The N. crassa FGSC#2489 wild-type strain and the FGSC#9718 (mat a mus-51:bar), FGSC#9568 (mat a mus-52:hph), FGSC#18393 (a NCU08772:hyg), FGSC#18394 (A NCU08772:hyg), FGSC#6103 (A his-3), FGSC#18932 (NCU06687:hyg), and FGSC#9518 (A, his-3⁺:Pccg-1-hH1⁺-sfgfp⁺) mutant strains were purchased from the Fungal Genetics Stock Center (FGSC, University of Missouri, Kansas City, MO, USA)¹ (McCluskey, 2003). The strain Δpcl-1 hH1-sfgfp was generated by crossing the FGSC#18393 and #9518 strains in 0.1 × SC Westergard medium (Westergaard and Mitchell, 1947) containing 0.5% sucrose. Gene knockout in all mutant strains was confirmed by PCR using specific oligonucleotides (Supplementary Table 1) by comparing to the amplification in the genomic DNA of the wild-type strain. The strains were maintained on solid Vogel’s minimal (VM) medium, pH 5.8 (Vogel, 1956) containing 2% sucrose at 30°C. Conidia from 10-day old cultures were suspended in sterile water, counted, and used in the experiments.

For stress analyses in plates, 5 μl of a 2 × 10⁷ conidia per mL suspension from wild-type, Δpcl and Δpcl-1 pcl-1⁺ (complemented) strains were inoculated on Petri dishes in solid VM medium containing 2% sucrose and either NaCl or sorbitol for osmotic stress assays, or H₂O₂ for oxidative stress assays, or Congo Red (CR) for plasma membrane/cell wall stress assay, or calcium chloride for calcium sensitivity assay, or at pH 7.8 for alkaline stress assay. Cultures were incubated at 30°C, and the colony diameter was measured after 24 h.

For complementation, the FGSC#18393 strain (Δpcl-1:hyg) was crossed with the FGSC#6103 strain (his-3) to generate the Δpcl-1 his-3 double mutant strain. A DNA fragment of 1,156 bp was amplified by PCR with the primers 8772sfGFP-F and 8772sfGFP-R (Supplementary Table 1) using genomic DNA from the wild-type strain as template. The 8772sfGFP-R oligonucleotide contains the sequence that codifies for 6-Gly between the nucleotide sequences encoding sfGFP and the PCL-1 proteins. PCR was performed using a Phusion High-Fidelity PCR kit (Finzymes), and the DNA fragment was purified with a QIAquick Gel Extraction Kit (Qiagen, Hilden, Germany) according to the manufacturer’s instructions. The purified DNA fragment was cloned into the PacI and XbaI sites of the pTSL91-A plasmid (a donation from N. L. Glass, University of California at Berkeley, Berkeley, CA, USA), generating the pTSL91A-pcl-1 plasmid. This plasmid allows the constitutive expression of the C-terminus PCL-1-sfGFP fusion protein, as the ccg-1 promoter drives the pcl-1 gene expression. To express the protein under the control of native promoter, the ccg-1 promoter was removed with the NotI and XbaI enzymes and replaced by the pcl-1 promoter (1,500 bp) amplified by PCR using the np8772-F and np8772-R oligonucleotides (Supplementary Table 1), generating the construction pTSL91A-Ppcl-1-pcl1-sfgfp. Plasmid constructions were used to transform by electroporation competent conidia from the recipient Δpcl-1 his-3 double mutant strain. The transformants were selected on VM media containing hygromicin (200 μg per mL) without histidine and confirmed by PCR using the primers np8772-F and 8772sfGFP-R (Supplementary Table 1). The progeny was analyzed by fluorescence microscopy, and the complemented strain (Δpcl-1 pcl-1⁺) was evaluated in different experiments.

The analyses were performed in submerged cultures, and 2 × 10⁷ conidia per mL (final concentration) from WT, Δpcl and Δpcl-1 pcl-1⁺ strains were inoculated into flasks containing 30 mL of liquid VM medium plus 2% sucrose, pH 5.8, 30°C, 150 rpm. Aliquots were removed every 2 h, and the conidia in germination were counted. Germination was also analyzed in conidia inoculated onto coverslips covered with liquid VM plus 2% sucrose, pH 5.8 and incubated at 30°C for different times. For nuclei analysis, cells were fixed [1% phosphate buffered saline (PBS), 3.7% formaldehyde, 0.1% Triton X-100], washed twice with PBS and stained with 100 μL DAPI (4′,6-diamidino-2-phenylindole, 0.5 mg per mL in DMSO) for 5 min. DAPI fluorescence was visualized using a fluorescence microscope with excitation and emission wavelengths of 358 nm and 463 nm, respectively. Images of conidia germination were captured using an AXIO Imager.A2 Zeiss microscope, at a magnification of 1000 X, coupled to an AxioCam MRm camera and processed with the using the AxioVision software.

