The High Osmolarity Glycerol (HOG) pathway is a major fungal signaling cascade. Initially found to control the response to osmotic stress (Brewster et al., 1993), the HOG pathway was later on shown to process a variety of stress signals and to contribute to the virulence of several fungal pathogens (Ma and Li, 2013; Román et al., 2020).

Most information on the HOG pathway originates from work done with Saccharomyces cerevisiae (Hohmann, 2002). In the absence of osmotic stress, the membrane-bound sensing kinase Sln1p phosphorylates the histidine phosphotransfer (HPt) protein Ypd1p that transfers phosphoryl groups to the response regulator Ssk1p. If a stress signal is perceived, Sln1p becomes inactive and this results in a dephosphorylation of Ssk1p and, in turn, in an activation of Ssk2p and the HOG MAP kinase module (Posas and Saito, 1998; Horie et al., 2008; Fassler and West, 2013). This organization explains the lethal consequences of a SLN1 or YPD1 deletion, since a permanent inactivation or the absence of either protein entails a persistent and lethal activation of Hog1p (Maeda et al., 1994). This concept is corroborated by the finding that YPD1 deletions in S. cerevisiae or Cryptococcus neoformans are not lethal in strains, in which the HOG pathway is inactivated by an additional mutation (Posas et al., 1996; Lee et al., 2011). The fact that some Ypd1p orthologs are not essential (Nguyen et al., 2000; Lee et al., 2011; Mavrianos et al., 2014; Jacob et al., 2015; Rodríguez-González et al., 2017) points to differences in the Ypd1-dependent signaling processes in different fungi.

Ypd1 is part of a multistep phosphorelay that also comprises one or more hybrid histidine kinases (HHKs) and two response regulators (RRs) (Defosse et al., 2015) (compare Figure 7). The perceived signal is relayed by a series of phosphorylation events: starting with an autophosphorylation of a His residue in the His Kinase acceptor (HisKA) domain, the phosphoryl group is then transferred intramolecularly to an Asp residue in the receiver domain of the HHK. In the next step, the phosphoryl group is transferred to a conserved His residue in the HPt protein Ypd1 and Ypd1 finally phosphorylates an Asp residue in the receiver domain of the RRs. This relay was first described for S. cerevisiae (Maeda et al., 1994; Posas et al., 1996) and represents the blueprint for the phosphorelays of other fungi.

HHKs represent a fungal-specific form of two-component systems (TCS) that represent sensor molecules that allow microorganisms to sense and respond to a plethora of environmental signals and stress conditions. Bacterial TCS consist of a membrane-bound histidine kinase and an intracellular response regulator protein, whereas fungal HHKs integrate both TCS functions in one protein (Bahn, 2008).

Most filamentous fungi harbor several HHKs and Catlett et al. (2003) defined 11 different groups. All HHKs share a common C-terminal signaling module, but each group is characterized by a distinct N-terminal sensing module that is specialized in the measurement of certain environmental signals. Remarkably, all fungi analyzed so far harbor one HPt protein and two RRs and the paradigms of these proteins were identified and characterized in S. cerevisiae (Brown et al., 1994; Li et al., 1998).

Aspergillus fumigatus is a major air-borne fungal pathogen that causes severe and often life-threatening infections in immunocompromised patients (McCormick et al., 2010) and patients suffering from severe viral infections, such as those caused by influenza or SARS-CoV-2 (Dewi et al., 2021). Although it lacks sophisticated virulence traits, A. fumigatus is a successful pathogen and this is largely due to its stress resistance and ability to rapidly adapt to a wide range of environmental conditions (Tekaia and Latgé, 2005; McCormick et al., 2010). This versatility is to a large extend attributable to a set of 13 HHKs that belong to eight different groups of HHKs (Defosse et al., 2015). One of them, the group III HHK TcsC, controls the activity of the HOG pathway (McCormick et al., 2012; Hagiwara et al., 2013) and thereby resembles Sln1p at the functional level. A remarkable difference exists with regards to the viability of respective deletion mutants: The group VI HHK Sln1p of S. cerevisiae is essential (Maeda et al., 1994), whereas TcsC and group III HHKs of other filamentous fungi are not (Defosse et al., 2015). The pivotal role of Sln1p reflects the fact that it is the sole HHK in baker's yeast and essentially required to prevent an uncontrolled and lethal activation of the HOG pathway. The existence of multiple HHKs in most other fungi seems to provide a level of functional redundancy that enables group III HHK mutants to thrive under standard growth conditions.

