The Ca2+/Calcineurin-Dependent Signaling Pathway in the Gray Mold Botrytis cinerea: The Role of Calcipressin in Modulating Calcineurin Activity

In the gray mold fungus Botrytis cinerea the Gα subunit Bcg1 of a heterotrimeric G protein is an upstream activator of the Ca2+/calmodulin-dependent phosphatase calcineurin. In this study we focused on the functional characterization of the catalytic subunit of calcineurin (BcCnA) and its putative regulator calcipressin (BcRcn1). We deleted the genes encoding both proteins to examine their role concerning growth, differentiation and virulence. The ΔbccnA mutant shows a severe growth defect, does not produce conidia and is avirulent, while the loss of BcRcn1 caused retardation of hyphal growth and delayed infection of host plants, but had no impact on conidiation and sclerotia formation. Expression of several calcineurin-dependent genes and bccnA itself is positively affected by BcRcn1. Complementation of the Δbcrcn1 mutant with a GFP-BcRcn1 fusion construct revealed that BcRcn1 is localized in the cytoplasm and accumulates around the nuclei. Furthermore, we showed that BcCnA physically interacts with BcRcn1 and the regulatory subunit of calcineurin, BcCnB. We investigated the impact of several protein domains characteristic for modulation and activation of BcCnA via BcRcn1, such as the phosphorylation sites and the calcineurin-docking site, by physical interaction studies between BcCnA and wild-type and mutated copies of BcRcn1. Based on the observed phenotypes we conclude that BcRcn1 acts as a positive modulator of BcCnA and the Ca2+/calcineurin-mediated signal transduction in B. cinerea, and that both proteins regulate fungal development and virulence.


Introduction
The necrotrophic ascomycete Botrytis cinerea is able to infect a broad range of dicotyledonous host plants including economically important vegetable and fruit crops such as tomatoes, beans, grape vine or strawberries [1,2]. The filamentous fungus is both a serious pathogen causing economically significant losses in agriculture, and an ideal model organism to study pathogen-host interactions. The recently published genome sequences of the two different B. cinerea strains B05.10 and T4 [3] makes application of molecular techniques like knock-out approaches and high through-put genomic tools much easier and stimulates research on this plant pathogen.
The fine-tuned regulation of all processes of life in fungal plant and human pathogens, such as growth, development, morphological differentiation and host infection, is essential for survival [4][5][6][7][8].
Fungi are able to sense changes in the environment, and to respond appropriately by cellular changes on transcript and protein levels. Components of signaling pathways, such as mitogen-activated protein kinase (MAPK) cascades, the adenylate cyclase/cyclic AMP (cAMP)/protein kinase A cascade and the calcium/calcineurin pathway, which are highly conserved in yeast, filamentous fungi and even higher eukaryotes, control fundamental aspects of fungal growth, development and reproduction [9]. However, their targets and biological functions may differ and require study in each organism.
Calcineurin (CN) is a highly conserved Ca 2+ /calmodulinregulated type 2B protein phosphatase that is crucial for mediating cellular stress responses. Stress-induced transient increases of intracellular Ca 2+ -concentrations from either intracellular stores or extracellular sources are sensed by calmodulin (CaM), a small protein containing four EF-hand motifs for binding Ca 2+ ions. The Ca 2+ -CaM complex then activates various target proteins including protein kinases and the phosphatase CN. Functional CN consists in its inactivated form of two subunits, a catalytic (CNA) and a regulatory subunit (CNB) [10][11][12][13][14]. CN differs from the socalled CN-like phosphatases in its C-terminal extension which contains domains important for regulation of enzymatic activity (autoinhibitory domain, AID) and for the interaction with the two regulatory proteins CNB and CaM. Activation occurs when the cytosolic Ca 2+ level is increased and free Ca 2+ is bound by CaM and CNB [15]. Ca 2+ -bound CaM interacts with the CaM-binding domain of CNA, releasing the autoinhibitory domain (AID) from the substrate-binding pocket of the catalytic domain by a conformational change that relieves autoinhibition and leads to activation of the phosphatase and dephosphorylation of CN substrates [13].
One of the well known CN substrates is the conserved zincfinger transcription factor Crz1 which is translocated to the nucleus to regulate expression of target genes, e.g. those involved in cell wall integrity and ion homeostasis [16]. In Saccharomyces cerevisiae [10,[17][18][19] and Schizosaccharomyces pombe [20], the importance of CN has been established in various stress responses including high temperature, high concentrations of ions, cell wall stress and exposure to mating pheromone. In the human pathogens Candida albicans, Candida dubliniensis and Aspergillus fumigatus, the CN-dependent cascade controls cell wall integrity, stress resistance and response, morphogenesis, serum survival and virulence [4,[21][22][23][24][25][26]. For C. albicans it has also been shown that CN is a key mediator of Hsp90-dependent azole resistance [6]. In plant pathogenic fungi such as Aspergillus oryzae, Magnaporthe oryzae, Ustilago maydis and Ustilago hordei, a role for CN was not only suggested in hyphal growth, but also in environmental stress adaptation, e.g. at alkaline pH or high NaCl concentrations, during formation of conidia and appressoria, for sexual development, cell wall integrity, and virulence [27]. Furthermore, in S. cerevisiae and U. maydis CN was shown to be a putative antagonist of the protein kinase A (PKA) demonstrating a connection between the CN-and the cAMPdependent cascades [28,29].
In B. cinerea, the impact of CN on cellular processes was first studied by Viaud et al. [30] using an inhibitor of calcineurin, cyclosporine A (CsA). These studies indicated that CN is probably involved in regulation of hyphal morphology, formation of infection structures and secondary metabolism and led to the identification of 18 CN-dependent (CND) genes. Among the CND genes are the botrydial biosynthesis genes which were previously shown to be controlled by the Ga subunit Bcg1 [31,32] suggesting a connection between both signaling pathways. Later we could confirm this functional link by demonstrating that Bcg1 does not only control the adenylate cyclase/ cAMP/PKA cascade [33], but also the Ca 2+ /CN-signaling cascade [34,35]. Further research demonstrated for the first time in a phytopathogenic fungus that the Ga subunit Bcg1 acts as an upstream-activator of the B. cinerea Ca 2+ -signaling cascade via phospholipase C (BcPlc1) [35]. Detailed studies on the B. cinerea Ca 2+ -signaling cascade resulted in the identification of the first homolog of the yeast transcription factor Crz1 (CRaZy, calcineurin-responsive zinc finger transcription factor) in a filamentous fungus [34]. BcCrz1 is involved in hyphal morphology, conidiation, sclerotia formation, and stress response. Deletion mutants are hypersensitive to external Ca 2+ and react when subjected to hypo-osmotic stress. Infection of host plants is significantly retarded in these deletion mutants compared to the wild-type. Interestingly, most of the described phenotypic characteristics could be restored by the addition of Mg 2+ . The subcellular localization of a GFP-BcCrz1 fusion product in yeast cells depends on the Ca 2+ level and CN activity indicating that CN mediates nuclear translocation of BcCrz1 by its dephosphorylation. The expression of a set of Bcg1-and CN-dependent genes is also affected in Dbccrz1 mutants, confirming that this transcription factor acts downstream of CN in B. cinerea, as shown in S. cerevisiae [34].
Very little is known about regulation of CN activity in B. cinerea and other filamentous fungi. In yeast and mammals CN activity is regulated by a family of conserved proteins named ''regulators of calcineurin'' (RCAN in mammals, RCN in yeast and filamentous fungi) [36]. These proteins, also called calcipressins, have been reported to have both positive and negative effects on CN activity indicating a negative feedback regulation [12,37]. The Rcn1 homologs Cbp1 (calcineurin-binding protein) and RcnA of the basidiomycete Cryptococcus neoformans and the ascomycete Aspergillus nidulans, respectively, were shown to bind to CN and modulate CN activity [38,39]. In A. fumigatus cbpA deletion mutants displayed improved Ca 2+tolerance, decreased virulence and increased expression of the genes encoding the Ca 2+ /H + -exchanger VcxA, the chitin synthase A and CnaA [40].
The function of Rcn1-family proteins is determined by several highly conserved sequence motifs. The stimulatory effect of Rcn1 on CN activity requires phosphorylation of Rcn1 at the conserved FxISPPxSPP motif. In yeast and mammals, Mck1, a Gsk3 protein kinase, phosphorylates Rcn1 at the first serine residue after priming phosphorylation at the second serine residue by a mitogen-activated kinase (MAPK) [37,41]. The second conserved motif in Rcn1 is a key feature of CN regulation, the CN-specific PxIxIT docking site [42][43][44][45]. By screening peptide libraries comprising different PxIxIT motif variations of mammalian CN targets and artificial sequences, a high-affinity CN-binding peptide, PVIVIT, was found [46]. Interestingly, a PxIxIT docking site is also present in most studied Crz1 proteins. In yeast, substitution of the native Crz1 sequence with an allele containing the high-affinity PVIVIT sequence resulted in increased survival rates under high salt conditions, but higher sensitivity to alkaline stress. The increased affinity of Crz1 for CN by modification of the PxIxIT motif resulted in a constitutively dephosphorylated and hyperactive transcription factor [47].
In this study we focused on the further characterization of the Ca 2+ /CN-signaling pathway in B. cinerea by studying the impact of CN and the calcipressin homolog, BcRcn1, on growth, development and pathogenicity. We identified and manipulated the highly conserved motifs in BcRcn1 and performed interaction studies between BcCnA and BcCnB as well as wild-type and mutated forms of BcRcn1 to elucidate the mechanism of CN regulation in B. cinerea.

