Extracellular calcium triggers unique transcriptional programs and modulates staurosporine-induced cell death in Neurospora crassa

Alterations in the intracellular levels of calcium are a common response to cell death stimuli in animals and fungi and, particularly, in the Neurospora crassa response to staurosporine. We highlight the importance of the extracellular availability of Ca2+ for this response. Limitation of the ion in the culture medium further sensitizes cells to the drug and results in increased accumulation of reactive oxygen species (ROS). Conversely, an approximately 30-fold excess of external Ca2+ leads to increased drug tolerance and lower ROS generation. In line with this, distinct staurosporine-induced cytosolic Ca2+ signaling profiles were observed in the absence or presence of excessive external Ca2+. High-throughput RNA sequencing revealed that different concentrations of extracellular Ca2+ define distinct transcriptional programs. Our transcriptional profiling also pointed to two putative novel Ca2+-binding proteins, encoded by the NCU08524 and NCU06607 genes, and provides a reference dataset for future investigations on the role of Ca2+ in fungal biology.


INTRODUCTION
Calcium (Ca 2+ ) is crucial for diverse processes in fungi, including tip growth, virulence, circadian rhythms, nutrient sensing, cell wall regeneration, chemotropic interaction and cell death [1]. Ca 2+ and cell death are intimately connected, especially because of the vital role of Ca 2+ as an intracellular messenger affecting numerous signaling pathways that determine cell fate. Therefore, it is not surprising that cells commonly respond to cell death stimuli with rises in the intracellular levels of this divalent ion. On one hand, these transient or stable modifications in the levels of Ca 2+ can operate as survival signals, leading to the activity of anti-death proteins or stimulating transcriptional defence responses. On the other hand, the loss of Ca 2+ homeostasis can behave as a pro-death signal, since in some systems overload or a simple perturbation of the distribution of the ion within storage organelles is sufficient to trigger cell death [2]. In yeast species like Saccharomyces cerevisiae, Candida albicans and Cryptococcus neoformans, cytosolic Ca 2+ increases are a common feature of the cellular response to a number of antifungal stresses, including pheromone [3], the plant essential oils carvacrol [4] and eugenol [5], amiodarone [3,6], and ER stressinducing agents like tunicamycin and azole drugs [7,8]. In N. crassa, a rise in the cytosolic levels of Ca 2+ was associated with cell death induced by chitosan [9], Penicillium chrysogenum protein PAF [10], PAF26 [11] and staurosporine [12].
We have been using N. crassa to investigate the molecular mechanisms underlying the cell death process in-duced by the bacterial alkaloid staurosporine. Treatment with staurosporine leads to a complex response that includes the efflux of reduced glutathione (GSH), with a concomitant change in the intracellular redox state to a more oxidative environment and the consequent accumulation of reactive oxygen species (ROS) [13,14]. The drug also activates a drug resistance pathway that comprises the zinc binuclear cluster transcription factor CZT-1 [15] and ABC-3, an ATP-binding cassette (ABC) transporter [16]. A combined genetic and pharmacological strategy coupled to measurements of cytosolic alterations in Ca 2+ levels with the photoreporter aequorin revealed a three-step signature response to the drug that involves mobilization of the ion from intracellular stores as well as uptake from the external medium. N. crassa is a good model for the study of Ca 2+ dynamics as sequencing of its genome revealed a rich and versatile assortment of molecules involved in Ca 2+ homeostasis, including channels, pumps and signaling transducers [17][18][19]. A substantial fraction of the genome, however, remains to be characterized and additional unknown members of Ca 2+ machinery may exist. The beststudied Ca 2+ -binding moiety is the classical EF-hand, in which a Ca 2+ -interacting loop is flanked by two helices [20]. However, there are variations on this motif, especially at the structural level, and it is now accepted that more than a dozen different types of Ca 2+ -binding domains exist [21]. In particular, several proteins classes were shown to carry convergently evolved Ca 2+ -binding Dx[DN]xDG motifs. The distinct evolutionary origins of the motif are evident from the variability of structural contexts flanking the Ca 2+binding region [22,23]. Though initially identified in bacteria [24], it was later perceived that this motif is widespread across all kingdoms [22,23], including in a fungal caleosin [23,25] and lectin from Psathyrella velutina [22,26].
