Using a strain expressing C-terminally TAP-tagged RpdA (RpdA^TAP) we conducted tandem affinity purification (TAP; Rigaut et al., 1999) coupled to liquid chromatography-tandem mass spectrometry (LC-MS/MS). Hits that were identified by at least two peptides were searched for homologs in budding and fission yeast databases via BLASTp. This suggested the presence of RpdA complexes corresponding to those described in yeast, i.e., RpdA-L and RpdA-S (Figure 1A and Supplementary Table 1; Lechner et al., 2000; Carrozza et al., 2005a; Nicolas et al., 2007; Baker et al., 2013; Zilio et al., 2014). However, not all known yeast complex members were detected, mainly due to the fact that the A. nidulans genome lacks the respective homologs. Based on the presented data, composition of a third RpdA complex homologous to Rpd3μ (Rpd3-Snt2-Ecm5; Baker et al., 2013; McDaniel and Strahl, 2013) is not entirely elucidated yet and requires further investigation (Supplementary Table 1).

Among the remaining hits found in two analyses, two highly ranked proteins, ScrC and AN4022, attracted our attention. To investigate their putative interaction with RpdA, strains expressing C-terminally TAP-tagged versions of these candidates, ScrC^TAP and FscA^TAP, were generated, which were then used as baits for vice versa purifications. Results actually confirmed the interaction of both proteins with each other and with RpdA (Table 1). Moreover, SinC, PrwA, and SdsC were highly enriched and also LafA was present in each purification using ScrC^TAP and FscA^TAP as baits. To exclude co-purification of the detected proteins by interaction with the TAP-tag only, GFP-trap experiments using a strain expressing C-terminally Venus-tagged RpdA, RpdA^Venus, were performed. Results confirmed those of the TAP experiments (data not shown). Data suggest that ScrC and FscA, together with the RpdA core (RpdA, SinA, PrwA) and two other proteins also found in RpdA-L complexes (LafA and SdsC) constitute a novel RpdA complex in A. nidulans. While ScrC was identified in a suppressor screen for ΔcrzA phenotypes (Almeida et al., 2013), the previously uncharacterized second novel interaction partner (AN4022) was designated FscA (friend of ScrC, see below). Consequently, we have termed the novel complex RpdA core LafA SdsC ScrC FscA (RcLS2F) complex (Figure 1B).

In order to estimate the native size of the RcLS2F complex, fungal extracts of a strain expressing FscA with a C-terminal Venus fluorescent protein tag, FscA^Venus, were loaded onto a Superose 6 size exclusion column. Western blots of elution fractions probed with antibodies against Venus revealed a FscA peak with retention corresponding to an apparent molecular mass of greater than 1 × 10⁶ Da (Supplementary Figure 1). Importantly, immunodetection of RpdA showed a retention profile similar to FscA, providing an independent confirmation of the TAP results (Supplementary Figure 1).

In order to investigate RcLS2F composition more precisely, scrC and fscA mutants in an RpdA^TAP strain were generated and used for purifications as described above. In the ΔfscA strain, all interaction partners apart from FscA were precipitated by the RpdA^TAP bait (Table 2). In the ΔscrC strain, however, neither ScrC nor FscA were identified in the RpdA TAP elution fraction (Table 2). This result indicates that ScrC directly recruits FscA to the RpdA complex and is critical for complete RcLS2F formation.

To test the cellular localization of ScrC and FscA, strains expressing C-terminally Venus-tagged versions, ScrC^Ven and FscA^Ven, under control of the tunable Penicillium chrysogenum xylanase promoter (Zadra et al., 2000; Tribus et al., 2010) were used for epifluorescence microscopy. These experiments revealed that, as shown for RpdA (Bauer et al., 2016), both proteins are mainly located within the nucleus (Figure 2A) confirming RcLS2F as a nuclear complex. In order to determine RcLS2F as catalytically active complex, we also used ScrC^TAP and FscA^TAP strains for IgG pull-downs. Both precipitates clearly showed deacetylase activity which amounts to about 30% of total RpdA activity that was pulled down by RpdA^TAP (Figure 2B). This result confirms histone deacetylase activity of RcLS2F in vitro.