Identification of the B. cinerea Calcineurin A Ortholog
A BlastP similarity search in the protein database of B. cinerea B05.10 (Broad Institute) using the calcineurin A (CNA1) protein sequence of S. cerevisiae as query revealed bc1g_02606.1 (ABN58724.1) as a potential ortholog (Fig. S1). The predicted open reading frame of bccnA contains 1,805 bp including two introns of 60 and 50 bp. The deduced protein has a size of 564 aa and shows a high level of overall amino acid identity to catalytic CN subunits of different organisms, e.g. 98% (E value 0.0) to CNA of Sclerotinia sclerotiorum, 85% (0.0) to CNA of A. fumigatus, and 60% (0.0) to CNA proteins of mammals.
CNA consists of a catalytic domain homologous to other serine/ threonine protein phosphatases (PP2B family) and three regulatory domains at the C-terminus that distinguish CN from other phosphatase family members: the CNB-binding domain, the CaM-binding domain, and the autoinhibitory domain AID. The latter domain binds in the active site cleft in the absence of Ca 2+ / CaM and acts in concert with the CaM-binding domain to confer CaM regulation [13]. All domains were present in BcCnA (Fig. S1). Notably, CNA proteins of filamentous fungi share higher similarities with mammalian proteins than with those of S. cerevisiae. Furthermore, we were able to identify the most conserved residues probably responsible for interaction of the immunophilin/ immunosuppressant complexes cyclophilin A/CsA and FKBP12/ FK506 with the catalytic CN subunit via hydrogen bonds or van der Waals' interactions.

Deletion of bccnA Results in Severe Growth Defects and Loss of Sporulation and Virulence
It was reported by Viaud et al. [30] that deletion of the catalytic subunit of calcineurin is lethal in B. cinerea. However, recently we were able to replace bccnA by a resistance cassette (Fig. S2A) using osmotically stabilized selection media for protoplast regeneration. Two independent transformations revealed two mutants with the same drastic phenotype. Homokaryotic DbccnA mutants were obtained by single spore isolation, and the absence of the bccnA wild-type allele was confirmed by PCR (Fig. S2A).
Both DbccnA mutants showed a severe growth defect. After several weeks colonies reached diameters of about one to two cm, while the wild-type has a growth rate of one cm per day (Fig. 1A). In addition, a characteristic feature of the mutants is the genetic instability: their appearance changed within weeks from darkcolored to bright-colored colonies. Accordingly, several single spore isolates of the same (heterokaryotic) primary transformant led to colonies exhibiting different phenotypes concerning colony color, growth rate, configuration, and sector formation (Fig. 1B). However, finally the differences between isolates adjusted to one stable phenotype, and after several months all mutants formed small compact bright-colored colonies. In contrast to the wild-type that forms conidia in the light and sclerotia in constant darkness, the mutants formed sclerotia only very rarely after incubation for long periods in constant darkness, whereas conidia were never produced, neither in light nor in the dark.
Microscopic studies showed that DbccnA mutants exhibit a similar hyphal morphology as the B. cinerea wild-type strain B05.10 treated with the CN-inhibitor CsA (Fig. 1C). Addition of CsA to solidified standard medium or spore suspensions results in inhibition of vegetative growth and retardation of conidia germination, respectively. The hyphal morphology was significantly altered: instead of long, steady-growing hyphae in the untreated wild-type control, highly branching mycelia were obtained in the presence of CsA. This increase in hyphal branching was also observed for DbccnA mutants. The hyphae of DbccnA mutants were misshaped due to abnormal septation and more frequent branching in addition to the already described slow growth rate. Septation occurred more often in small, irregular intervals, and branches emerge through dichotomous splitting at the hyphal tips. The untreated wild-type nuclei are aligned like pearls in a chain, normally about three to six nuclei per septum section (Fig. 1D). Hyphae of the CsA-treated wild-type strain or the DbccnA mutant showed similar numbers of nuclei which were arranged almost equispaced. However, as hyphae of these samples are inherently more thickish, not all nuclei are located in a long line, but slightly staggered from each other (Fig. 1D).
To show whether the deletion of bccnA has an impact on virulence, wounded and unwounded primary leaves of bean plants were inoculated with agar plugs of the non-sporulating DbccnA mutant and the wild-type. In contrast to the wild-type, the DbccnA mutants did not penetrate the bean leaves, neither wounded nor unwounded, and were completely non-virulent (Fig. 1E).