We have recently shown that staurosporine-induced cell death in N. crassa involves a precise sequence of Ca 2+ signaling events which are completely abolished by the inhibition of extracellular Ca 2+ uptake [12]. Thus, we investigated the effects of the drug in N. crassa cells growing in limited or excessive amounts of the ion. The results presented here revealed that extracellular Ca 2+ modulates cell death and the transcriptional alterations induced by staurosporine, and led to the identification of two novel putative Ca 2+ -binding proteins, encoded by the NCU08524 and NCU06607 genes.

RESULTS
The availability of extracellular Ca 2+ impacts staurosporine-induced cell death and intracellular Ca 2+ signaling Intracellular Ca 2+ dynamics play an important role during staurosporine-induced cell death in N. crassa, eliciting a complex response that includes Ca 2+ release from internal stores as well as uptake from the extracellular space [12]. Thus, we tested the response to staurosporine of cells growing with different Ca 2+ concentrations. We prepared a culture medium, designated 'no Ca 2+ ', by removing CaCl 2 from Vogel's minimal medium (MM) [27]. Small amounts of Ca 2+ as impurities from the other reagents in the solu-tion allowed N. crassa to grow well under such conditions (Fig. 1A, middle panel). In contrasting experiments, we supplemented medium with 20 mM CaCl 2 ('20 mM CaCl 2 ' medium), which represents an approximately 30 fold (but not toxic) increase as compared with the 0.68 mM CaCl 2 present in standard Vogel's medium (Fig. 1A, lower panel). We inoculated wild type cells on the centre of Petri dishes containing standard, no Ca 2+ or 20 mM CaCl 2 solid Vogel's MM supplemented with staurosporine and measured radial growth during a 104 hour time course. The drug inhibited growth in all media ( Fig. 1A-C). However, the staurosporine inhibitory effect was amplified in the absence of Ca 2+ (e.g., ~39% of inhibition after 32 hours of treatment with 1 μM staurosporine in no Ca 2+ versus ~14% in standard MM). On the other hand, growth inhibition was partially overcome in 20 mM CaCl 2 medium (e.g., ~37% of inhibition after 32 hours of treatment with 2.5 μM staurosporine in 20 mM CaCl 2 versus ~65% in standard MM). In GFS medium, which contains sorbose that promotes colonial growth [28], the outcome of modifying the extracellular concentration of Ca 2+ was similar: the effects of staurosporine were exacerbated in no Ca 2+ medium whereas growth inhibition was partially suppressed in the presence of 20 mM CaCl 2 , (Fig. 1D). Enhancement and suppression of inhibition of growth by absence of Ca 2+ and 20 mM CaCl 2 , respectively, was not reproduced when cells are treated with phytosphingosine (data not shown), another cell death inducer in N. crassa, supporting previous observations that staurosporine and phytosphingosine act by distinct mechanisms [14,29,30].
The production of ROS is an essential event during the response of N. crassa to staurosporine, as the addition of antioxidants blocks cell death [13]. We asked if the availability of Ca 2+ in the culture medium influenced cellular ROS production induced by staurosporine. While in standard MM cells stressed for 30 minutes with staurosporine displayed a ~4.2 fold-increase in ROS accumulation, the equivalent accumulation was ~10.4 (p-value = 0.001) and ~1.3 (p-value = 0.014) in no Ca 2+ and 20 mM CaCl 2 media, respectively (Fig. 1G). Given the importance of ROS formation for cell death provoked by staurosporine, this Ca 2+ -ROS dependence is likely associated with the distinct cell death phenotypes in the different culture media.
To further stress the modulatory effect of altering the levels of Ca 2+ in staurosporine-induced cell death, we combined the drug with the Ca 2+ chelator BAPTA and the Ca 2+ ionophore A23187 (calcimycin) in GFS plates with standard MM. As expected, extracellular Ca 2+ blockage with BAPTA  synergized whereas elevation of intracellular Ca 2+ with A23187 antagonized the effects of staurosporine (Fig. 1H).