The cell cycle progression of the Δpcl-1 hH1-sfgfp strain was analyzed by inoculating 2 × 10⁷ conidia per mL (final concentration) into flasks containing 30 mL of liquid VM medium plus 2% sucrose, pH 5.8, 30°C, 150 rpm. The FGSC#9518 (A, his-3⁺:Pccg-1-hH1⁺-sfgfp⁺) strain was used as control. Aliquots were removed every 2 h, and the GFP fluorescence was analyzed in the AXIO Imager.A2 Zeiss fluorescence microscope coupled to an AxioCam MRm camera and processed using the AxioVision 4.8.2 software with excitation and emission wavelengths of 490 and 525 nm, respectively. Images were captured at different times, and six images from each strain were used for quantification of the nuclei in the different phases of mitosis (approximately 200) (Hong et al., 2014).

Glycogen and glycogen synthase activity were quantified in mycelia pads grown by inoculating 10⁷ conidia per mL into 500 mL of liquid VM, 250 rpm at 30°C. Mycelia samples were collected every 12 h and frozen in liquid nitrogen. Mycelia pads were extracted in lysis buffer (50 mM Tris-HCl, pH 8.0, 50 mM NaF, 1 mM EDTA, 0.5 mM PMSF, 0.1 mM TCLK, 25 mM benzamidine, and 1 μg per mL of each pepstatin and aprotinin). Cellular extracts were clarified by centrifugation at 3,000 X g, 10 min, 4°C, and the supernatants were used for glycogen, glycogen synthase activity, and protein quantification. Glycogen quantification was performed after precipitation with cold ethanol and digestion with α-amylase and amyloglucosidase (Freitas et al., 2010). Free glucose was quantified with a glucose oxidase kit (Labtest, Brazil), and glycogen concentration was normalized to the total protein concentration. Glycogen synthase activity was measured in the presence and in the absence of G6P (glucose-6-phosphate, 7.2 mM final concentration) and normalized to the total protein concentration (Thomas et al., 1968). Total protein was quantified using BSA as standard (Hartree, 1972).

Glycogen synthase phosphorylation was analyzed by two-dimensional electrophoresis in mycelial pads from the wild-type and Δpcl-1 strains grown for 24 h (Candido et al., 2014). Briefly, total protein was fractionated by IEF using the pH 4–7 gradient Immobiline™ DryStrip (13 cm, linear, GE Healthcare, Chicago, USA) in an Ettan IPGphor 3 system (GE Healthcare, Chicago, USA), according to the manufacturer’s instructions. IEF was carried out at 50 μA per strip at 20°C, using the following steps: 100 V (10 h), 500 V (500 Vh), 1,000 V (750 Vh), 8,000 V (11,325 Vh), and 8,000 V (5,067 Vh). After electrofocusing, the strips were equilibrated and subjected to a 9% SDS-PAGE, using a Hoefer™ SE600 system (GE Healthcare, Chicago, USA). To analyze the GSN phosphorylation status, proteins were transferred to nitrocellulose membrane after 2-DE and blotted with anti-GSN antibody raised in rabbit. Blots were subsequently probed with HRP-conjugated secondary antibodies (Sigma) and developed with luminol reagent. As a control of phosphorylation, total protein from the wild-type strain was treated with λ-protein phosphatase according to the suppliers’ protocol (#P0753S, New England Biolabs, Ipswich, USA).

The pcl-1 cDNA (ORF NCU08772, 1,110 bp) was amplified by RT-PCR with the oligonucleotides Nc8772-F/Nc8772-R (Supplementary Table 1) using cDNA synthesized from total RNA and the SuperScript III reverse transcriptase kit (Invitrogen, Waltham, USA). The DNA fragment was cloned into NdeI and BamHI sites of pET28a expression vector leading to the pET-pcl-1 construction for the expression of a 369 amino acids protein fused to His-tag at N-terminus. The pho85-1 cDNA (ORF NCU07580, 1,014 bp) was amplified by RT-PCR with the oligonucleotides Nc7580-F/Nc7580-R (Supplementary Table 1) and cloned into NdeI and BamHI sites of pET28a expression vector leading to the pET-pho85-1 construction for the expression of a 337 amino acids protein fused to His-tag at N-terminus. The full gsn cDNA was amplified from the pET28a-gsn plasmid (Freitas et al., 2010) with the GSN-BamHI-F and GSN-EcoRI-R oligonucleotides and subcloned into BamHI and EcoRI sites of a modified pET28a expression vector, leading to a His-tagged SUMO fusion protein at N-terminus. Site-directed mutagenesis at serine/threonine amino acids in SUMO-GSN was performed using the QuickChange mutagenesis kit (Stratagene, San Diego, CA, USA). The S632A, S636A, T641A, and T645A single mutations were introduced with the S632A-F/R, S636A-F/R, T641A-F/R, and T645A-F/R oligonucleotides pairs, respectively. The quadruple mutant protein S632A/S636A/T641A/T645A was constructed by previously constructing the double mutant S632/636A using the S636ADM-F/R oligonucleotides and the S632A mutant cDNA as model. The double mutant cDNA was used to construct the triple mutant S632/636AT641A using the T641A-F/R oligonucleotides, which was used to construct the quadruple mutant using the T645ADM-F/R oligonucleotides. All plasmid constructions were confirmed by DNA sequencing. Recombinant proteins were produced according to Campanella et al. (2021) with slight modifications. The His-SUMO protein was also produced under the same induction conditions and used as control in the in vitro protein phosphorylation experiments.