To determine the conservation of ScrC and FscA, BLASTp¹ searches against available proteomes were performed. These analyses revealed that both proteins are conserved within the Ascomycetes subclade Eurotiomycetidae only, a fungal group containing a number of medically, biotechnologically, and agriculturally relevant genera such as Aspergillus, Penicillium, Coccidioides, or Histoplasma. As determined applying the fungiDB² genome browser, both genes are arranged in similar syntenic surroundings. As searches of the Pfam database³ (El-Gebali et al., 2019) did not yield any known protein domains, orthologs of nine fungal species were used for MEME⁴ analysis (Bailey and Elkan, 1994). Interestingly, a putative conserved domain of 150 aa (Figures 3A,C) preceded by a glycine-arginine-rich (GAR) region (Figures 3A,B) was identified in ScrC orthologs. Alignment of FscA orthologs revealed presence of two conserved domains in all nine species (Figures 3D–F), a GAR region, however, could only be detected for Aspergillus and Penicillium orthologs (Figure 3D). Regions comprising a remarkably high proline content in both proteins (>35% within their 100 C-terminal residues) prompted us to use PrDOS⁵ (Ishida and Kinoshita, 2007) for prediction of intrinsically disordered regions. These analyses revealed that major fractions of the tested proteins have high disorder propensity (Figures 3A,D). Remarkably, major fractions of the residues below a disorder probability of 0.6 were identified within the indicated conserved domains (Figures 3A,D).

To further characterize FscA and ScrC, we generated deletion mutants by replacing the corresponding coding sequences with the pyrG auxotrophic marker. Mutant strains displayed normal colony morphology with reduced diameter compared to the wild type (Figure 4A). Radial growth was reduced to approximately 80% of the wild type in both mutants, regardless of the medium used (Figure 4B). Complementation by reintegration of genes coding for the expression of Venus-tagged ScrC and FscA at the respective deletion loci reverted the growth phenotypes (Figures 4A,B) and confirmed the functionality of the Venus-tagged versions. Moreover, also conidiation of the mutant strains was compared to that of the wild type. ΔscrC and ΔfscA strains showed a significant reduction of conidiation that was restored in the complemented strains (Figure 4C). Furthermore, selfing experiments were conducted to analyze, if lack of RcLS2F would interfere with the capability of sexual reproduction. Interestingly, both mutants were not able to form cleistothecia but instead, only produced empty nests of Hülle cells (Figure 4D). This phenotype was fully restored in the reconstituted strains (Figure 4D).

CrzA is the transcription factor targeted by the calcium (Ca²⁺) signaling cascade in fungi and its deletion becomes deleterious under several types of stresses, like exposure to Ca²⁺, manganese, or alkaline pH (Hagiwara et al., 2008; Spielvogel et al., 2008). In a ΔcrzA suppressor screen, the RcLS2F-defining component ScrC was identified among others (Almeida et al., 2013). Comparable phenotypes of ΔscrC and ΔfscA as well as their mutual co-purification indicated similar functions within the RcLS2F complex. To examine this assumption in more detail, the coding sequence of crzA was deleted in a wild type, a ΔscrC, and a ΔfscA background and plate growth analyses were performed under Ca²⁺, manganese and alkaline pH stress. Intriguingly, both ΔscrC and ΔfscA strains attenuated the deleterious effects of the crzA deletion (Figure 5A). Interestingly and in contrast to low Ca²⁺, manganese, or pH, high Ca²⁺ revealed a reduced suppressive capacity of ΔfscA when compared to ΔscrC (Figure 5A), which is in line with the presumption of a recruitment of FscA via ScrC. Interestingly, calcium clearly increased radial growth in the wild type but neither in ΔfscA nor in ΔscrC strains (Figure 5A).

The results shown above suggested a connection of the suppressor function of ΔscrC and ΔfscA via the RcLS2F complex. It was tempting to speculate that disruption of this nuclear complex might interfere with proper regulation of transcription. To test if the expression of known Ca²⁺-induced CrzA-dependent genes would be affected by disruption of RcLS2F, two genes were selected for northern analysis under Ca²⁺-induction: vcxA, a previously characterized vacuolar Ca²⁺/H⁺ exchanger (Spielvogel et al., 2008), and AN9206, a gene reported to be strongly induced by Ca²⁺ (Hagiwara et al., 2008). Ca²⁺ induction of vcxA and AN9206 was not affected by scrC or fscA deletion in the crzA+ background. In the absence of Ca²⁺, however, de-repression of both genes was observed in both mutants (Figures 5B, 6A). Densitometric quantification of northern blotting results revealed an increase of transcript abundance of 2.4/3.3 × and 3.0/4.4 × for vcxA/An9206 in ΔfscA and ΔscrC mutants, respectively (Supplementary Table 2). In the ΔcrzA background, none of the strains could induce expression of these genes, but again both vcxA and AN9206 were clearly de-repressed in the RcLS2F deletion mutants (Figure 5B). Remarkably, release of repression generally was more pronounced in ΔscrC, which fits to the crucial role of ScrC in the assembly of the RcLS2F complex.