Expression of a Truncated bccnA Copy Lacking the AID Domain
As the deletion of the whole bccnA gene resulted in severe defects in fungal growth, development and virulence that make further studies with these mutants extremely difficult, we generated mutants with a truncated bccnA gene, bccnA DAID , controlled by its native promoter but lacking the last 250 bp encoding the AID (Fig. S2A). The bccnA wild-type allele was replaced by the truncated bccnA DAID allele expecting a mutant with a dominantly active calcineurin A subunit. The transformation resulted in several independent hygromycin-resistant transformants, all showing the homologous integration at the bccnA locus (Fig. S2A). Mutants were purified by single spore isolation and screened for absence of the AID domain.
Three homokaryotic bccnA DAID mutants displayed severe growth defects on all tested media. On CM medium, they grew as small compact colonies with a significantly reduced growth rate of about 1.5 to 2.5 cm per week. However, in contrast to the DbccnA mutants, sclerotia formation (constant darkness) and conidiogenesis (light-dark conditions) were unaffected ( Fig. 2A). Expression of the truncated bccnA DAID gene resulted in reduced virulence: lesion formation on living bean plants was retarded, and in planta conidiation was only rarely observed (Fig. 2B). The retarded infection process was in correlation with the reduced growth rate on solid media, compared to the wild-type.
Surprisingly, the obtained phenotype of the bccnA DAID mutants resembled that of the wild-type after inhibition with CsA or FK506 [34]. To explain this unexpected phenotype, expression of the bccnA DAID copy in the mutant background was compared with that of the wild-type bccnA gene, both controlled by the same native promoter. Interestingly, the transcript level of the truncated gene was undetectable by northern blot analyses (data not shown). Therefore, the transcript level of bccnA DAID mRNA was determined by quantitative RT-PCR for all three independent transformants (T1-T3) (Fig. 3). Although the wild-type and truncated gene copies are controlled by the same native promoter, the mutants revealed a dramatically reduced transcript level (about 20 to 51% of the wild-type level) suggesting a low transcript stability of the truncated gene compared to the wild-type bccnA gene. The reason for low transcript stability might be due to the missing terminator of the truncated bccnA DAID gene copy. Therefore, a new vector containing the bccnA DAID allele with the native terminator was transformed into the B. cinerea wild-type. In several transformants, the wild-type bccnA gene was replaced by the bccnA DAID gene. However, these mutants showed a wild-type like phenotype, normal growth, conidiation and sclerotia formation indicating that the encoded BcCnA DAID protein does not act as a dominant active calcineurin.

Identification of the B. cinerea Calcipressin Ortholog
To further study the regulation of calcineurin activity we analyzed the role of calcipressin, which has been shown to have both activating and inhibiting function on calcineurin activity in S. cerevisiae [41].
We identified BC1G_07084.1 (CCD44521) as a potential ortholog of the S. cerevisiae regulator of calcineurin RCN1, based on highest similarity to the yeast protein. The predicted open reading frame of bcrcn1 contains no introns and encodes 255 aa. Standard protein blast revealed an overall amino acid identity of 92% (E value 3e-152) to the S. sclerotiorum ortholog, 48% (6e-55) to A. fumigatus CbpA, 34% (2e-12) to C. neoformans Cbp1 and 27% (8e-08) to S. cerevisiae RCN1. As all important regions, such as the Nterminal domain, the SP-linker domain for phosphorylation and the C-terminal domain, were present with a significant level of sequence conservation in the potential ortholog, we assumed that bc1g_07084.1 encodes the calcipressin ortholog BcRcn1 (data not shown).

BcRcn1 is Important for Vegetative Growth, but not for Sporulation or Sclerotia Formation
To analyze the function of Rcn1 in B. cinerea, bcrcn1 was disrupted. The bcrcn1 replacement fragment (Fig. S2B) was transformed into protoplasts of B. cinerea B05.10 resulting in three independent Dbcrcn1 mutants. For verification that the phenotype of Dbcrcn1 was due to the gene deletion, Dbcrcn1 mutant T3 was transformed with a construct comprising the entire bcrcn1 wild-type allele and its 59-and 39-non-coding regions. Three independent complementation transformants (bcrcn1com) were obtained for further studies.
Radial growth of Dbcrcn1 and bcrcn1com strains was analyzed on different solid media. As all deletion mutants on one hand, and all complementation strains on the other hand showed a similar phenotype, only one transformant per mutation was used for subsequent studies. Deletion of bcrcn1 resulted in formation of small, compact colonies ( Fig. 2A) and markedly decreased growth rates. The growth defect of Dbcrcn1 mutants was completely restored in the complementation mutants (data not shown). The bcrcn1 deletion strains were not impaired in conidiation, conidia germination or sclerotia formation.

High Osmolarity Improves the Growth of Dbcrcn1 and DbccnA Mutants
On media with high-osmolarity caused by addition of salts (NaCl) or sugars (sorbitol, glucose, sucrose), growth rates of Dbcrcn1 mutants were slightly improved. The most striking increase in growth rate was observed for Dbcrcn1 mutants on CM medium supplemented with NaCl (Fig. 4). The growth rates of DbccnA mutants were increased by addition of 1 M sorbitol, glucose or NaCl to solid complete medium. In contrast to the deletion mutant of the Ca 2+ /CN-dependent transcription factor BcCrz1 whose growth was almost fully restored by Mg 2+ ions [34], the growth rate of DbccnA mutant was not significantly improved by MgCl 2 supplementation (Fig. 4).

BcRcn1 is Essential for Full Virulence
In contrast to bccnA mutants which were completely avirulent ( Fig. 1), Dbcrcn1 mutants were able to penetrate, but spreading lesion formation was retarded. Complemented transformants (bcrcn1com) were fully pathogenic (Fig. 2B). However, using young non-sporulating mycelia of the Dbcrcn1 strain for inoculation, no disease symptoms developed, and conidiophores and conidia formed only on the agar plugs (data not shown).

BcRcn1 is Localized Basically in the Cytoplasm, but Accumulates at the Nuclei
To study the subcellular localization of BcRcn1, bcrcn1 was fused to codon-optimized gfp [48]. The Dbcrcn1 mutant transformed with the gfp-bcrcn1 fusion construct showed complete restoration of growth and pathogenicity, demonstrating full functionality of the fusion protein (data not shown). Fluorescence microscopy of this GFP-marked strain showed both, a homogenous distribution of a basic level of BcRcn1 in the cytoplasm, but also a stronger fluorescence around or in the nuclei, when monitored by simultaneous staining of the nuclei with Hoechst 33342 (Fig. 5). This localization pattern of GFP-BcRcn1 was shown under standard cultivation conditions (GB5 medium), but also under stress conditions, e.g. upon addition of 20 mM CaCl 2 or at alkaline pH (data not shown).