The sensing module of group III HHKs consists of several HAMP domains that control the activity of the C-terminal signaling or kinase module, which after autophosphorylation may transfer a phosphoryl group to the HPt protein Ypd1. Expression of a truncated form of TcsC lacking its sensing module causes a dysregulation of the osmotic homeostasis in A. fumigatus and results in a fatal influx of water (Spadinger and Ebel, 2017), which emphasizes the necessity for a tight control of the activity of the TcsC kinase module.

In A. nidulans, a model organism for filamentous fungi, the HOG pathway controls the response to several stress conditions (Vargas-Pérez et al., 2007; Hagiwara et al., 2009). Mutants in genes encoding the group III HHK NikA, which is assumed to phosphorylates YpdA, are viable (Hagiwara et al., 2007; Vargas-Pérez et al., 2007), but attempts to generate an ypdA deletion mutant failed, suggesting that YpdA is an essential protein in A. nidulans (Furukawa et al., 2005; Vargas-Pérez et al., 2007; Yoshimi et al., 2021) demonstrated very recently that downregulation of ypdA expression results in growth inhibition and aberrant hyphal morphology. Although related, A. nidulans and A. fumigatus show differences in the phenotypes of their respective group III HHK mutants (Böhmer et al., 2020) that may derive from differences in the biological activities of their HPt proteins.

To characterize the phosphorelay in A. fumigatus, we have analyzed its individual components using deletion mutants, point mutations and GFP fusion proteins. Our data demonstrate that Ypd1 is an essential protein in A. fumigatus and that antifungal agents, such as fludioxonil, kill A. fumigatus by a TcsC-mediated inactivation of Ypd1. Further evidence suggests that the activity of A. fumigatus Ypd1 is controlled by at least two different HHKs.

The HOG pathway enables fungi to adapt to a hyperosmotic environment and to other stress conditions. This pathway and in particular its group III HHK is, moreover, an attractive target for novel therapeutic agents, for which the bacterial compound pyrrolnitrin and its derivative fludioxonil are lead substances. A. fumigatus is a major fungal pathogen and its group III HHK TcsC is both, the main sensor protein of the HOG pathway and a key factor in the antifungal activity of fludioxonil and pyrrolnitrin (McCormick et al., 2012; Hagiwara et al., 2013).

The HOG pathway was described in detail in the model organism S. cerevisiae. A major difference between baker's yeast and A. fumigatus exists at the level of the HOG sensor kinase: S. cerevisiae Sln1p is a group VI and TcsC a group III HHK. Loss of Sln1p is lethal (Maeda et al., 1994), whereas a ΔtcsC mutant can grow normally under ambient conditions (McCormick et al., 2012; Hagiwara et al., 2013). Another distinctive feature is that S. cerevisiae lacks group III HHKs. TcsB, the A. fumigatus ortholog of Sln1p, shares 38.76% identical residues and all functional domains with Sln1p: an N-terminal section containing two transmembrane regions, a histidine kinase domain and a receiver domain (Ota and Varshavsky, 1993). In contrast to Sln1p, TcsB is not essential for A. fumigatus and a tcsB deletion mutant had no discernible phenotype apart from a slightly enhanced sensitivity to SDS (Du et al., 2006).

The general organization of the HOG pathway seems to be well-conserved in the fungal kingdom. In A. fumigatus, the initial multistep phosphorelay comprises of TcsC, the HPt protein Ypd1 and the two RRs SskA and Skn7. The basic concept of the phosphorelay is that the HPt protein phosphorylates and thereby inactivates the two RRs under ambient conditions, although S. cerevisiae Skn7p was shown to be activated, rather than inactivated by phosphorylation (Ketela et al., 1998; Li et al., 1998).