In agreement, pre-incubation with BAPTA significantly enhanced apoptotic levels induced by staurosporine, whereas they were partially inhibited by A23187, as measured by YOPRO-1 staining (Fig. 1I). Altogether, these data consistently indicate that the absence of Ca 2+ enhances staurosporine-induced cell death, whereas excess (but not toxic) concentrations of Ca 2+ partially prevent it. Using cells expressing the cytosolic Ca 2+ reporter gene aequorin, we described recently that staurosporine induces a three-step Ca 2+ response comprising three main signals called "A", "B" and "C". Peak "A" occurs immediately after the addition of the drug and lasts approximately 20 minutes; peak "B" appears after 35-40 minutes and lasts approximately 80 minutes; phase "C" is a prolonged and continuous elevation of cytosolic Ca 2+ [12]. In the absence of extracellular Ca 2+ , the Ca 2+ response to staurosporine was compromised, with peaks "A" and "B" showing substantial reduction ( Fig. 2A-B). In medium with 20 mM CaCl 2 there was a strong boost in peak "A", while peak "B" remained similar to standard MM. These alterations in intracellular Ca 2+ dynamics are in agreement with our findings that extracellular Ca 2+ uptake is important for the Ca 2+ signaling response to staurosporine [12].

Distinct transcriptional programs in staurosporine-treated cells depend on the extracellular Ca 2+ availability
Treatment with 20 μM staurosporine led to significant alterations in the expression of ~28% of the N. crassa genes (1921 genes) [15]. The extent of this transcriptional response was amplified in the absence of Ca 2+ (~36% of genes (2749 genes) were altered by the drug), and markedly suppressed by 20 mM of CaCl 2 (~6% genes (454 genes) showed altered profiles) ( Fig. 3A-B). In standard medium, approximately half of the genes were induced and approximately half were repressed by staurosporine [15]. In no Ca 2+ medium, 43.1% of the altered genes were upregulated, whereas 56.9% where downregulated. Remarkably, in the presence of excessive Ca 2+ most of the altered genes were repressed (~68%). The complete dataset obtained from these RNA-seq experiments is presented as File S1.
We analysed the distribution of the individual genes induced and repressed by staurosporine in the different culture media. In the absence of Ca 2+ , 792 (66.8%) of the induced genes were specific to this condition while in 20 mM CaCl 2 this corresponded to 67 (46.2%) genes ( Fig. 3C and File S1). Among repressed genes, 875 (56.0%) and 157 (50.8%) genes were specific to no Ca 2+ and 20 mM CaCl 2 media, respectively ( Fig. 3D and File S1). Only minor fractions of the altered genes (30 and 74 genes, induced and repressed, respectively) were common to all conditions.
FunCat was used to examine the enrichment of functional categories in the different gene sets (File S2; Table S1-S2). The most interesting Ca 2+ concentration-specific differences were found among the downregulated genes. In the absence of Ca 2+ , staurosporine led to the downregulation of genes included in numerous categories. These encompass various groups within the 'Metabolism' supercategory such as 'metabolism of the aspartate family', 'purine nucleotide/nucleoside/nucleobase metabolism', 'phosphate metabolism' and 'tetracyclic and pentacyclic triterpenes metabolism'. Also, there was a specific repression of genes involved in 'Cell cycle and DNA processing' and 'Transcription', namely 'mRNA synthesis' and 'mRNA processing'. Categories related to stress responses like 'Protein fate', 'unfolded protein response', 'cellular sensing and response to external stimulus', 'cell growth', 'antiapoptosis' were also repressed. Finally, while in standard MM there was a downregulation of 'aerobic respiration' genes, in no Ca 2+ medium, other bioenergetic pathways were repressed: 'glycolysis and gluconeogenesis' and 'pentose-phosphate pathway'. These data indicates that a highly complex transcriptional program is established by staurosporine-treated cells growing in the absence of Ca 2+ resulting in the repression of several genes related with an anti-stress response. It seems that, when Ca 2+ is limited, staurosporine causes the downregulation of cellular de-fences and this may be related to the high susceptibility of these cells to the drug.
RNA-seq analysis of the fungal reaction to staurosporine revealed that the drug can trigger very distinct transcriptional responses depending on the concentrations of Ca 2+ in the culture medium. This is probably associated with the observed differences in the levels of staurosporine-induced cell death in the three media.