Motif Characterization of BcRcn1
Calcipressins have two phosphorylation sites (serine residues). In S. cerevisiae, their phosphorylation status decides on whether they have an inhibiting or activating function on CN [12,41]. Potential incubated for 48 h in liquid CM medium with and without 10 mg/ml of the calcineurin inhibitor cyclosporine A (CsA). DbccnA mycelia were incubated for 4 weeks in liquid CM medium. E: Pathogenicity assay: living bean leaves were inoculated with agar plugs with non-sporulating mycelia of the wild-type (WT) and the DbccnA mutant. Images were taken 3 and 6 days post infection (dpi). doi:10.1371/journal.pone.0041761.g001 phosphorylation sites were identified in the amino acid sequence of BcRcn1 by a multiple sequence alignment (Fig. 6A). The most conserved motif, the so-called SP-motif FxISPPxSPP, is displayed in the central span of about 20 aa. As previously described by Hilioti and Cunningham [41], the SP-linker region containing these two conserved serine-proline repeats is essential for phosphorylation of Rcn1 by the kinase Gsk3. In yeast and mammals, the protein kinase Gsk3 phosphorylates Rcn1 at the first serine residue in this domain after priming phosphorylation has taken place at the second serine residue. Beside the potential phosphorylation motif, a second conserved motif was found which is similar to the CN-specific PxIxIT docking site [42,45,47].
To study the function of these conserved motifs in more detail, we performed site-directed mutagenesis of bcrcn1 and expressed the mutated gene copies individually in the Dbcrcn1 background (Fig. 6B). First of all, we mutated the SPPPS motif to an APPPA motif which can probably no longer be phosphorylated. Furthermore, to show whether BcRcn1 and BcCnA interact and whether the suggested interaction site is functional, we generated an altered PxIxIT CN-docking site in BcRcn1. By this procedure, the motif was changed from PAIAVE to PVIVIT, a motif which was shown to have the highest affinity to CNA when different synthetic peptides were tested as substrates [46].
Vectors carrying the two mutated bcrcn1 copies under control of the native bcrcn1 promoter were transformed into the Dbcrcn1 strain. Interestingly, transformants carrying rcn1mut APPPA with the mutated phosphorylation sites showed exactly the same phenotype as the bcrcn1 deletion mutant or the wild-type when treated with the CN-inhibitors CsA or FK506 [34]. These data correlate with the suggestion that non-phosphorylated BcRcn1 functions as an inhibitor of CN activity.
In contrast, strains expressing the bcrcn1 allele with the proposed higher-affinity motif PVIVIT revealed a wild-type-like phenotype. A higher affinity of BcRcn1 to BcCnA seems to have no effect on vegetative growth or pathogenicity.  To experimentally prove whether the wild-type and the two mutated proteins of BcRcn1 interact with BcCnA, a split-ubiquitin based yeast two-hybrid approach was performed. Therefore, a bait-vector with bccnA encoding the catalytic subunit of CN and several prey vectors containing the bcrcn1 wild-type gene, one of the bcrcn1 mutant gene copies or the bccnB gene encoding the regulatory subunit of CN were cloned and transformed into yeast for interaction studies. As expected, BcCnA clearly interacts with its putative regulator BcRcn1 and the regulatory subunit BcCnB (Fig. 6C). The mutated proteins (BcRcn1mut APPPA , BcRcn1mut P-VIVIT ) also showed an interaction with BcCnA. BcRcn1mut PVIVIT revealed even significantly higher interaction rates with BcCnA than the wild-type protein (comparable to those of BcCnB) demonstrating that the mutated PVIVIT motif has indeed a higher affinity to CNA.

BcRcn1 Regulates BcCnA-dependent Gene Expression
To study the possible involvement of BcRcn1 in CN-dependent regulation of gene expression we studied the expression of previously identified CN target genes in both the wild-type and the Dbcrcn1 mutant by northern blot analyses (Fig. 7). Both strains were cultivated in liquid cultures in the presence or absence of the CN inhibitor CsA (as described by Viaud et al. [30]). All chosen CN-dependent genes showed similar expression patterns in the bcrcn1 deletion mutant and the CN-inhibited wild-type. Only some genes were up-regulated by inhibition of CN (e.g. CND8) and in the Dbcrcn1 mutant. Most of the CN-dependent genes were downregulated in the wild-type upon CsA treatment and also less expressed in the mutant. Interestingly, the expression level of bccnA itself was lower in the Dbcrcn1 mutant. For quantification of the bccnA transcript in the Dbcrcn1 background, qRT-PCR analyses were carried out confirming that the bccnA transcript level is significantly reduced to 42% in the Dbcrcn1 mutant compared to the wild-type (Fig. 3). In contrast, the bccnA transcript level was not significantly decreased in the Dbccrz1 mutant under the same conditions. However, the bcrcn1 expression level was reduced by addition of CsA to the wild-type indicating an interdependent transcriptional regulation of BcRcn1 and BcCnA.

Discussion
In this work, we studied the function and regulation of two components of the Ca 2+ /CN-dependent signaling pathway in the phytopathogenic fungus B. cinerea, calcineurin and its putative regulator, Rcn1. Previous CN inhibitory studies already demonstrated that CN is required for growth, morphogenesis, secondary metabolism and virulence in B. cinerea [30,49] and that Bcg1 and CN share a set of target genes, e.g. those for botrydial and botcinic acid [35].
Here we present the successful generation of bccnA knock-out mutants by using osmotically stabilized selection medium. Additionally, we generated a bccnA DAID mutant expressing a truncated bccnA gene copy without the autoinhibitory domain. Furthermore, we identified the only member of the calcipressin family in B. cinerea, BcRcn1. With these mutants available, we were able to study the complex Ca 2+ /CN-dependent regulatory system by comparing the characteristic features of these mutants regarding growth, differentiation, expression pattern, and virulence with those of the wild-type. Calcineurin has an Impact on Growth and Development of B. cinerea Previous attempts to delete the two CN subunits failed, suggesting that both genes are essential in B. cinerea [30]. The most obvious phenotype of the DbccnA mutants was the severe growth defect. The hyphal morphology was heavily modified and comparable to that obtained for the wild-type treated with CsA. DbccnA mutants were absolutely avirulent on wounded and unwounded bean plants. These data fit to previous penetration assays on onion epidermis which showed that reduced CN activity in wild-type conidia prevented the formation of infection structures in B. cinerea strain T4 [30]. Furthermore, deletion of Mammalian and yeast as well as the synthetic peptide sequences derive from [45]. Black letters/stars: highly conserved residues; dark grayish letters: amino acids with strongly similar properties; light gray letters: weakly similar properties. B: Phenotype of mutated strains: growth analyses with wild-type B05.10 (WT), Dbcrcn1, Dbcrcn1 + bcrcn1mut APPPA , Dbcrcn1 + bcrcn1mut PVIVIT after 1 week on described media. Inoculation was made with agar plugs of equal size (5 mm). C: Yeast strain NMY51 was transformed with the bait vector containing the bccnA gene and different prey vectors: + control: pAI-Alg5, -control: pPR3-N, BcRcn1/BcCnB/ BcRcn1mut APPPA /BcRcn1mut PVIVIT : pPR3-N_BcRcn1/_BcCnB/_BcRcn1mut APPPA /_BcRcn1mut PVIVIT . Strains were incubated in different concentrations (non-diluted, 1:10, 1:100, 1:1000-diluted) on SD-L-W (selection for vectors) and on SD-L-W-H-adenin + X-Gal-plates (interaction of tested proteins). Percentage of growth was calculated by counting of colonies in comparison to number of colonies on SD-L-W. The test was performed three times. doi:10.1371/journal.pone.0041761.g006 bccnA or inhibition of CN by CsA probably results in inactivation of the downstream-acting transcription factor BcCrz1 that was shown to have an impact on virulence. Similarly to DbccnA mutants, bccrz1 deletion mutants were unable to penetrate bean leaves when agar plugs were used for inoculation [34].
A characteristic feature of bccnA mutants is their genetic instability, for their morphology changed within weeks from dark-colored to brightly-colored colonies. All formed sectors were homokaryotic deletion mutants, indicating that genetic instability is due to a process of suppression of the mutation. The extremely low growth rate was slightly improved in media with high osmolarity indicating that the mutant is affected in cell wall stability. The other characteristic features, the altered hyphal morphology and septa formation, support the suggestion that the cell wall composition of bccnA mutants might differ from those in the wild-type. However, high-osmolarity conditions did not fully restore wild-type growth and morphology indicating that the mutant has additional defects. As the Dbcrcn1 mutants also react on these conditions with an improvement of growth we conclude that the Ca 2+ /CN-signaling pathway influences the cell wall composition in general.
In yeast cells, Mg 2+ starvation induces the translocation of the transcription factor Crz1 from the cytoplasm into the nucleus. Mg 2+ depletion causes enhanced cellular Ca 2+ concentrations which activate the CN/Crz1-pathway [50]. Interestingly, in B. cinerea, the role of Mg 2+ ions seems to be different. Whereas addition of external Mg 2+ almost fully restored growth rate, conidiation and penetration, but not sclerotia formation of Dbccrz1 mutants [34], addition of external Mg 2+ failed to overcome the growth or differentiation defects of bccnA mutants leading to the conclusion that Mg 2+ -dependent metabolism acts downstream of CN.
Surprisingly, the deletion of the AID caused a reduced transcript level of bccnA that was confirmed by qRT-PCR. The bccnA DAID vector was cloned without the 39-UTR to avoid loss of the resistance cassette by homologous recombination at the terminator region. Transcript instability due to the missing terminator might be the reason for the unexpected decrease of the transcript level of bccnA in the bccnA DAID mutant. Subsequently, the mutant shows a phenotype resembling that of the Dbcrcn1 mutant and the CsA-inhibited wild-type. On the other hand, the less severe phenotype of the bccnA DAID mutant compared to the DbccnA mutant must be due to a partially functional BcCnA DAID protein as the bccnA DAID allele is the only available bccnA gene copy in these mutants. In addition, transformants carrying the bccnA DAID allele with the native terminator showed a wild-typelike phenotype. In conclusion, deletion of the AID did not result in the expected dominantly active calcineurin.
It might be possible to generate constitutively active forms of CN by deleting both, the calmodulin binding domain and the AID, as shown by O'Keefe et al. [51]. The regulatory role of the C-terminal domains of CNA in mammals was explored by Wang et al. [52,53]. They showed that not only the AID, but two distinct segments are responsible for inhibiting the enzymatic activity by Figure 7. Expression studies in the wild-type and the Dbcrcn1 mutant strain. The wild-type B05.10 and the Dbcrcn1 mutant were grown for 72 h in liquid culture (SISLER medium) and then moved to fresh cultures without (2) or with (+) 20 mg/ml of the calcineurininhibitor cyclosporine A (CsA). The northern blot was hybridized to radioactively labeled probes of several BcCnA/BcCrz1-dependent genes (CND-genes [30], CND5 = bcbot1 as a botrydial biosynthesis gene, bcboa4 as a botcinic acid biosynthesis gene), bccnA and bcrcn1. Loading controls: bcactA and rRNA. doi:10.1371/journal.pone.0041761.g007 binding to the catalytic site and to the CNB binding helix. Future experiments will show if the deletion of both domains as well as site-directed mutagenesis on specific residues result in a dominantly active CnA in B. cinerea.