In this study, we have initially analyzed the localization of Ypd1, SskA and Skn7. SskA, which controls the activity of a downstream MAPK cascade, was found to reside in the cytosol, whereas Skn7 was highly enriched in the nuclei. This complementary pattern was not affected by fludioxonil and resembles that previously reported for S. cerevisiae Ssk1p and Skn7p in the presence and absence of hyperosmotic stress (Lu et al., 2003). GFP-Ypd1 was detected both, in the nucleus and the cytoplasm and thereby resembles S. cerevisiae Ypd1p (Lu et al., 2003). Treatment with fludioxonil did not result in a dramatic change of this nucleocytoplasmic pattern. The presence of its two target molecules in the cytosol and the nucleus requires Ypd1 to shuttle between both cellular compartments. After bleaching of the cytosolic pool of GFP-Ypd1, we indeed found that the resulting nuclear enrichment of the GFP signals declined over time and disappeared after 10–15 min. This indicates that GFP-Ypd1 is able to shuttle between the nucleus and the cytoplasm.

Evidence from the related fungus A. nidulans suggested that Ypd1 could be an essential protein. We took this into account and generated ypd1^tet−on mutants in which the native ypd1 promotor was replaced by a tet-on construct. Under non-inducing conditions (in the absence of doxycyclin), the wild type (AfS35) ypd1^tet−on strain was hardly able to form colonies and its hyphae were swollen and often showed signs of cellular lysis. This suggested that Ypd1 is an essential protein in A. fumigatus.

To characterize the role of other proteins of the multistep phosphorelay, we established ypd1^tet−on derivatives of several deletion mutants. Strains with a tcsC or skn7 deletion showed a similar phenotype as AfS35 ypd1^tet−on demonstrating that neither TcsC nor Skn7 are essentially required for the deleterious process that results from a reduced ypd1 expression level. In C. neoformans, ypd1 is an essential gene, but a deletion strain can be obtained in a strain lacking hog1 to prevent a lethal activation of the HOG pathway (Lee et al., 2011). In this study, we found that an ypd1^tet−on mutant is viable in a strain that also lacks sskA. This suggests that the low levels of Ypd1, which are present in an ypd1^tet−on strain under non-inducing conditions, can control Skn7 but not SskA. The phenotype associated with a depletion of Ypd1 comprises a dramatic swelling of the fungal cells, an enhanced septation, an elevated number of nuclei per compartment as well as an increased chitin content in the cell wall, all features previously described for fludioxonil-treated A. fumigatus wild type hyphae (McCormick et al., 2012; Wiedemann et al., 2016). Very recently, Yoshimi et al. (2021) also reported that a downregulation of ypdA expression results in growth inhibition and aberrant hyphal morphology in A. nidulans.

We have recently shown that the antifungal activity of fludioxonil is largely dependent on Skn7 and only to a minor extend on the SskA - SakA axis (Schruefer et al., 2021). The ΔsskA ypd1^tet−on double and the ΔsskAΔskn7 ypd1^tet−on triple mutant both grow well without doxy, but only the former is sensitive to fludioxonil, which confirms the important role of Skn7 in this context. The fact that this lethal effect can be overcome by osmoprotection may be explained by the moderate increase of the internal glycerol concentration previously found in a ΔsskA strain (Schruefer et al., 2021).

Similar to its counterpart in S. cerevisiae, A. fumigatus Ypd1 is apparently required to control the activity of SskA and Skn7 and the fact that the conserved His residue at position 89 is important for this biological activity suggests that Ypd1 transfers phosphoryl groups to the RRs and most likely to conserved aspartate residues in their receiver domains. For this activity, Ypd1 itself has to be constantly phosphorylated under ambient conditions. In S. cerevisiae, this is done by Sln1p, which consequently is an essential protein. In A. fumigatus, TcsC is dispensable under ambient conditions and has therefore a clearly different function than Sln1p. Since individual knock-out strains in all 13 HHK genes of A. fumigatus are viable (Chapeland-Leclerc et al., 2015), it appears likely that Ypd1 receives phosphoryl groups from more than one HHK; whether TcsC is involved remains an open question.