For the phosphorylation assays, 25 μg of the wild-type and mutants His-SUMO-GSN proteins were incubated with 10 μg of His-PHO85-1 and 10 μg of His-PCL-1 proteins in reacting buffer (1 mM ATP, 25 mM MgCl₂, [γ-³²P]-ATP, activity 330 mCi/mmol, ∼1200 cpm/mmol) in 25 μL of reaction volume for 30 min at 30°C. Crude cellular extract of a N. crassa mutant strain (Δgsn), which does not synthesize glycogen synthase, was also used as a source of protein kinases. After incubation, His-SUMO-GSN proteins were immobilized on Ni-NTA agarose beads (Qiagen, Hilden, Germany) under low agitation. The beads were collected by centrifugation, washed three times in washing buffer (50 mM Tris, pH 8.0, 30 mM imidazole, 500 mM NaCl) and suspended in elution buffer (50 mM Tris, pH 8.0, 500 mM imidazole, 500 mM NaCl). After protein elution, Laemmli buffer (Laemmli, 1970) was added, the protein samples were boiled for 5 min and separated on a 12% SDS-PAGE gel. After electrophoresis, the gel was stained with Coomassie Brilliant Blue (CBB), dried and exposed to X-ray film.

For protein expression and cellular localization, strains expressing at C-terminus either V5- or GFP-tagged CRZ-1 proteins were constructed using the Honda and Selker (2009) methodology, as previously described (Campanella et al., 2021). The pZERO-hph-V5 and pZERO-hph-gfp plasmids (a donation from M. Freitag, Oregon State University, Corvallis, OR, USA) were used as templates and the splitmarkers were amplified by PCR using the CrzGly-F/R, and CrzLox-F/R oligonucleotide pairs, listed in Supplementary Table 1. The hph-V5 and hph-gfp nucleotide sequences were amplified from pZERO plasmids using the 10xGly-F/LoxP-R pair (Supplementary Table 1). The PCR fragments were fused to the hph-V5 or hph-gfp fragments by PCR using the CrzGly-F/Hph-R, Hph-F/CrzLox-R oligonucleotide pairs (Supplementary Table 1). Fusion DNA fragments were individually transformed into #9718 strain, and transformants were selected in the presence of hygromycin. Homokaryons were isolated by crossing with either the FGSC#2489 or FGSC#18394 strains in 0.5% Westergaard’s medium leading to the crz-1:crz-1-V5 and crz-1:crz-1-gfp in the wild type and in the Δpcl-1 background strains. The segregants from each strain were confirmed by PCR using the Nc8772-F/R, CrzGly-F, V5-R, and GFP-R primers (Supplementary Table 1). N. crassa transformation and other molecular techniques were performed using the protocols available at the Neurospora homepage.²

For gene expression analysis, conidia from the WT and Δpcl-1 strains were grown in VM media containing 2% sucrose for 24 h with and without 300 mM CaCl₂. The mycelia were harvested by filtration, washed with sterile water, and submitted to total RNA extraction (Sokolovsky et al., 1990). Total RNA (10 μg) samples were first treated with RQ1 RNase-free DNase (Promega, Madison, WI, USA) and subjected to cDNA synthesis by using SuperScript III First-Strand Synthesis kit (Invitrogen, Waltham, USA) and an oligo (dT) primer, according to manufacturer’s instructions. The cDNA libraries were subjected to RT-qPCR on a StepOnePlus™ Real Time PCR System (Applied Biosystems, Waltham, USA) using the Power SYBR^® Green PCR Master Mix (Applied Biosystems, Waltham, USA) and specific primers for each gene amplicon (Supplementary Table 1) as previously described (Virgilio et al., 2017). Data analysis was performed by the StepOne Software (Applied Biosystems, Waltham, USA), the comparative CT (ΔΔCT) method (Livak and Schmittgen, 2001) was used to estimate the Log2 fold expression. At least three biological replicates, with three experimental replicates each were performed, and reactions with no template were used as negative control. The PCR products were subjected to melting curves analysis to verify the presence of a single amplicon. All reaction efficiencies varied from 94 to 100% and the results were expressed relative to the expression of the beta-tubulin reference gene (β-tub-2 gene, NCU04054).