BcRcn1 is Accumulated Around the Nuclei
Little is known about the subcellular localization of components of the Ca 2+ /CN-signaling pathway, such as the CN subunits, or the potential CN regulator Rcn1, in filamentous fungi. For the transcription factor Crz1 it has been shown that it is translocated to the nucleus upon its dephosphorylation by CNA [26,34]. A. fumigatus was the first filamentous fungus in which the catalytic subunit of calcineurin was shown to localize at the hyphal septum where it complexes with CnaB for normal CN function [54].
In this work, we investigated localization of BcRcn1 by expressing a GFP fusion construct. The GFP-BcRcn1 fusion protein was distributed homogenously in the cytoplasm, as described for A. nidulans [39]. However, partial accumulation of BcRcn1 was also observed around or in the nuclei under all tested conditions. This is the first indication that BcRcn1 may modulate CN activity in the cytoplasm in near proximity to or in the nucleus. Its modulation can take place either by specific regulation of CN activity to mediate the Ca 2+ signaling response, or by its targeting to the nucleus for further facilitation of substrate recognition. Future localization studies of BcCnA would give further insight into the complex interaction network of these diverse signaling components.

BcRcn1 is a Regulator of Calcineurin Activity
To show whether CN activity in B. cinerea is regulated in a similar way as in yeast and mammals, we studied the role of calcipressin, BcRcn1, on Ca 2+ /CN-dependent signaling. Deletion of bcrcn1 resulted in reduced hyphal growth as previously reported for the corresponding mutant in S. cerevisiae [12] and the human pathogen A. fumigatus [40].
The Dbcrcn1 phenotype showed a strong similarity to the wildtype treated with CN inhibitors suggesting that BcRcn1 acts as a positive regulator of CN activity and the Ca 2+ /CN-dependent pathway in B. cinerea. In contrast, Rcn1 functions as an inhibitor as well as an activator of CN signaling in yeast [12,41].

Expression Studies Confirm the Activating Role of BcRcn1
To confirm that BcRcn1 is an activator of CN activity in B. cinerea, we studied the expression of CN-dependent genes in the wild-type and bcrcn1 mutants, treated or non-treated with CsA. Deletion of bcrcn1 resulted in the same effect as inhibition of CN in the wild-type. Among the CN target genes were also those involved in secondary metabolism, like botrydial biosynthesis genes, e.g. bcbot1 (CND5) [30,55]. While the inhibition of CN by CsA did not affect the level of bccnA transcript, the latter was significantly decreased in the Dbcrcn1 mutant independently of the presence of CsA. On the other hand, the bcrcn1 transcript itself was differentially expressed in a CN-dependent manner, as a lower transcript level was observed after treatment with CsA. Furthermore, we found a putative CDRE (calcineurin-dependent response element) in the promoter region of bcrcn1 at position 746 bp upstream of the start codon. The identified motif GTGGCTGGG fits in the model element G[T/G]GGC[T/ A]G[T/G]G predicted as binding motif for Crz1 in A. nidulans [56]. These results indicate an interconnection of BcCnA and BcRcn1 at transcriptional level via binding of BcCrz1 to the promoter of bcrcn1 similarly to A. nidulans. Dependent on the environment, active BcCnA enhances bcrcn1 expression via BcCrz1. On the other hand, BcRcn1 positively regulates bccnA expression.
In A. fumigatus, CbpA is probably also involved in the transcriptional regulation of CN, while Cbp1 in C. neoformans did not significantly affect the transcriptional level of cna1 [40,57]. In S. cerevisiae, neither the deletion nor the overexpression of rcn1 affected the expression of an epitope-tagged Cna1-Myc fusion under non-signaling conditions. However, under high-calcium conditions, the expression of the cna1 fusion consistently declined to lower expression levels in Drcn1 mutants [12]. In contrast, the co-regulation of all CND genes in B. cinerea indicated an exclusively activating function of BcRcn1 on the CN signaling cascade.