Limited and excess of Ca 2+ alters basal gene expression
The transcriptional response to the lack or excess of Ca 2+ was analysed by comparing basal gene expression in no Ca 2+ , 20 mM CaCl 2 and standard MM (File S3). The fraction of genes with altered expression in no Ca 2+ and 20 mM CaCl 2 compared with standard MM was 3.9% and 1.8%, respectively (Fig. 4A-B), corresponding to 281 and 129 genes, respectively. While half of the genes altered in no Ca 2+ were induced and half were repressed, in 20 mM CaCl 2 most of the altered genes were upregulated (71.3%). Most of the genes were specifically induced or repressed depending on the particular culture medium: 97.2% and 97.1%, respectively, for no Ca 2+ , and 95.7% and 89.2%, respectively, for 20 mM CaCl 2 (Fig. 5A-B). Not surprisingly, categories related to 'ion transport', 'homeostasis of cations' and 'homeostasis of anions' were enriched in the set of genes induced in the absence of Ca 2+ (Fig. 5A and File S4), which is perhaps linked to a cellular attempt to recover proper conditions of osmolarity. Consistently, these categories were also significantly enriched among the genes repressed in the presence of 20 mM CaCl 2 (Fig. 5B). Lack of Ca 2+ also seems to be associated to a metabolic adaptation of the cells, since there was enrichment in the induction of genes falling on the 'metabolism' super-category, like 'amino acid metabolism', 'nitrogen, sulfur and selenium metabolism' and 'anaerobic respiration'. We looked for alterations in the expression of known Ca 2+ channels, Ca 2+ -ATPases, Ca 2+ -exchangers, Ca 2+dependent signaling molecules and other Ca 2+ binding proteins [18,19,31,32] (Fig. S1 and File S5). NCU11680, NCU04736 and NCU07075, encoding the TRP channel YVC-1, the Ca 2+ -ATPase NCA-2 and the Ca 2+ /H + antiporter CAX, respectively, were repressed in the absence of Ca 2+ . NCU04265, encoding the invertase enzyme (INV) was induced in the absence of Ca 2+ . The induction of INV, an enzyme that hydrolyses extracellular sucrose into glucose and fructose, is a marker of the process of carbon catabolite repression [33,34]. NCU08147, encoding the stress-related Ca 2+ -ATPase ENA-2 [35] was repressed in the presence of 20 mM CaCl 2 . NCU05046 and NCU07966, encoding ENA-1 and TRM-1, respectively, were induced by no Ca 2+ and repressed in 20 mM CaCl 2 medium. These results further suggest an intracellular metabolic remodeling in response to Ca 2+ availability.

Two putative novel components of the N. crassa Ca 2+ homeostasis machinery
We found that 188 genes with unknown function ("hypothetical proteins") had altered expression in no Ca 2+ or in 20 mM CaCl 2 medium. We screened the respective protein sequences for the presence of Ca 2+ -binding patterns using CaPS [20] and identified 6 candidates (NCU01697, NCU08524, NCU06116, NCU07582, NCU06607 and NCU03647). We further considered NCU08524 and NCU06607, because the others did not display convincing similarities with known proteins, even around the predicted Ca 2+ -binding motif. NCU08524 was repressed in Ca 2+free medium whereas NCU06607 was induced by 20 mM CaCl 2 (Fig. 6A). Fig. 6B shows the alignment of the predict-ed Ca 2+ -binding motifs of NCU08524 and NCU06607 with those of proteins containing the Dx[DN]xDG motif [22,23]. The NCU08524 motif sequence resembles that of the rhamnogalacturonan lyase YesW from Bacillus subtilis. YesW exemplifies β-propeller structures in which the Dx[DN]xDG motif can be embedded in one of the propeller blades, forming a 'calcium blade' structure [22]. Such a calcium blade was recently described in a fungal lectin from Psathyrella velutina [22]. NCU08524 is similar to a fucose-specific lectin from Macrophomina phaseolina (Table S3) and matches from HHpred suggest that it contains a 6-bladed propeller structure in which β-strands flank the putative Ca 2+ -binding motif. Taking advantage of the similarity between fucose-specific lectins and NCU08524, molecular models were built for the latter in a situation of Ca 2+ binding ( Fig. S2A-C). In proteins containing the Dx[DN]xDG motif, interactions with bound Ca 2+ ions are supplemented by the interaction of side chains from at least one downstream acidic residue [22]. Therefore, three molecular models were prepared, considering Glu 502 , Glu 503 or Asp 504 as the possible additional acidic residue. The models show that NCU08524 can accommodate a Ca 2+ -binding geometry by combining the Dx[DN]xDG motif with any of these downstream residues (Fig. S2A-C).