The observed antagonism between CrzA and ScrC/FscA raised the question if expression scrC/fscA themselves would be affected by Ca²⁺. Indeed, northern analysis showed that both fscA and scrC were transcriptionally repressed upon exposure to Ca²⁺ in the wild type. In the deletion strains, however, this repression was replaced by a strong up-regulation of scrC in ΔfscA and vice versa, indicating an autoregulatory feed-back loop epistatic to Ca²⁺-based repression (Figure 5C). We then examined if this autoregulation would affect other RcLS2F components as well. In contrast to fscA and scrC, however, expression of the remaining complex partners (rpdA, sinC, prwA, sdsC, lafA) was not affected in the two mutants (Figure 5D). Taken together, these results indicate transcriptional autoregulation of scrC and fscA and a RcLS2F-mediated repression of vcxA and AN9206 in the absence of Ca²⁺ stress conditions.

The lack of transcriptional repression observed in both RcLS2F mutants prompted us to investigate whether this regulation depends on RcLS2F deacetylase activity. Due to the pleiotropic phenotype of rpdA knock-down mutants resulting in severe growth defects, pharmacological inhibition of RpdA by trichostatin A (TSA) treatment was used to address this question. Indeed, TSA treatment for 30 min caused de-repression of vcxA and AN9206 in wild type mycelia comparable to that of the ΔfscA and slightly lower than that in the ΔscrC mutant (Figure 6A). Remarkably, TSA treatment also led to de-repression of fscA and scrC in the wild type (Figure 6B), which is in line with the observed autoregulation of RcLS2F (Figure 5B).

Taken together, our data suggest that the novel mold-specific RcLS2F KDAC complex, containing two poorly characterized proteins as defining and crucial components, is functionally implicated in Ca²⁺-mediated transcriptional regulation via its co-repressor activity.

Strains used in this study are listed in Supplementary Table 3. Strains were grown on glucose minimal medium [MM: 1% (m/v) glucose, 10 mM di-ammonium tartrate, salt solution, and trace elements] or glucose complete medium [CM: 2% (m/v) glucose, 0.2% (m/v) peptone, 0.1% (m/v) yeast extract, 0.1% (m/v) casamino acids, salt solution, and trace elements] as described (Cove, 1966). If not stated otherwise, MM was supplemented with 0.1 μg/ml biotin and 4 mM arginine. All other auxotrophs were supplemented as described elsewhere (Todd et al., 2007). Alleles driven by the xylP promoter were induced by addition of 0.05–1% of xylose to MM.

For cloning, amplification, digestion, and propagation of DNA fragments and vectors, standard molecular techniques were used (Sambrook and Russell, 2001). Oligonucleotides and plasmids used for strain construction are listed in Supplementary Table 4. Cassettes for genetic manipulation were generated by PCR fusions as described (Shevchuk et al., 2004) or by In-Fusion HD Cloning (Clontech) in accordance with the manufacturer’s instructions. Expression constructs for FscA^Venus and ScrC^Venus were generated by cloning the respective coding sequences amplified from genomic DNA into NcoI and NotI sites of pIB97, which descends from pIB92 (Bauer et al., 2016) but with a functional argB allele. Fungal transformations were performed as described previously (Szewczyk et al., 2006) using an nkuA disruptant that was generated as follows. First, strain RIB1.22 harboring biA1, riboB2, pyrG89, and argB2 alleles was isolated after genetic crossing (Todd et al., 2007). Next, the sequence coding for the C-terminus of NkuA was deleted using the Af_riboB marker. The resulting strain TIB24.12 was prone to homologous recombination due to a lack of functional NkuA and was used as recipient in subsequent transformations. Deletions of fscA and scrC were performed employing the Af_pyrG marker, crzA deletions with the Af_bioDA marker (Magliano et al., 2011). A. fumigatus (Af) orthologs were used as auxotrophic marker genes (argB, bioDA, pyrG, riboB) for transformation of non-homologous end-joining deficient strains.

For the complementation of fscA and scrC, a CRISPR-Cas9 system as described by Nødvig et al. (2015, 2018) was used to render the Af_pyrG marker non-functional by insertion of a premature termination codon (Supplementary Figure 3). The resulting pyrG auxotrophs were then transformed with cassettes composed of sequences encoding for C-terminally Venus-tagged versions of both deleted proteins, flanked by their 5′ UTR and part of the wild type Af_pyrG marker allele. By using a truncated pyrG, ectopic integration was excluded. fscA^Venus/scrC^Venus fusions to pyrG were performed by PCR. To generate RpdA^TAP strains with scrC and fscA deletions, the pyrG auxtotroph ANOB486 was transformed with respective Af_pyrG marker constructs (see above). Transformants were confirmed by PCR screening and single integration of the deletion constructs was verified by Southern analysis as described elsewhere (Graessle et al., 2000).