Modulatory Effect of BcRcn1 through Direct Interaction with BcCnA
In this study we showed that the catalytic subunit BcCnA not only physically interacts with its regulatory subunit, BcCnB, but also with BcRcn1 further indicating the role of calcipressin as regulator of calcineurin. To study the role of putative functional domains of BcRcn1, different mutations were introduced into the bcrcn1 gene. Mutation of both putative phosphorylation sites in BcRcn1 did not result in any change of the interaction rate with BcCnA, but in a phenotype resembling the knock-out of bcrcn1 or the CsA-treated wild-type [34]. Apparently, the activating effect of BcRcn1 on the Ca 2+ signaling cascade is missing in this mutant, confirming our model of BcRcn1 as an activator of CN activity. In yeast, an activating function of Rcn1 was suggested for the doublephosphorylated form, whereas dephosphorylation causes inhibition of CN [37]. Mutations of only one of the two phosphorylation sites could help to elucidate the role of the phosphorylation status in modulating CN activity.
In addition, we were able to demonstrate that the PAIAVE (PxIxIT)-sequence element in BcRcn1 is a functional interaction domain with CNA. Remarkably, the replacement of the potential PxIxIT-CNA-binding motif in BcRcn1 by a motif with postulated higher affinity to CNA [46] resulted indeed in much higher interaction rates up to the same strength as for the two CN subunits, BcCnA and BcCnB. However, this much higher affinity did not result in any visible phenotype: mutants expressing the high-affinity BcRcn1 protein showed wild-type-like characteristics.
There are reports on the presence of additional interaction sites in CN substrates, such as those proposed for NFAT transcription factors in mammals, the analogue CN target of the yeast and filamentous fungal transcription factor Crz1, or on alternative mechanisms for CN-substrate interaction (reviewed in [45,47]). There are indications that the LxVP-motif which was described as additional motif for interaction with CNA in these reports, is also present in BcRcn1 (data not shown), which may be responsible for interaction with CNA as well. Therefore further studies are needed to unravel the mechanism of regulation of CN by BcRcn1 in B. cinerea.

Conclusions and Perspectives
Despite the high level of sequence conservation of components of signaling pathways between yeasts, filamentous fungi and higher eukaryotes, several differences have been found regarding the function of these signaling cascades in processes of life, such as vegetative growth, development, reproduction, and virulence. Nothing was known so far about mechanisms of regulation of CN in phytopathogenic fungi. In this work we demonstrated on transcriptional and protein levels that the putative CN regulator BcRcn1 is an activator of CN activity in B. cinerea under all tested conditions. This is the first report of calcipressin acting as a positive regulator of CN and CN target genes, such as those involved in secondary metabolism. Future studies will show under which conditions the Ca 2+ /CN-dependent signaling cascade is activated in B. cinerea and whether BcRcn1 has an inhibiting function on CN as well under those conditions.
Mutants containing a GFP-BcCrz1 fusion protein will be used for studying factors affecting nuclear translocation of this transcription factor.
Yeast cells were grown either in complete YP(A)D medium consisting of 2% glucose, 2% peptone, (0.004% adenine sulfate) and 1% yeast extract or in minimal SD medium containing 2% glucose, 0.67% yeast nitrogen base, and the amino acids required by the strains. Solid media contained 2% agar. S. cerevisiae strains were transformed by the lithium acetate procedure as described previously [60]. For maintenance of plasmids, yeast transformants were pre-cultured in selective media.
Escherichia coli strains TOP10F' (Invitrogen, Groningen, Netherlands) and E. coli XL1-Blue (Agilent Technologies) were used as hosts for plasmid construction, site-directed mutagenesis and propagation.

Germination Assays and Microscopic Analyses
Fluorescence microscopy was carried out after incubation of germinating conidia for 16-24 hours post inoculation (hpi) in liquid GB5 medium supplemented with 2% glucose and 0.132 g/l (NH 4 ) 2 HPO 4 . To study hyphal morphology, the B. cinerea strains were grown in liquid CM supplemented with cyclosporine A or CaCl 2 as indicated. After incubation for 24-48 hpi at 20uC, the colonies were incubated for 5 min in 1% (wt/vol) Calcofluor White solution. For detection of nuclei, nucleic acids were stained using the fluorescent dye Hoechst 33342; 10 ml freshly prepared Hoechst solution, prepared according to Kangatharalingam et al. [61], were added to the germinated conidia prior to microscopy. Fluorescence and light microscopy was performed with an Axio Imager.M2 microscope (Carl Zeiss MicroImaging GmbH, Jena, Germany). Differential interference contrast (DIC) microscopy was used for bright field images. Staining of hyphae by Hoechst 33342 and Calcofluor White was examined using the filter set 49 DAPI shift free (excitation G 365, beam splitter FT 395, emission BP 445/50) and GFP fluorescence with filter set 38 (excitation BP 470/40, beam splitter FT 495, emission BP 525/50). Images were captured with an AxioCam MRm camera and analyzed using the Axiovision Rel 4.5 software package (both Carl Zeiss MicroImaging GmbH, Jena, Germany).

Pathogenicity Assays on Bean Plants
Infection assays were performed with conidia from 7 to 10-dayold PDAB agar cultures as described previously [32]. In addition, non-sporulating agar plugs taken from 3-day-old CM agar cultures were used to inoculate primary leaves of Phaseolus vulgaris on the adaxial surface. The infected plants were incubated in a plastic propagator box at 20uC under natural illumination. Disease symptoms were scored until 10 days after inoculation.

Standard Molecular Methods
Fungal genomic DNA was isolated as described previously [62]. Plasmid DNA was isolated using a plasmid DNA preparation kit (Genomed, Bad Oeynhausen, Germany). For Southern blot analyses, the fungal DNA was transferred to Hybond N + filters (Amersham Biosciences, Freiburg, Germany) after digestion with restriction enzymes and size separation on a 1% agarose gel according to the method of Sambrook et al. [63]. Hybridization was carried out in 6x SSC (1x SSC is 0.15 M NaCl plus 0.015 M sodium citrate), 5x Denhardt's solution, 0.1% SDS, and 50 mM phosphate buffer, pH 6.6, at 65uC in the presence of a randomprimed [a-32 P]dCTP-labeled probe. The membranes were washed once (2x SSPE [1x SSPE is 0.18 M NaCl, 10 mM NaH 2 PO 4 , and 1 mM EDTA {pH 7.7}], 0.1% SDS) before being exposed to autoradiographic film. Total RNA was isolated from mycelial samples using the Trizol procedure (Invitrogen, Groningen, The Netherlands). Samples (25 mg) of total RNA were transferred to Hybond N + membranes after electrophoresis on a 1% agarose gel containing formaldehyde, according to the method of Sambrook et al. [63]. Blot hybridizations were done in 0.6 M NaCl, 0.16 M Na 2 HPO 4 , 0.06 M EDTA, 1% N-lauroylsarcosine (Sigma-Aldrich, St. Louis, MO), 10% dextran sulfate (Eppendorf AG, Hamburg, Germany), 0.01% salmon sperm DNA, pH 6.2, as described for Southern blots; 1 mg of total RNA was taken for cDNA synthesis using the oligo(dT)12-18 primer and SuperScript II reverse transcriptase (Invitrogen, Groningen, The Netherlands) according to the manufacturer's instructions. PCR mixtures contained 25 ng DNA, 5 pmol of each primer, 200 nM concentrations of deoxynucleotide triphosphates, and 1 unit of BioThermDNA polymerase (GeneCraft GmbH, Lüdinghausen, Germany). The reactions started with 4 min at 94uC, followed by 35 cycles of 1 min at 94uC, 1 min at 56 to 65uC, and 1 min at 70uC, and a final 10 min at 70uC. PCR products were cloned into pCRH2.1-TOPOH (Invitrogen, Groningen, The Netherlands). For sequence analyses, Lasergene v6 software (DNAStar, Madison, WI) was used. Quantitative real-time PCR was carried out using a one-tenth dilution of the cDNA template in a MyiQ2 Two-Color Real-Time PCR Detection system (Bio-Rad, Hercules, CA, U.S.A) with the Bio-Rad iQ SYBR Green supermix. The genes encoding actin A, glyceraldehyde-3-phosphate dehydrogenase and tubulin showed the same expression pattern in the wild type strain, as well as in the mutant strains used in this work. All three genes were used to normalize the cDNA amount in each sample using the primers 1/2 (for actin A), 3/4 (glyceraldehyde-3-phosphate dehydrogenase) and 5/6 (tubulin). All primers used in this study are listed in the Table S1. To study the expression of bccnA in the bccnA DAID mutants, primers 7/8 were used. The annealing temperature in every PCR was in a range from 58uC to 62uC, while the time extension was always 20 s. To avoid genomic DNA amplification, RNA samples were treated with DNAseI (Promega) and every pair of primers was designed in such way that one of them hybridized to an exon-exon split. Therefore, for each gene, the PCR efficiency was between 90 and 110%. The relative expression of bccnA was calculated following the DDCt (cycle threshold) method, from the mean of two different determinants of Ct values.