The data obtained with the ypd1^tet−on construct already suggested that Ypd1 is essential in A. fumigatus. The viability of the ΔsskA ypd1^tet−on mutant either suggests that an ypd1 deletion is tolerated in this genetic background or that the low amounts of Ypd1 in this mutant are sufficient to enable its growth. To address this point, we tried to establish an ypd1 deletion of the ΔsskA strain. This attempt failed in four independent experiments, which strongly suggests that, in a ΔsskA background, a certain amount of Ypd1 is required to control the Skn7 activity. This concept is corroborated be the fact that we easily obtained ypd1 deletion mutants in a ΔsakAΔskn7 background.

While the core HOG pathway is well conserved in the fungal kingdom, it is still unclear which HHK(s) control Ypd1 (Day and Quinn, 2019). An activation of the HOG pathway, either by fludioxonil or environmental stress, requires TcsC. The fact that fludioxonil treatment and an Ypd1 depletion result in similar phenotypes strongly suggests that TcsC activation entails an inactivation of Ypd1 and consequently a dephosphorylation of the two RRs. This implies that the RRs are active in their de-phosphorylated form as indicated in the model depicted in Figure 7. This concept is well established for Ssk1/SskA, but not for Skn7. Data from S. cerevisiae rather indicate a positive regulation of the Skn7 activity by a phosphorylation of the conserved Asp residue at position 427 (Ketela et al., 1998; Li et al., 1998). Further experiments are clearly required to address this point properly.

Our data show that the His₈₉ residue in Ypd1 is essentially required to keep the HOG pathway in an inactive state. If TcsC is indeed a histidine kinase, its activation must result in a different phosphorylation, e.g., of an alternative His residue in Ypd1. This could result in a conformational change that prevents a phosphorylation at His₈₉ or it may lead to a different specificity of Ypd1 resulting in a phosphorylation of Skn7 instead of SskA. The Ypd1 of A. fumigatus harbors three His residues at positions 89, 106 and 112. His₁₀₆ is not well conserved in the Ypd1 proteins of different Aspergilli, leaving only His₁₁₂ as a potential alternative target site. Another possible way to inactivate Ypd1 is de-phosphorylation. Recent data published for the group III HHK Drk1 of Blastomyces dermatitidis indicate that this protein possesses a histidine kinase activity under ambient conditions, but can adopt a phosphatase activity under fludioxonil treatment (Lawry et al., 2017).

A role of TcsC in the activation of the HOG pathway is evident, but its target molecules and its precise mode of action remain elusive. Whether TcsC has a functional role under steady state conditions is also unclear, but the distinct morphology of the ΔtcsC colonies (McCormick et al., 2012) suggests this. Skn7 is a crucial player in the antifungal activities trigged by fludioxonil (Schruefer et al., 2021). TcsC modulates the activity of Skn7 most likely via Ypd1. The similar phenotypes of fludioxonil-treated and Ypd1-depleted hyphae suggest that both processes are characterized by an inactivation of Ypd1. A switch between a kinase and a phosphatase activity as described for B. dermatitis Drk1 could explain how TcsC inactivates Ypd1. In the absence of stress, Ypd1, is most likely controlled by two or multiple HHKs (Figure 7); the identification of these HHKs and the analysis of their individual contribution will be an important task for future studies. Hence, even after more than 25 years of research on the HOG pathway (Brewster and Gustin, 2014), many questions are still awaiting an answer.

The original contributions presented in the study are included in the article/Supplementary Material, further inquiries can be directed to the corresponding author/s.

SS performed most experiments. AS performed the localization experiments with GFP-Ypd1. CK performed the localization experiments with GFP-Skn7 and SskA-GFP. LS constructed the ypd1 tet-on construct. SS and FE designed the experiments and wrote the manuscript. All authors contributed to the article and approved the submitted version.

This work was funded by a grant of the Deutsche Forschungsgemeinschaft (EB-184-1) to FE.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.