For protein expression, conidia from crz-1:crz-1-V5 and Δpcl-1 crz-1:crz-1-V5 strains were grown for 24 h in VM containing 2% sucrose at 30°C and 180 rpm. After 24 h, the mycelia were harvested by filtration and divided in samples. One sample was frozen in liquid nitrogen and stored at −80°C until use (control sample). The remaining samples were individually transferred into 150 ml of fresh VM liquid medium containing either 0.5% sucrose and 10 or 300 mM CaCl₂ for up to 60 min at 30°C and 180 rpm. The stressed mycelia were harvested, washed with sterile water and frozen in liquid nitrogen. Protein extraction was performed according to Campanella et al. (2021). For analysis of phosphorylated proteins 1 mM NaF, 1 mM sodium orthovanadate, and 1x EDTA-Free Phosphatase Inhibitor Cocktail (Sigma) was added to the lysis buffer. For protein phosphatase treatment, samples prepared in lysis buffer without phosphatase inhibitor were added of 2 mM MnCl₂ and treated with 400 U of λ-protein phosphatase (#P0753S, New England Biolabs, Ipswich, USA) for 40 min at 30°C followed by boiling for 5 min in 1 X Laemmli sample buffer (Laemmli, 1970). Separation of CRZ-1-V5 was accomplished in 4–15% Mini-PROTEAN^® TGX™ Precast Gels using 30 μg of total protein (Hartree, 1972). The proteins were electro-transferred to nitrocellulose blotting membrane (BioRad, Hercules, USA) and probed with a monoclonal anti-V5 antibody (Invitrogen, Waltham, USA). Blots were subsequently probed with HRP-conjugated secondary antibodies (Invitrogen, Waltham, USA) and developed with ECL reagent (BioRad, Hercules, USA). Images were acquired using the ChemiDoc MP Imaging System (BioRad, Hercules, USA).

To determine the cellular localization of the CRZ-1-GFP protein under calcium stress, 200 μL of a conidial suspension (2 × 10⁶ conidia per mL) from the crz-1:crz-1-gfp and Δpcl-1 crz-1:crz-1-gfp strains were inoculated onto coverslips, covered with VM media plus 2% sucrose and incubated at 30°C for 8 h (control sample). After incubation, the coverslips were transferred to fresh VM media containing 0.5% sucrose and 300 mM CaCl₂, and incubated for 30 min. The PCL-1 localization during germination was performed by inoculating 2 × 10⁷ conidia per mL (final concentration) of Δpcl-1 pcl-1⁺ complemented strain, which express the sfGFP-tagged PCL-1 under the native promoter. Coverslips were removed and protein location was analyzed every 2 h. For nuclei analysis, all samples were immediately fixed in 3.7% formaldehyde, 1% phosphate buffered saline (PBS), pH 7.0, 0.2% (v/v) Tween 80, washed twice with PBS and stained with Hoechst (10 μg per mL) for 15 min, followed by washing with PBS. Confocal Laser Scanning Microscopy (CLSM) was performed using a CARLS ZEISS LSM 800 Confocal Microscopy with a Plan Apochromat 63x/1.40 Oil DIC M27 objective. The detection parameters for all experiments were fixed at (i) nuclei analysis–hoechst—laser 405 nm: 35% 2.21 AU/79 μm, detection wavelength 400–450 nm, detection gain 894 V; (ii) CRZ-1-GFP–laser 488 nm: 35% 1.89 AU/84 μm detection wavelength 488–574 nm, detection gain 850V. The z-stack increments were 0.3 μm. The images were analyzed using the ZEN Blue 2.3.

PCL-1 and PHO-85-1 sequences were submitted to AlphaFold³ to obtain their respective predicted models separately and also in complex, using AlphaFold Multimer.⁴ The atomic coordinates of ATP and Mg²⁺ were retrieved from the crystallographic structure of the S. cerevisiae Pcl10/PHO85 complex available in Protein Data Bank (PDB ID: 4KRC)⁵ after superposition of the complexes. Then, the PCL-1/PHO85-1 complex was submitted to molecular dynamics (MD) simulations using GROMACS v.2020.4 (Abraham et al., 2015) under the CHARMM36m force field (Huang et al., 2017). The protonation state of the complex was determined according to the PROPKA3 web server (Olsson et al., 2011), setting the residue His290 of PCL-1 as positively charged. The complex was placed in a rhombic dodecahedric box of 12 Å distant from the farthest atom in XYZ directions. The system was then solvated and equilibrated with 0.15 M of NaCl, adding two Cl^– to obtain the net charge of zero. Following, the system was minimized using the Steepest Descent algorithm until reaching an energy gradient below 100 kJ/mol/nm². An NVT ensemble was applied generating the initial velocities randomly following a Maxwell–Boltzman distribution at 300 K during 1 ns, using the V-Rescale thermostat (Bussi et al., 2007) with a time constant of 0.1 ps. An NPT ensemble of 1 ns was then applied, defining 1 bar as reference pressure, using the Berendsen barostat (Eslami et al., 2010) with a time constant of 1.0 ps. These steps were performed restraining the backbone atoms of both proteins, ATP heavy atoms and Mg²⁺ under a force constant 1.000 kJ/mol/nm². An unconstrained MD of 300 ns was applied using the Nose–Hoover thermostat (Hoover, 1985) and Parrinello–Rahman barostat (Parrinello and Rahman, 1981), using a time constant of 0.5 and 5.0 ps, respectively. Non-bonded interactions were calculated considering atoms within 10 Å using the PME method, with a switching force function between 10 and 12 Å. Three independent replicas were performed collecting frames every 20 ps. After processing the PBC conditions of all frames, RMSD and RMSF calculations were performed for backbone atoms of the complex, using GROMACS built-in tools. Further, the prevalence of contacts for each residue of PCL-1 or PHO85-1 was measured considering heavy atoms of pairs of residues within a 3.5 Å cutoff, using a homemade tcl script implemented in VMD (Humphrey et al., 1996).