Vector Cloning
For construction of the vector pDbccnA-hyg R , the plasmid pOliHP [64], carrying the E. coli hygromycin B phosphotransferase gene hph under control of the A. nidulans oliC promoter, was used as basal vector (Fig. S2A). The gene flanks were amplified by PCR with primers, derived from the genomic sequence of B. cinerea B05.10 (ABN58724), containing artificial restriction sites for further cloning. A 1.4-kb PCR fragment was amplified from the bccnA 59-region using the primers 9&10, and a 0.6-kb fragment of the 39-untranslated region was generated as second flank using primers 11&12. Both PCR products were cloned into pCRH2.1-TOPOH (Invitrogen, Groningen, The Netherlands), isolated with KpnI/SalI and HindIII/EcoRI, respectively, and then cloned into the corresponding restriction sites of pOliHP, creating pDbccnAhyg R . Prior to transformation, the whole replacement cassette was isolated by restriction with KpnI and EcoRI. Hygromycin Bresistant transformants were analyzed by PCR for homologous integration using the primers 13&14 and 15&16. Putative knockout mutants were purified by several rounds of single spore isolation and screened by PCR for the absence of the bccnA wildtype allele using the primers 17&18 (Fig. S2A). Southern blot analyses of these transformants was not feasible as mutants were so slow growing that genomic DNA was obtained from two to three week-old liquid cultures. The by this procedure gained DNA was so instable and difficult to obtain that just PCR was performable, but the quality was not good enough to perform restrictions for Southern blot analyses.
For expression of a truncated version of bccnA, pDbccnA-hyg R was used as basis for cloning and for mediation of homologous integration of the truncated gene at the bccnA locus. The gene including the promoter sequence was amplified by PCR in two steps using primers which contain restriction sites if necessary for further cloning. A 1.6-kb fragment including 1 kb of the 59 noncoding region and 0.6 kb of the bccnA open reading frame by using the primers 19&20 was amplified (Fig. S2A). The second part of the gene lacking the last 250 bp was amplified by using the primers 21&22, yielding a PCR fragment with a size of 1.0 kb. Both PCR products were cloned into pCRH2.1-TOPOH, separately, then sequenced and finally isolated with KpnI/BglII and BglII/SalI, respectively. Both fragments were cloned into the KpnI and SalI restriction sites of pDbccnA-hyg R , creating pbccnA-D AID. By this approach the former 59-flank of bccnA was replaced by the truncated version of bccnA which functions as new 59-flank for homologous integration at the bccnA locus, and therefore the wild-type allele is replaced by the truncated version. For transformation of B. cinerea B05.10 wild-type, pbccnA DAID was linearized with SmaI, cutting behind the bccnA-39-flank. The hygromycin resistant transformants were analyzed by PCR for homologous integration using the primers 13&14 (59-flank) and 15&16 (39-flank). Putative bccnA DAID mutants were purified by single spore isolation and screened by PCR for the absence of the bccnA wild-type allele using the primers 17&18. For Southern blot analysis, the genomic DNA of bccnA DAID mutant and wild-type strains was digested with KpnI, and hybridized with the bccnA-39flank after electrophoresis and blotting. The replacement of the wild-type allele by the truncated form (bccnA DAID ) leads to a hybridizing fragment of approx. 4 kb, in comparison to the wildtype fragment which has a size of 2 kb. The shift is due to the insertion of the hygromycin resistance cassette between bccnA and the 39-non-coding region of bccnA (Fig. S2A).
The knock-out construct for deletion of bcrcn1 was obtained using the homologous recombination system in yeast [65]. About 1 kb of the 59-and the 39-flanking regions of bcrcn1 (CCD44521) were amplified from genomic DNA using the primer pairs 23&24 and 25&26, respectively (Fig. S2B). These primers are composed of one part for binding to the genomic DNA and the other part with homologous regions to the resistance cassette or the shuttle vector pRS426, respectively [66]. The hygromycin resistance cassette of pOliHP [64] containing the hph gene of E. coli under control of the oliC promoter of A. nidulans was amplified by PCR using primers 27&28. After restriction of pRS426 using EcoRI/ XhoI transformation into yeast strain FY834 [67] with all three PCR fragments together was carried out. Positive transformants were selected on SD-uracil plates. Total DNA from uracilprototrophic yeast colonies was isolated with the GeneJET TM Plasmid Miniprep Kit (Fermentas), using glass beads for lysation of yeast cells in the first step. The complete recombined replacement fragment for deletion of bcrcn1 was re-amplified from yeast DNA by PCR using the primers 15&18 and taken for transformation of B. cinerea B05.10. Successful deletion was verified via diagnostic PCR using primers 29&30 for integration of the hph R cassette at the 59 region, and 14&31 for the 39 region. Absence of a wild-type copy of bcrcn1 was proved by PCR with primers 32&33. The complete absence of any additional ectopic integration was confirmed by Southern blot analysis after digestion of genomic DNA with BamHI and hybridization with the bcrcn1-59-flank. The replacement of the wild-type bcrcn1 allele by the hph R cassette leads to a hybridizing fragment of approx. 1.4 kb, in comparison to the wild-type fragment which has a size of more than 8 kb (Fig. S2B).
For complementing the Dbcrcn1 mutant, a fragment containing about 1 kb of the native bcrcn1 promoter, the complete open reading frame, and about 500 bp of the bcrcn1 terminator was amplified by PCR using primers 34&35. The PCR product was cloned into pCRH2.1-TOPOH, resulting in pTOPO_rcn1com, sequenced and finally isolated with SpeI/XhoI. The fragment was cloned into pDbcniaD-nat R , containing the nourseothricin resistance cassette, with nat1 from Streptomyces noursei under the control of the A. nidulans trpC promoter, creating pbcniaD_nat R _rcn1com. This plasmid also contains the non-coding 59 and 39 flanking regions of bcniaD, encoding the nitrate reductase, to mediate homologous integration of the complementation construct at the bcniaD locus [35]. Prior to transformation the plasmid was linearized by restriction with SacI/PvuI. The resulting fragment of 5.9 kb was used for transforming mutant Dbcrcn1 T3. Presence of the complete bcrcn1 complementation construct at the bcniaD locus was confirmed by diagnostic PCR using primer pairs 36&37, 38&39 and 40&41 (data not shown). As deletion of bcniaD is not essential for the complementation, just one round of single spore isolation was done to stabilize the integration event. The wild-type copy of bcniaD is still present in the complementation mutants.
The gfp-bcrcn1 fusion fragment was generated using the homologous recombination system in yeast [65]. The coding region of bcrcn1 was amplified using the primers 42&43 that contain overlapping sequences homologous to the glucanase terminator and gfp (encoding codon-optimized GFP [48]), respectively, of the vector pNAN-OGG [68]. This vector contains flanks mediating the replacement of the gene encoding the nitrite reductase, and by this ensuring integration at a known locus. After amplification, the PCR product and the NotI-digested plasmid pNAN-OGG were co-transformed into the uracil-auxotrophic S. cerevisiae strain FY834. DNA of pooled yeast colonies was isolated as described above and transformed into E. coli. Plasmid DNA from single colonies was isolated and sequenced, resulting in the final vector pNAN-OG bcrcn1 G. The whole transformation construct, containing the bcniiA flanks, nat R and the gfp-bcrcn1 fusion construct flanked by the oliC promoter and the glucanase terminator was transformed after restriction of pNAN-OG bcrcn1 G with SacII. Integration of the transformation construct into WT: B05.10 and Dbcrcn1 T3 was analyzed by diagnostic PCR (data not shown).
For generation of mutated versions of bcrcn1, pTOPO_rcn1com which was cloned for complementation of Dbcrcn1 was used as basic vector. Site-directed mutagenesis was carried out as described by the manufacturer using the QuikChangeH II Site-Directed Mutagenesis Kit (Agilent Technologies). By performing the PCR using the primers 44&45 (APPPA-mutation) and 46&47 (PVIVIT-mutation), respectively, via pTOPO_rcn1com as template, the mutation was inserted in bcrcn1. As for the complementation, the constructs were then cloned into the destination vector pDbcniaD_nat R resulting in pDbcniaD_nat R _rcn1mut APPPA and pDbcniaD_nat R _rcn1mut PVIVIT and subsequently transformed into mutant Dbcrcn1 T3. Transformants were further on analyzed by diagnostic PCR (data not shown).