The predicted Ca 2+ -binding motif of NCU06607 (Fig. 6C) is comparable to the Dx[DN]xDG motif of a calmodulin from Paramecium tetraurelia [23]. Molecular modeling of NCU06607 was not feasible because the Dx[DN]xDG motif is not located within a folded domain of known structure, but at the beginning of a predicted intrinsically disordered region. Anyway, the Dx[DN]xDG motif is sometimes found in highly disordered regions [36]. NCU06607 has a predicted extracellular localization and there is a strong bias in the localization of proteins possessing the Dx[DN]xDG motif towards the cell surface or secretion [22]. Plasma membrane localization together with a transmembrane domain was predicted for NCU08524 (Table S3).
The growth of ΔNCU08524 and ΔNCU06607 was compared with the wild type strain under different conditions of Ca 2+ availability. The growth of both mutants was significantly reduced in no Ca 2+ medium and in medium containing the Ca 2+ chelator BAPTA (Fig. 6D). ΔNCU06607 also showed a significant reduction in growth in the presence of 20 mM CaCl 2 as compared with wild type. The observed differences in growth were not merely due to non-specific osmotic stress, since all strains were similarly affected by the addition of 4% NaCl. ΔNCU06607 was slightly more sensitive to staurosporine than wild type, suggesting that the respective protein may be involved in the response to the drug (Fig. 6E). In summary, the RNA-seq dataset and further analysis revealed two putative novel components of the Ca 2+ -handling machinery in N. crassa.

Several members of the Ca 2+ homeostasis machinery are involved in the fungal response to staurosporine
To further substantiate the crucial role of Ca 2+ during the N. crassa response to staurosporine and to identify mediators of the process, we assayed the drug sensitivity profile of approximately 50 deletion strains encoding proteins in-volved in Ca 2+ homeostasis. We showed recently that deletion of genes encoding phospholipase C enzymes, especially PLC-2, results in mutants resistant to staurosporine, while strains lacking the Ca 2+ channels CCH-1 and MID-1 are hypersensitive [12]. Several other deletion strains that lack Ca 2+ -permeable channels, Ca 2+ -ATPases or other Ca 2+ binding molecules involved in cellular signaling showed a resistance phenotype to staurosporine that was dissimilar from the wild type strain (Fig. 7). A resistance phenotype was particularly strong in deletion mutants for genes encoding the Ca 2+ -ATPases NCA-2 and PMR-1, and deletion of other Ca 2+ -ATPases, namely ENA-1, ENA-2 and NCU10143 also led to increased resistance. The disruption of the signaling molecules calmodulin, CAMK-4, and CSE-1 aug- mented resistance, whereas strains carrying deletions in genes encoding STK-16 and PRD-4 resulted in increased susceptibility to staurosporine. Table 1 lists strains with altered sensitivity to staurosporine and Table S4 lists all strains tested. The deletion strains for NCU05360, NCU03804, NCU03833, NCU09871, NCU06347, NCU06617, NCU02411 and NCU09265 were not tested due to strain unavailability or their heterokaryotic nature. This data further stresses the importance and complexity of the fungal response to staurosporine, implicating numerous Ca 2+related proteins of different function and subcellular localization in the process.

DISCUSSION
Here we show that controlling the amount of Ca 2+ available in the extracellular milieu is enough for a robust modula-tion of the effects of staurosporine in N. crassa. Cell death is strongly enhanced by Ca 2+ limitation whereas it is partially suppressed by a ~30-fold excess of external Ca 2+ . A similar behavior was observed in S. cerevisiae cells treated with miconazole or terbinafine, inhibitors of ergosterol biosynthesis [37], suggesting common stress-responsive pathways in the two fungi, even when the stimuli have apparently distinct targets. Supplementation of the medium with high Ca 2+ concentrations (20 mM CaCl 2 , as used here) protected N. crassa cells from toxicity induced by PAF, an antifungal protein from Penicillium chrysogenum that also disturbs intracellular Ca 2+ homeostasis [10]. Our results substantiate the view that Ca 2+ is an important player during cell death.