In a previous study, we identified the ortholog of the S. cerevisiae Pho85p cyclin-dependent protein kinase, the ORF NCU07580 product, as a regulatory protein of the glycogen metabolism (Candido et al., 2014). In S. cerevisiae, the Pho85p protein kinase together with Pcl6p, Pcl7p, Pcl8p, and Pcl10p (for Pho85p cyclins) cyclins has been described as a protein that controls glycogen metabolism by controlling glycogen synthase (Gsy2p) phosphorylation (Wilson et al., 1999; Wang et al., 2001b). To identify the partner cyclin of PHO85-1 involved in the glycogen metabolism control in N. crassa, a BLASTp search in the N. crassa genome database⁶ using as queries the S. cerevisiae cyclins, retrieved a homolog protein encoded by the ORF NCU08772, which shares an average of 24% identity to the yeast cyclins mainly in the cyclin domain (Figure 1A). The predicted protein is annotated in the fungus database as nuclear division-60, and is named here as PCL-1 consistent with the N. crassa nomenclature (Perkins et al., 1982) and considering that it is the first PCL described in N. crassa. PCL-1 is a non-essential protein, and the Δpcl-1 mutant strain exhibits wild-type phenotypes suggesting that it is not involved in processes related to growth and development (results not shown).

PCL-1 is composed of 369 amino acid residues and contains a characteristic cyclin domain encompassing the amino acid region from 156 to 340 (Figure 1A; Gibson et al., 1994). Sequence alignment of PCL-1 with the orthologous Pcls from yeast (Figure 1B) showed that they share a C-terminus conserved region only in the amino acid sequence corresponding to the cyclin box domain (CBD, 24% identity and e-value 3.10^–14 compared with Pcl10). This region shows high divergence regarding the conserved amino acid residues, there are large insertions and only a few residues are identical in all proteins.

To confirm if PCL-1 has a folding characteristic of a cyclin, its structure was predicted using AlphaFold. As shown in Figure 1C, PLC-1 model presents a globular region formed by the arrangement of different α-helices predicted with high confidence. Such a region corresponds to a classical cyclin domain (Tatum and Endicott, 2020), which is composed of two N-terminal helices (αNT1 and αNT2) followed by five helices (α1–α5) and one C-terminal helix (αCT1). Furthermore, PCL-1 was superposed to Pcl10 (Zheng and Quiocho, 2013) to corroborate the presence of such a cyclin-like domain, evidenced by the structural homology of all α-helices described above with Cα atoms RMSD value of only 0.63 Å.

We investigated if PCL-1 is involved in the N. crassa glycogen metabolism control similar to the S. cerevisiae Pcl6p, Pcl7p, Pcl8p, and Pcl10p counterparts. Glycogen quantification in the Δpcl-1 strain at different times of vegetative growth revealed that the mutant strain accumulates higher levels of glycogen, approximately twice more as much as the wild-type strain (Figure 2A). We further quantified the glycogen synthase (GSN) phosphorylation status as an attempt to correlate the enzyme phosphorylation with the glycogen stored by the mutant strain. Phosphorylation inhibits glycogen synthase activity (Téllez-Iñón et al., 1969). Enzymatic activity was quantified in the same mycelial pads used for glycogen quantification in the presence and in the absence of glucose-6-phosphate (G6P), the glycogen synthase allosteric modulator. The −/+ G6P activity ratio is considered as an index of phosphorylation, with higher levels being correlated with lower phosphorylation and thus, higher activity. The activity ratio in the Δpcl-1 strain was higher than that in the wild-type strain at the initial growth times, suggesting that GSN is less phosphorylated in the mutant strain (Figure 2B). This result may explain the higher glycogen levels observed in the mutant strain under the same growth condition. As the −/+ G6P activity ratio correlates with different isoforms of phosphorylated GSN, we decided to investigate the phosphorylated forms of GSN. For that, 24 h mycelial pads from the wild-type and mutant strain were analyzed by 2D-gels followed by Western blot using anti GSN antibody (Figure 2C). Five GSN differentially phosphorylated isoforms were observed in the wild-type strain, and treatment with protein phosphatase reduces to two isoforms, as we previously demonstrated (Candido et al., 2014). Only three GSN phosphorylated isoforms were visualized in the Δpcl-1 strain, suggesting that GSN is less phosphorylated in this strain and that PCL-1 influences the phosphorylation status of GSN.

