotech, China) to form the plasmid Blunt-ribo. After that, 5'-and 3'-flanking region fragments were ligated into the Pst1 and Not1 sites of the plasmid Blunt-ribo, respectively. The resulting plasmid was transformed into the strains of TN02A7, ΔpomA and sepH1, respectively, to construct the ΔpbsB, ΔpbsB ΔpomA and ΔpbsB sepH1 strains and then the expected progenies were obtained by diagnostic PCR.
To generate the RFP-labeled histone H2A strain (ZXA05), A. fumigatus pyrG, a selectable nutritional marker, and the RFP fragment (except for the termination codon) were amplified from the plasmid pXDRFP4 using primers pyrG-F/R and RFP-F/R, respectively. The H2A DNA fragment (including the termination codon) was amplified using primers H2A-F/R, and the gDNA of TN02A7 was used as a template. Next, the RFP fragment and the H2A DNA fragment were combined by using RFP-F/H2A-R for PCR to form the RFP-H2A fragment. The aforementioned pyrG marker and the RFP-H2A fragment were then cloned into the reconstituted vector, pBARGPE1. The AngpdA promoter was included sequentially and then the resulting plasmid was transformed into the recipient strain TN02A7 to obtain strain ZXA05. For the generation of RFP-labeled histone H2A in the relative mutant background strains, mutants sin11, sin53 and ΔpomA were crossed with the RFP-labeled strain ZXA05, respectively, and then progenies embedding RFP labeling in relative mutants were obtained by diagnostic PCR and microscopic examination.
To generate the alcA(p)::GFP-pomA strain, an approximate 0.9 kb fragment of the pomA gene was amplified from TN02A7 gDNA using primers alc-pomA-F/R. The aforementioned fragment was then cloned into the corresponding sites of pLB01 in which a GFP tag was tagged at the N-terminus of pomA. The resulting plasmid was transformed into the recipient strain TN02A7. A similar strategy was used to construct alc(p)::GFP-mobA strain. For the generation of GFP tag tagged MobA in strains of sepH1 or ΔpomA, alc(p)::GFP-mobA strain was crossed with sepH1 or ΔpomA, respectively.
To generate constructs for the pomA overexpression strain, a pomA DNA fragment was amplified by using cla1-pomA-F/R as primers and the gDNA of TN02A7 as a template. The PCR fragment was then cloned into the aforementioned AngpdA promoter-included vector pBARGPE1. The resulting plasmid was transformed into the recipient strain TN02A7.
To generate constructs for the site-directed mutant PomA L1265S , upstream and downstream flanking sequences were amplified from TN02A7 gDNA by using primers S-pomA-P1/P3 and S-pomA-P4/P6. The upstream flanking sequence with a site directed mutation (Leucine 1265 was replaced by Serine) was amplified following two steps. First, two fragments were amplified from TN02A7 gDNA with primers S-pomA-P1/S-R and S-F/S-pomA-P3, respectively. Second, the two aforementioned fragments were combined by fusion PCR with primers S-pomA-P1/ P3. Finally, a 4.7 kb DNA fragment was generated using primers S-pomA-P2/P5 and the up-and downstream flanking fragments and the pyrG marker were used as PCR templates, which were then cloned into the pEasy-Blunt Zero (TransGen Biotech, China) vector. The resulting plasmid was transformed into the recipient strain TN02A7 to obtain mutant PomA L1265S . For construction of the double mutant (sepH and PomA L1265S ), the sepH1 mutant was crossed with PomA L1265S and then the expected progenies were obtained by phenotypic screening and diagnostic PCR.
To generate the HogA phosphorylation sites deletion strains (HogA-ΔP), two DNA fragments were amplified from TN02A7 genomic DNA by using primers HogA-site-p1/HogA-Lde-down and HogA -site-p3/HogA -R-de-up and then these fragments were combined by using HogA-site-p1/HogA-site-p3 for PCR to form the upstream flanking sequence. The downstream flanking sequence was amplified using primers HogA-site-p4/HogA-site-p6 using the genomic DNA (gDNA) of TN02A7 as a template. Lastly, the upstream and downstream flanking sequences and the aforementioned pyrG marker were used as templates for amplification using primer pairs HogA-site-p2/HogA-site-p5 to generate the cassette and was then transformed into the recipient strain TN02A7. The strain HogA-ΔP was crossed with sepH1 to generate constructs of HogA phosphorylation sites deletion in background of sepH1.

Fluorescence microscopy observations
To visualize septa of the indicated strains, conidia were inoculated onto sterile glass coverslips at related temperatures prior to observation. At the detection time point, media on the coverslips were removed and germlings were washed three times with PBS. Cultured cells were then fixed with 4% paraformaldehyde (Polysciences, Warrington, PA, USA) and washed three times with PBS. CFW (Sigma-Aldrich, St. Louis, MO, USA) was used to stain hyphal septa in the dark for 5 min. CFW solution was then removed and the glass coverslip was washed three times with PBS. Images were captured with a Zeiss Axio imager A1 microscope (Zeiss, Jena, Germany) and managed with Adobe Photoshop (Adobe, San Jose, CA, USA).
For visualization of the localization of GFP-PomA and RFP-H2A, a similar strategy procedure was used as described in references.

RNA isolation and quantitative real-time PCR
To isolate RNA from relative strains, fresh conidia were inoculated on minimal medium in the dark for 24 h at 37˚C and then mycelia were immediately harvested and frozen in liquid nitrogen. Total RNA was extracted using TRIzol (Roche) as described in the manufacturer's instructions. For gDNA digestion and cDNA synthesis, the HiScriptII Q RT SuperMix for qPCR (+gDNA wiper) Kit (Vazyme) was used following the procedures of the protocol manual [46].