The effects of staurosporine in media containing different amounts of Ca 2+ are remarkably paralleled by distinct  cytosolic Ca 2+ dynamics, in line with our previous findings that extracellular Ca 2+ uptake is required for the development of the cytosolic Ca 2+ signature induced by staurosporine [12]. We have recently demonstrated that during the staurosporine-induced Ca 2+ signature, peaks "A" and "B" precede cell death, suggesting that they represent signaling events. In the presence of 20 mM CaCl 2 the peak "A" is enhanced, whereas in medium lacking Ca 2+ peaks "A" and "B" of the staurosporine-induced cytosolic Ca 2+ response are abolished. Thus, intracellular Ca 2+ signaling that is triggered by staurosporine is markedly modulated by the concentration of Ca 2+ in the external medium and might have a role in determining cell survival. However, the intracellular response to staurosporine is very complex and direct conclusions on the susceptibility of the cells based solely on the cytosolic Ca 2+ profile, although tempting, are not possible. For instance, a Δplc-2 mutant shows increased resistance to the drug despite that its cytosolic Ca 2+ response to staurosporine is nearly obliterated [12]. Thus, it seems that events downstream of the cytosolic changes in Ca 2+ levels ultimately define cell fate in response to staurosporine.
The screening for staurosporine sensitivity of Ca 2+handling mutant strains revealed a number of proteins likely involved in the cell death response to the drug. Further studies are needed to contextualize them functionally in the mechanism of action of the drug. NCA-2 is a Ca 2+ -ATPase which pumps excess of cytosolic Ca 2+ to the vacuoles or extracellular space [38,39], and the respective knockout cells are very resistant to staurosporine. The knockout strain for PMR-1, a Golgi-localized Ca 2+ -ATPase [40], is also very resistant to staurosporine and yeast ΔPMR1 knockout cells accumulate much less Ca 2+ in the ER [41]. Lack of these functions in Δnca-2 and Δpmr-1 cells is likely related to the phenotype of the strains upon treatment with staurosporine, as both the ER and the vacuoles play an important role during the fungal response to staurosporine [12]. It is plausible that the lack of NCA-2 or PMR-1 results in an increased accumulation of Ca 2+ in the cytosol and this is consistent with the observations that high levels of Ca 2+ (below a toxicity threshold) protect against the effects of staurosporine. The lack of NCA-2 has been recently associated with increased sensitivity to UV exposure, further implicating the molecule in cell survival [42]. Other Ca 2+ -related deletion strains present altered resistance to staurosporine, including Δcse-1, Δcamk-4 and Δstk-16. Knockout of cse-1 has been shown to lead to impaired germling communication and fusion [43], as well as increased sensitivity to UV irradiation and Ca 2+ [44]. A previous study has shown that disruption of camk-4 results in benomyl resistance, while disruption of stk-16 leads to increased sensitivity to fludioxonil, sodium chloride and benomyl [45].
The N. crassa transcriptional response to staurosporine is robustly reduced in the presence of 20 mM CaCl 2 when compared with standard MM, whereas limited extracellular Ca 2+ causes alterations in an increased number of genes. Thus, different concentrations of extracellular Ca 2+ not only lead to distinct intracellular Ca 2+ dynamics, ROS accumulation and cell death levels, but also seem to be coupled to unique transcriptional profiles. It appears that, in the absence of Ca 2+ , cells treated with staurosporine turn off a diverse array of biological processes. We discovered several enriched functional categories in the set of no Ca 2+ medium-specific repressed genes, including the repression of genes involved in cell cycle, signal transduction, cell growth, anti-apoptosis, cellular polarization, protein fate and the unfolded protein response. It is possible that the decreased activity of various intracellular pathways explains the inability of cells to cope with the staurosporine insult and the consequent increased cell death.
N. crassa clearly translates changes in extracellular Ca 2+ into a transcriptional response, since alterations in gene expression are caused alone by extracellular Ca 2+ limitation or overload. This likely results from the homeostatic adaptation to the abnormal Ca 2+ environment, i.e., the observed differences in gene expression might be triggered after the uptake or release of Ca 2+ to balance the intracellular levels of the ion. Alternatively, these alterations may be secondary to the activity of a surface-localized Ca 2+ sensor, although a sequence homologue of the extracellular Ca 2+sensing receptor of animals and plants [46,47] has not been found so far in N. crassa. Given the predicted cell surface localization of the two putative novel Ca 2+ -binding proteins, NCU08524 and NCU06607, and the growth phenotype of the respective deletion strains under Ca 2+ stress, it is tempting to speculate that they could play a role in Ca 2+ sensing. The chemotropic interaction during conidial anastomosis tubes during cell fusion is Ca 2+ -dependent [43] and both NCU08524 and NCU06607 genes have been recently shown to be under the control of the PP-1 transcription factor, a regulator of cell fusion [48,49].