As mentioned, in S. cerevisiae, the serine/threonine Pho85p protein kinase, together with Pcls, phosphorylates Gsy2p (Huang et al., 1996, 1998; Wilson et al., 1999). In addition, we previously identified the ORF NCU07580 product, the ortholog of the S. cerevisiae Pho85p, as a kinase protein regulating glycogen metabolism (Candido et al., 2014). This protein is annotated in the N. crassa database as a cyclin-dependent protein kinase, and based on the high identity (64%) between the N. crassa and yeast kinase proteins we named the NCU07580 product as PHO85-1, according to the N. crassa nomenclature. The N. crassa protein exhibits the PSTAIRE and the glycine-rich domains (Supplementary Figure 1), which characterize CDKs (Pines, 1995; Nishizawa et al., 1999), as well as the T-loop region encompassing the threonine residue (Thr164 in PHO85-1), which needs to be phosphorylated for activation in CDKs other than Pho85p (Nishizawa et al., 1999). Using an in vitro GSN phosphorylation assay, we investigated whether PCL-1 would be the PHO85-1 partner cyclin that directs the kinase to glycogen metabolism. For this, both His-tagged recombinant proteins (His-PCL-1 and His-PHO85-1) were produced in E. coli and partially purified by affinity chromatography. The His-tagged SUMO-fused GSN wild-type and mutant proteins were also produced in E. coli and partially purified by affinity chromatography.

Based on an amino acid sequence alignment with the rabbit muscle (Gys1) and the S. cerevisiae (Gsy2p) glycogen synthases, we identified four C-terminus putative phosphorylation sites in GSN, the Ser632, Ser636, Thr641, and Thr645 amino acid residues (Figure 3A). Like the yeast enzyme, the N. crassa enzyme lacks the N-terminus phosphorylation sites. We next investigate whether PCL-1 could be a partner cyclin of the PHO85-1 protein kinase in the in vitro GSN phosphorylation, and to identify the amino acid residue target of phosphorylation by the PHO85-1/PCL-1 complex. Site-directed mutagenesis was used to introduce point mutations in the His-SUMO-GSN by replacing the putative serine and threonine phosphorylation residues with alanine, generating the single S632A, S636A, T641A, and T645A and the quadruple S632A/S636A/T641A/T645A mutant proteins. We first analyzed the wild-type His-SUMO-GSN phosphorylation using as source of protein kinases the crude cellular extract from a knockout strain that does not synthesize GSN (FGSC#18932, gsn:hyg). BSA was used as a negative control in the assay. We observed that GSN was phosphorylated in vitro by the kinases present in the cellular extract (Figure 3B, lane 2). In addition, GSN was phosphorylated only when the PHO85-1 protein kinase and the PCL-1 cyclin were combined in the reaction mixture (Figure 3B, lane 6). This result confirms that PHO85-1 is the N. crassa ortholog of the yeast Pho85p protein, and that PCL-1 acts as a partner cyclin of the PHO85-1 protein kinase in the GSN phosphorylation. No phosphorylation was observed in the absence of either the cyclin (Figure 3B, lane 4) or the protein kinase (Figure 3B, lane 5). We further assayed the phosphorylation of the GSN mutants by the PHO85-1/PCL-1 complex. The GSN wild-type protein incubated in absence of the complex (Figure 3C, lane 1) and the recombinant SUMO protein (Figure 3C, lane 2) were used as negative controls. We observed that only the GSNS636A mutant was not phosphorylated in vitro, indicating that the Ser636 amino acid residue is the amino acid residue target of phosphorylation by the kinase/cyclin protein complex (Figure 3C, lane 5). No phosphorylation was observed when the quadruple mutant GSN (S632A/S636A/T641A/T645A) was used as substrate, indicating that no additional sites are target of phosphorylation by the PHO85-1/PCL-1 complex (Figure 3C, lane 8).

As shown in Figure 3, the PHO85-1 kinase activity depends on the presence of PCL-1, suggesting that these proteins interact to result in a functional CDK/cyclin complex. Therefore, the structure of PHO85-1 was predicted using AlphaFold as previously applied to PCL-1. Five models of PHO85-1 were generated, which converged to only two conformations (Supplementary Figure 2, models 1 and 2). In general, the predicted PHO85-1 structures presented high confidence, except for the PSTAIRE helix and the T-loop. Interestingly, these regions were predicted with high confidence using AlphaFold Multimer (Supplementary Figure 2, model 3) only when PCL-1 was present, suggesting that these regions may assume stable conformations only when in complex with their partner, resulting in a salt-bridge between the Lys39 and Glu56 residues of PHO85-1.