Tandem affinity purification was performed as described (Bayram et al., 2012b; Bauer et al., 2019b). Briefly, C-terminally TAP-tagged strains were grown for 14 h in MM (including 0.05% xylose for TIB32.1) at 37°C and 180 rpm. Fungal mycelia were lyophilized and extracted as described (Bauer et al., 2019b). Purifications were monitored by western blotting using antibodies against RpdA or calmodulin binding protein and silver staining (Blum et al., 1987). For protein identification, samples were precipitated with TCA and then subjected to SDS-PAGE, together with a pre-stained protein ladder. To minimize separation of large proteins in the stacking gel, percentage was reduced to 3.5%. Shortly after the 300 kD marker band had left the stacking gel, electrophoresis was stopped and gels were stained with Coomassie brilliant blue. Gel slices from the lowest stained bands up to the end of the separating gel were used for in-gel trypsin digestion of the protein mixture followed by liquid chromatography-tandem mass spectrometry (LC-MS/MS).

Method 1 (used for all identifications except RpdA TAP in TIB32.1).

Proteins in gel slices excised from SDS-PAGE gels were reduced with dithiothreitol, alkylated with iodoacetamide, and digested with trypsin from porcine pancreas (Promega) as described previously (Faserl et al., 2019). Tryptic digests were analyzed using an UltiMate 3000 RSCLnano-HPLC system coupled to a Q Exactive HF mass spectrometer (both Thermo Scientific), equipped with a Nanospray Flex ionization source. The peptides were separated on a homemade fritless fused-silica micro-capillary column (100 μm i.d. × 280 μm o.d. × 20 cm length) packed with 2.4 μm reversed-phase C18 material (Reprosil). Solvents for HPLC were 0.1% formic acid (solvent A) and 0.1% formic acid in 85% acetonitrile (solvent B). The gradient profile was as follows: 0–4 min, 4% B; 4–57 min, 4–35% B; 57–62 min, 35–100% B, and 62–67 min, 100% B. The flow rate was 300 nl/min.

The Q Exactive HF mass spectrometer was operated in the data dependent mode selecting the top 20 most abundant isotope patterns with charge state > 1 from the survey scan with an isolation window of 1.6 mass-to-charge ratio (m/z). Survey full scan MS spectra were acquired from 300 to 1750 m/z at a resolution of 60,000 with a maximum injection time (IT) of 120 ms, and automatic gain control (AGC) target 1 × 10⁶. The selected isotope patterns were fragmented by higher-energy collisional dissociation with normalized collision energy of 28 at a resolution of 30,000 with a maximum IT of 120 ms, and AGC target 5 × 10⁵.

Data Analysis was performed using Proteome Discoverer 2.1 (Thermo Scientific) with search engine Sequest. The raw files were searched against “FungiDB-36_AnidulansFGSCA4_ AnnotatedProteins” database (Galagan et al., 2005; Basenko et al., 2018). Precursor and fragment mass tolerance was set to 10 ppm and 0.02 Da, respectively, and up to two missed cleavages were allowed. Carbamidomethylation of cysteine was set as static modification and oxidation of methionine was set as variable modification. Acetylation, methionine-loss, and methionine-loss plus acetylation were set as N-terminal dynamic modification of proteins. Peptide identifications were filtered at 1% false discovery rate.

Method 2 (used for initial RpdA TAP).

Protein bands were excised from gels and digested with trypsin from porcine pancreas (Sigma-Aldrich) as described previously (Faserl et al., 2019). Tryptic digests were analyzed using an UltiMate 3000 nano-HPLC system coupled to an LTQ Orbitrap XL mass spectrometer (both Thermo Scientific) equipped with a nanospray ionization source. The peptides were separated on a homemade fritless fused-silica microcapillary column (75 μm i.d. × 280 μm o.d. × 10 cm length) packed with 3 μm reversed-phase C18 material (Reprosil). Solvent for HPLC were 0.1% formic acid (solvent A) and 0.1% formic acid in 85% acetonitrile (solvent B). The gradient profile was as follows: 0–2 min, 4% B; 2–55 min, 4–50% B; 55–60 min, 50–100% B, and 60–65 min, 100% B. The flow rate was 250 nl/min.