In summary, we show that the effects of the antifungal [13][14][15][16]50] and anticancer [51] agent staurosporine are robustly modulated by the accessibility to Ca 2+ in the culture medium. The amount of extracellular Ca 2+ affects staurosporine-triggered intracellular events like Ca 2+ signaling, the production of ROS and cell death. Our results indicate that these differences are linked to the stimulation of unique transcriptional programs and further highlight the importance of Ca 2+ during fungal cell death. In addition, our RNA-seq dataset will be a useful resource for investigations on the role of Ca 2+ on different aspects of fungal biology.

Growth assays
Hyphal growth in solid medium at 26°C was obtained by measuring colony elongation after the inoculation of 20 μl containing 5x10 4 conidia on the centre of large Petri dishes (ϕ 14.2 cm). Growth in liquid medium was examined by incubating 1x10 4 conidia/ml at 26°C, 100 rpm, under constant light in 96-well plates (200 μl/well) and following absorbance at 450 nm [53] during 24 hours. The % growth versus the respective control and the growth rate were calculated for each condition. For the spot assays, 3-fold serial dilutions up to 1x10 4 cells/ml were prepared for each strain. 5 μl from each dilution were spotted on plates containing glucose-fructose-sorbose (GFS) [28] plus the relevant chemicals and incubated for 3 days at 26°C.

Flow cytometry measurements of cell death and reactive oxygen species (ROS)
For each experiment and strain, a control without staining was prepared to define autofluorescence. Samples were read in a BD FACS Calibur and data analyzed with FlowJo (Tree Star). The levels of cell death were determined by staining cells with YOPRO-1 (Life Technologies) or propidium iodide (PI; Sigma-Aldrich). An inoculum of 10 6 conidia/ml in the appropriate medium was incubated for 4 hours at 26°C (140 rpm, constant light). Staurosporine (20 μM) was added and growth resumed for further 2 hours. The cells were harvested by centrifugation, washed twice with PBS and incubated either with 0.1 μM YOPRO-1 (samples were kept 20 min on ice before reading) or 2 μg/ml PI (samples were immediately ready for cytometry). The production of ROS was measured using the fluorescent probe dihydrorhodamine 123 (Sigma-Aldrich). After growing 10 6 conidia/ml for 4 hours in the indicated medium at 26°C, 20 μg/ml dihydrorhodamine 123 and staurosporine were added for further 30 minutes. Samples were harvested by centrifugation and washed twice with PBS before being resuspended in PBS and read in the cytometer.

Cytosolic Ca
2+ measurements A bioluminescent method based on the reaction of cytosolic Ca 2+ with the genetic probe aequorin in wild type cells was employed as previously described [12]. Briefly, 100 μl of conidia at a concentration of 2x10 6 cells/ml, 5 μM coelenterazine (Santa Cruz Biotechnology) and minimal medium were added to white opaque 96-well plates and incubated for 6 hours at 26°C, in the dark, without agitation. Luminescence (in RLU, relative light units) was captured over time on a Bio-Tek Synergy HT microplate reader. Maximum levels of luminescence, determined in extra wells by measuring luminescence during 3 mins after the injection of 100 μl of 3M CaCl 2 in 20% ethanol, were used to normalize each experiment (cytosolic Ca 2+ levels (arbitrary units) = experimental RLU values / total emitted luminescence). Quantification was performed by summing the normalized values and data are expressed as mean ± SEM. The values from control DMSO-treated samples were subtracted from staurosporine-treated samples to obtain the "staurosporine-induced amplitude of response".

Statistical and protein sequence analyses
The non-parametric Mann-Whitney test was used for comparisons between two groups using SPSS 20 (SPSS Inc.) and pvalues ≤ 0.05 were considered statistically significant. Clus-talW2 [54] was used to align protein sequences and PSI-BLAST [55] was used for iterative searches. Secondary structure was predicted with MEMSAT3 [56], conserved domains with Inter-ProScan [57], transmembrane domains with TMHMM 2.0 [58], subcellular localization with WoLF PSORT [59] and Ca 2+binding motifs with CaPS [20].