Figure 4A shows the predicted PHO85-1/PCL-1 complex, considering the presence of ATP and Mg²⁺ molecules (see Section “Materials and methods”), which was further submitted to MD simulations to verify its structural stability and to determine the main contacts between them. Some important regions at the interface of the complex are shown, such as the PSTAIRE helix and the T-loop of PHO85-1 and the helices α3, α5, and αCT1 of PCL-1. When analyzing these regions in detail, a salt-bridge is observed between the residue Asp280, located in the α3 helix of PCL-1, and three Arginine residues of PHO85-1, named Arg55 (PSTAIRE helix) and Arg132 and Arg156 (T-loop). Another important region found to stabilize the complex was determined by hydrophobic interactions between the PSTAIRE helix residues, such as Ile54, and the PCL-1 residues placed in the α3 and α5 helices, such as Phe312 and Val273 (Figure 4A, zoom-in-view panels).

Molecular dynamics (MD) simulations of the PHO85-1/PCL-1 complex showed considerable stability and reproducibility in all three replicas, presenting a root-mean-square deviation (RMSD) average value of 2.2 ± 0.49 Å (Figure 4B) and root mean-square fluctuation (RMSF) values for each protein below 3 Å in every replica (Figure 4C), except for the first N- and last C-terminal residues. Considering the whole amount of data, it was possible to determine the prevalence of contacts of the PHO85-1/PCL-1 interface. The salt-bridge led by Asp280 residue of PCL-1 is strongly stable, presenting contacts to Arg55, Arg132, and Arg156 during 74.83, 99, 70, and 75.02% of the time. Asp280 is surrounded by these three Arginine residues, allowing an environment with high complementarity of charges. Moreover, a hydrophobic-rich region formed by the complex was also considered important for stabilizing the PHO85-1/PCL-1 complex (Figure 4A, zoom-in-view panels), highlighting the contacts of Phe312 (PCL-1) to Ile57 (PHO85-1) during 99.77%, and Leu314 (PCL-1) to Ile54 (PHO85-1) during 81.72% of time.

As N. crassa synthesizes a limited number of cyclins (Borkovich et al., 2004; Hong et al., 2014), and that in yeast different cyclins direct the Pho85p kinase to different functions, we decided to investigate whether PCL-1 is involved in additional cellular functions. To verify whether PCL-1 influenced growth, we first analyzed conidia germination of the WT and Δpcl strains and the Δpcl-1 pcl-1⁺ complemented strain in samples collected from liquid cultures under agitation. The amount of conidia in germination was quantified at different times and we observed a delay in the germination rate of the Δpcl strain at the initial times of culture, mainly up to 6 h of growth. However, normal growth rate was achieved at later times of culture (8 h) (Figure 5A, left and right panels). Conidia germination was also analyzed in conidia inoculated onto coverslips and covered with liquid medium (Figure 5B). In this assay, the delay in germination is more evident because the mutant strain shows shorter hypha, mainly at the time of 6 h, when compared to the wild-type strain. The germination delay was not observed in the complemented strain indicating that the phenotype was due to the PCL-1 lacking. It is important to mention that no changes in the hypha morphology was observed, and that the complemented strain expresses the protein under the control of the native promoter.

To investigate the PCL-1 involvement in cell cycle progression, we constructed the Δpcl-1 hH1-sfgfp strain, a Δpcl-1 mutant strain that synthesizes the sfGFP-fused H1 histone (Figure 6A) and allows quantification of the nuclei in different mitotic phases (Figure 6A, right panel). Fewer nuclei under division were counted in the mutant strain compared to the non-mutant strain (hH1-sfGFP) indicating that PCL-1 may also control cell cycle progression (Figure 6B). We have not observed accumulation of cells in any specific mitotic phase, suggesting that PCL-1 is continuously synthesized during the cell cycle. The germination and cell progression data suggested that PCL-1 may play a role in fundamental cellular processes.

Since PCL-1 is involved in different cellular processes, we next investigated the cellular distribution pf PCL-1 at different times of conidia germination using the Δpcl-1 pcl-1⁺ complemented strain, which express sfGFP-fused PCL-1. The conidia of the strain showed a strong fluorescent signal throughout the cytoplasm (Figure 7, 0 h). Interestingly, during germination, cells displayed a noticeably subcellular accumulation of the protein in the cytosol of hyphae (Figure 7, later times, white arrowheads) indicating that PCL-1 may accumulate in certain cell regions during germination as granular clusters. In addition, accumulation in nuclei was also observed (Figure 7, red arrowheads), suggesting that the protein may also translocate into nucleus, although classical nuclear localization signals (NLS) were not identified in the protein primary sequence.

In S. cerevisiae, the Pho85 protein kinase, together with Pho80 partner cyclin, is a kinase that phosphorylates the zinc finger Crz1 transcription factor, which is involved in the regulation of calcium metabolism and activated by dephosphorylation by the calcineurin phosphatase (Carroll and O’Shea, 2002; Sopko et al., 2006). The fact that N. crassa synthesizes low number of cyclins (Borkovich et al., 2004), prompted us to investigate if PCL-1 could be a cyclin involved in calcium metabolism regulation. Hyphal growth in Petri dishes of the Δpcl-1 strain was strongly increased under high calcium chloride concentration (200–300 mM) suggesting that PCL-1 participates in the calcium stress response (Figure 8A; Supplementary Figure 3). The Δpcl-1 pcl-1⁺ complemented strain exhibited comparable growth to the wild-type strain (Figure 8A; Supplementary Figure 3), confirming that the growth defect in the mutant strain was indeed due to the pcl-1 deletion.