Transformation of B. cinerea
Protoplasts were generated using a mixture of Glucanex 200G (Novozymes, Denmark), lysing enzyme (Sigma Aldrich, St Louis, MO, USA), and Yatalase (Takara Bio Ins, Shiga, Japan) added to 15-20 mg of the linearized vector, and transformed according to Siewers et al. [55]. Resistant colonies were transferred to plates containing CM agar complemented with 70 mg/ml of hygromycin B (Invitrogen, San Diego, CA, USA) or 70 mg/ml of nourseothricin (Werner-Bioagents, Jena, Germany). Single conidial isolates were obtained by spreading conidial suspensions on GB5 plates containing 70 mg/ml of hygromycin B or nourseothricin. Single, germinated conidia were transferred individually to new plates containing the selection marker. Homokaryotic transformants were generated by several rounds of single spore isolation.

Interaction Studies
Cloning of interaction vectors, yeast transformation and quantification of interaction rates were performed as described in the manufacturer's manual of the DUALSystems Pairwise interaction Kit. The pDHB1-vector was used as bait-the pPR3-N-vector as prey-basis-vector (DUALSystems Biotech). cDNA of B. cinerea was used as template for a PCR with primers containing additional SfiI restriction sites. The pDHB1_cnA vector was cloned using primers 48&49 for PCR, digestion of the PCR-product with SfiI and ligation into the linearised pDHB1-vector using kanamycin for E. coli selection. For pPR3-N_cnB and pPR3-N_rcn1 primer pairs 50&51 and 52&53 were used, respectively, followed by restriction, ligation into pPR3-N and selection of transformed E. coli with ampicillin resistance. Mutated versions (pPR3-N_rcn1mut APPPA & pPR3-N_rcn1mut PVIVIT ) were generated using the mutated TOPO-vectors as template. Interaction studies were performed on SD-L-W-(selection for plasmid existence), SD-L-W-H-adenin-(quantification of interaction rates) or on SD-L-W-H-adenin+X-Gal-plates (containing 80 mg/l X-Gal for visualization of interaction).

Supporting Information
Fig. S1 Multiple alignment and domain structure of calcineurin A protein sequences from yeast, filamentous fungi and mammals. Shown are B. cinerea BcCnA (ABN58724), S. sclerotiorum Cna1 (ABB13418), A. fumigatus CnaA (XP_753703), S. cerevisiae calcineurin A1 (AAA34465), R. norvegicus (CAA40398) and H. sapiens calcineurin A catalytic subunit (AAB23769). The proteins contain catalytic phosphatase domains at the N-termini, the calcineurin B (CNB)-binding domain, the calmodulin (CaM)binding domain, and the autoinhibitory (AID) domain (reviewed in [13]). Identical residues are indicated in black, similar amino acids in gray. Putative binding domains via hydrogen bonds or van der Waals' interactions with the immunophilin/immunosuppressant complexes cyclophilin A/CsA and FKBP12/FK506 as described in [69][70][71] are indicated with colored fonts (green: CsA, blue: FK506, violet: FK506 and cylophilin A, red: CsA and FK506, orange: cylcophilin A). (TIF) Fig. S2 Gene replacement of bccnA, deletion of the autoinhibitory domain of bccnA and gene replacement of bcrcn1 using the hph resistance cassette. All primers used for cloning of the replacement vectors and the diagnostic PCR analyses for homologous integration are indicated and further described in the materials and methods section. A: Physical maps of bccnA (wild-type, WT), the DbccnA-hyg R and the bccnA DAID locus. The wild-type B. cinerea B05.10 was transformed with the bccnA replacement fragment (upper panel) or the truncated bccnA gene (lower panel) by homologous recombination and insertion of the hph resistance cassette. The last 250 bp (autoinhibitory domain, AID) of bccnA was replaced creating the bccnA DAID mutant. Diagnostic PCRs of two different DbccnA mutants (H5 and N1) and three bccnA DAID mutants (T1, T2, T3) in comparison to the WT are depicted just as the Southern blot analysis of bccnA DAID mutants. 59 and 39 prove homologous integration of the hph R cassette at the bccnA locus. WT shows the absence of the bccnA wild-type allele (the reverse primer lies in the AID domain). B: Physical maps of the bcrcn1 wild-type (WT) gene locus and the Dbcrcn1 gene locus. The bcrcn1 gene was replaced by the hph R cassette in the opposite direction. Diagnostic PCR and Southern blot analysis of Dbcrcn1 mutants (T3, T6, T7) in comparison to the WT are depicted as well. 59 and 39 prove homologous integration of the hph R cassette at the bcrcn1 locus. WT shows the absence of the bcrcn1 wild-type allele. (TIF)