Protein extraction and Western blot analysis
Fresh conidia were inoculated into minimal medium PGRUU or PDRUU and incubated for 24 h in the dark at related temperatures and then mycelia were harvested and immediately frozen in liquid nitrogen. For protein extraction, the following lysis buffer recipe was used: 10 mM Tris-HCl, pH 7.5, 150 mM NaCl, 0.5 mM EDTA, 0.01% Triton X-100, 1 mM DTT, 1 mM PMSF, and 1:100 protease inhibitor cocktail. A BCA kit was used to quantify the amount of protein according to the manufacturer's instructions. Lastly, protein samples were loaded onto a 10% SDS polyacrylamide gel and transferred to a PVDF membrane (Immobilon-P, Millipore) in transfer buffer (384 mM glycine, 50 mM Tris, pH 8.4, 20% methanol) at 350 mA for 1.5 h [47]. Next, the membrane was blocked with PBS, 5% milk, and 0.1% Tween 20 and then probed with the following antibodies: anti-phospho-p38 MAPK (The180/Tyr182) antibody (4511T; Cell Signalling Technology; dilution 1:1000) against phosphorylated HogA and anti-p38 (9212S, Cell Signaling Technology; dilution 1:1000) antibody against total MAPK protein and anti-actin antibody (ICN Biomedicals, Inc., clone C4; dilution 1:10000).

Next-generation sequencing
Conidial spores from A. nidulans strains S11 and S53 were inoculated into liquid medium and shaken at 42˚C for 18 h at 220 rpm. After harvesting, dried mycelia were then extracted to obtain gDNA and analyzed by NGS, which was performed at Shanghai OE Biotechnology Co., Ltd. as a commercial service. The NGS sequencing raw data submitted to SRA (http://www. ncbi.nlm.nih.gov/sra) at NCBI with accession number SRS4286007.

Quantitative phosphoproteomics
For quantitative phosphoproteomics, fresh conidia were inoculated into liquid minimal medium PDRUU and incubated for 18 h at 42˚C and shaken at 220 rpm. The harvested mycelia were then dried and extracted to obtain total proteins. The quantitative phosphoproteomics experiment was performed at Hangzhou PTM Biolabs Co., Ltd. as a commercial service. In brief, lysis buffer (8 M urea, 1% Triton-100, 10 mM DTT, and 1% protease inhibitor cocktail) was used and the protein concentration was determined with a BCA kit according to the manufacturer's instructions. Proteins were digested with trypsin (Promega) and then labeled with a TMT kit (Thermo) according to the manufacturer's protocol. Next, the labeled tryptic peptides were fractionated by high pH reversed-phase HPLC using a Thermo Betasil C18 column (5 μm particles, 10 mm ID, 250 mm length). After fractioning, the enrichment of phosphorylation was carried out based on biomaterial IMAC microspheres (for proteomics of total proteins, this step was skipped). Next, peptide mixtures were first incubated with an IMAC microsphere suspension and vortexed in loading buffer (50% acetonitrile/6% trifluoroacetic acid) and centrifuged to collect pellets with enriched phosphopeptides in the IMAC microspheres. The IMAC microspheres were then washed with 50% acetonitrile/6% trifluoroacetic acid and 30% acetonitrile/0.1% trifluoroacetic acid to sequentially remove nonspecifically adsorbed peptides. Elution buffer containing 10% NH 4 OH was used to collect the supernatant containing phosphopeptides. For LC-MS/MS analysis, the tryptic peptides were dissolved in 0.1% formic acid (solvent A) and directly loaded onto a homemade reversed-phase analytical column (15-cm length, 75 μm i.d.).
Proteins with more than a 1.2-fold increase or decrease in phosphorylation were analyzed for GO terms and KEGG pathway analysis using the UniProt-GOA database (http://www.ebi. ac.uk/GOA/) and the KEGG database, respectively. The raw phosphoproteomic data have submitted to ProteomeXchange (http://www.ebi.ac.uk/pride/archive/) with accession number PXD012514.  Fig. Screening and isolation of the sepH1 suppressors of septation and conidiation. S11 (A) and S53 (B) were crossed with the wild type strain R21 and then the isolated progenies sin11 and sin53 were respectively crossed with sepH1 in minimal media PDR. All plate colonies used to examine the phenotypes of isolated progenies were cultured in rich media YUU at 30 or 42˚C for 2 days. (C) Hyphal cells stained with CFW for the wild type (R21) and isolated progeny sepH-1, sepH-3, sepH sin11-2 and sepH sin53-2 strains cultured in liquid media at 42 C for 20 h. Arrows indicate the locations of septa. Bars, 10 μm. (TIF) S3 Fig. Quantitative data for septation, colony size and conidia production in relative strains. (A) Quantitative data of the conidia production for the WT (TN02A7), sepH1, S11 and S53 strains cultured in rich media YUU at 42˚C for 2 days. (B) Quantitative data of septation for the WT (TN02A7), S11, S53, sepH1 ΔpomA, sepH1 pomA L1265S and sepH1 ΔankA strains cultured in liquid rich media YUU at 42˚C for 20h. (C) Quantitative data of colony size for the WT (TN02A7), sepH1, S11 and S53 strains and (D) WT (R21), sin11 and sin53 cultured in rich media YUU at 30˚C and 42˚C for 2 days.