The expression of three genes encoding Ca²⁺⁺-ATPases, pmr-1 (NCU03292), nca-2 (NCU04736), and nca-3 (NCU05154) (Bowman et al., 2011, 2012) was analyzed after growing the wild-type and mutant strain under 300 mM calcium chloride for 24 h. Gene expression of all three genes was increased in the mutant strain compared to the wild-type strain indicating a role for PCL-1 in the expression regulation of these genes and, therefore, in the control of the calcium stress response in N. crassa (Figure 8B). To investigate if PCL-1 influences the cellular response to other stressors, we analyzed growth of the wild-type, mutant, and complemented strains in VM pH 7.8 (alkaline pH stress) and in VM containing NaCl and sorbitol (osmotic stress), H₂O₂ (oxidative stress), and Congo Red (plasma membrane/cell wall stress). The mutant strain showed sensitivity only to osmotic stress induced by NaCl (Figure 8C).

As the mutant strain exhibited increased growth and upregulation of calcium responsive genes under high calcium chloride concentration, we decided to extend the investigation on the calcium-responsive CRZ-1 transcription factor. We first analyzed the protein expression in the wild-type and Δpcl-1 strains both endogenously expressing C-terminus V5-tagged CRZ-1 at different times of growth under low (10 mM) and high (300 mM) calcium chloride concentrations. In this experiment, the protein samples were treated with phosphatase due to the high number of putative phosphorylation sites that might exist in the N. crassa protein. CRZ-1-V5 migrates as an apparent molecular mass between 90 and 100 kDa. Interestingly, we observed variation in the CRZ-1 expression both in a strain- and calcium chloride dependent manner (Figure 9A). Under high calcium chloride concentration the CRZ-1 expression in the wild-type strain increased in the time range from 0 to 30 min compared with low concentration. However, very low levels of protein were detected at 60 min under the same condition. On the other hand, the expression in the Δpcl-1 strain decreased under high calcium chloride concentration in the time range from 0 to 60 min compared with low calcium concentration. In summary, CRZ-1 expression is modulated by calcium concentration and by PCL-1.

It was previously mentioned here that the S. cerevisiae Pho85 kinase, together with Pho80 cyclin, is a kinase that phosphorylates and inactivates the Crz1p transcription factor (Sopko et al., 2006). Phosphorylated Crz1p localizes in the cytosol (Cyert, 2003). Based on the multiple functionalities of PCL-1, we hypothesized that CRZ-1 would be less phosphorylated in the Δpcl-1 strain than in the wild-type strain, and therefore, being able to be translocated to the nucleus. The A. fumigatus CrzA orthologous transcription factor has been reported as a highly phosphorylated protein in vivo, and a total of twenty phosphorylation sites were identified in the CrzA by mass spectroscopy; some of them predicted as potential target motifs for CDKs by bioinformatic tool (Shwab et al., 2019). Comparison between the A. fumigatus CrzA and the N. crassa CRZ-1 proteins by sequence alignment (Supplementary Figure 4) showed that many putative phosphorylation sites are conserved in the N. crassa protein (highlighted in red), including two potential CDK targets. We decided to analyze the phosphorylation status of CRZ-1-V5 in the wild-type and Δpcl-1 strains under high (300 mM) calcium chloride concentration (Figure 9B). Again, we observed in the mutant strain lower CRZ-1 expression than the expression in the wild-type strain in the presence of calcium. In addition, both strains exhibited similar expression levels in the absence of calcium (control, compare to Figure 9A at 0 time). However, we did not observe any differences in the phosphorylation levels of CRZ-1, even after protein phosphatase treatment. Using PhosTag gel as a different experimental approach to find out whether phosphorylation of CRZ-1 is affected by PCL-1 we were not able obtain a conclusive result, most likely due to the many putative phosphorylation sites in CRZ-1 (results not shown).

We decided to analyze the CRZ-1-GFP cellular localization in the wild-type and Δpcl-1 background strains, both expressing the protein under the control of the native promoter, in the presence of calcium chloride (300 mM) (Figure 9C). In the wild-type strain, strong fluorescence was visualized throughout the cytoplasm in the absence of calcium, as expected (Figure 9C, control). However, upon calcium stress, the protein translocates into nucleus (Figure 9C), as described for the A. fumigatus and A. nidulans CrzA orthologs (Soriani et al., 2008; Hernández-Ortiz and Espeso, 2013). Interestingly, in the Δpcl-1 strain, in the absence of calcium we observed protein localization both in the nuclei and cytoplasm, depending on the hyphae analyzed. On the other hand, upon calcium stress, the protein mainly localized in the nuclei (Figure 9C). These results strongly support our hypothesis, in which CRZ-1 is less phosphorylated in the Δpcl-1 background strain compared to the wild-type background strain, being able to translocate into nucleus and to regulate the expression of calcium responsive genes.