The LTQ Orbitrap XL mass spectrometer was operated in the data dependent mode selecting the top 10 most abundant isotope patters with charge state 2 + and 3 + from the survey scan with an isolation window of 2 mass-to-charge ratio (m/z). Survey full scan MS spectra were acquired from 300 to 2000 m/z at a resolution of 60,000 with a maximum IT of 20 ms, and AGC target 1 × 10⁶. The selected isotope patterns were fragmented by collision induced dissociation (CID) with normalized collision energy of 35, and a maximum injection time of 55 ms.

Data analysis was performed using Proteome Discoverer 1.4 (Thermo Scientific) with search engine Sequest. The raw files were searched against “aspergillus_nidulans_ fgsc_a4_1_proteins” database (BROAD Institute; Galagan et al., 2005). Precursor and fragment mass tolerance was set to 10 ppm and 0.8 Da, respectively, and up to two missed cleavages were allowed. Carbamidomethylation of cysteine and oxidation of methionine were set as variable modifications. Peptide identifications were filtered at 1% false discovery rate.

Polyacrylamide gels were blotted onto nitrocellulose membrane using the Trans-Blot Turbo system (Bio-Rad). Blots were then blocked in TBS including 5% skim milk powder or 4% Ficoll 400 and probed with antibodies directed against RpdA (1:1300 dilution, Trojer et al., 2003), HdaA (1:1000 dilution, Trojer et al., 2003), GFP (Merck 11814460001, 1:5000 dilution), or CBP (Roche 07-482, 1:1333 dilution). Antibodies were detected with alkaline phosphatase conjugate antibodies directed against rabbit or mouse IgG (Sigma A3687 and A3562, respectively, 1:30000 dilution) and visualized using the BCIP/NBT colorimetric substrate (Promega S3771). Alternatively, primary antibodies were detected with fluorescent secondary antibodies (IRDye800/IRDye680, LI-COR Biosciences) according to the manufacturer’s instructions and visualized using the Odyssey imaging system (LI-COR Biosciences).

If not otherwise stated, strains were grown for 18 h in MM with an inoculation density of 2 × 10⁶/ml. Total RNA was prepared from lyophilized fungal mycelia using TRI reagent (Sigma) according to the manufacturer’s instructions. RNA was electrophoresed in 1.2% agarose gels as described (Graessle et al., 2000) and then transferred to nylon membranes (Amersham Hybond-N) by down-ward capillary transfer (Chomczynski and Mackey, 1994). Digoxigenin-dUTP-labeled DNA probes specific for the analyzed transcripts were amplified with primers shown in Supplementary Table 4. Hybridized DNA probes were detected with alkaline phosphatase-conjugated anti-digoxigenin Fab fragments (Roche) and developed with CSPD chemiluminescent substrate (Roche) according to the manufacturer’s instructions. Signals were visualized with the Fusion-SL 3500 WL imaging system (Vilber Lourmat).

Enzymatic activity of protein fractions was measured using [³H] acetate-labeled chicken histones as substrate as described (Bauer et al., 2019b). Briefly, fractions of 50 μl were mixed with 10 μl of labeled histones (4 mg/ml) and incubated at 25°C for 60 min. The reactions were stopped and extracted with ethylacetate. The organic phase was counted in an AccuFLEX LSC-8000 (HITACHI) liquid scintillation counter.

To determine the subcellular localization of FscA, ScrC, and RpdA, strains expressing Venus-tagged proteins (Gsaller et al., 2014) were grown on cover slips in 6 well plates at 30°C overnight under conditions of induction of the xylP promoter (1% xylose). A strain expressing RpdA-Venus under the control of the xylanase promoter (TIB92n1; Bauer et al., 2016) was used as reference. Chromatin was stained with DAPI or by RFP-tagged H2A (Bayram et al., 2012a).

To monitor fungal growth on solid medium, images of samples were taken with a Nikon D5100 SLR camera or a Leica MZ16 stereomicroscope (Leica Microsystems GmbH) equipped with an AxioCam MRc camera (Carl Zeiss GmbH). For fluorescence microscopy, an Axioplan microscope (Carl Zeiss GmbH), equipped with (excitation/emission filters 365/420 nm for blue fluorescence, 500/535 nm for yellow fluorescence, and 565/620 nm for red fluorescence, Carl Zeiss GmbH) was used for imaging.

Image processing and editing was achieved with the programs Axio Vision (Carl Zeiss GmbH), the ImageJ distribution Fiji (Schindelin et al., 2012; Schneider et al., 2012), Affinity Photo (Serif, Inc.), and Affinity Designer (Serif, Inc.).