Ambient pH Controls Glycogen Levels by Regulating Glycogen Synthase Gene Expression in Neurospora crassa. New Insights into the pH Signaling Pathway

Glycogen is a polysaccharide widely distributed in microorganisms and animal cells and its metabolism is under intricate regulation. Its accumulation in a specific situation results from the balance between glycogen synthase and glycogen phosphorylase activities that control synthesis and degradation, respectively. These enzymes are highly regulated at transcriptional and post-translational levels. The existence of a DNA motif for the Aspergillus nidulans pH responsive transcription factor PacC in the promoter of the gene encoding glycogen synthase (gsn) in Neurospora crassa prompted us to investigate whether this transcription factor regulates glycogen accumulation. Transcription factors such as PacC in A. nidulans and Rim101p in Saccharomyces cerevisiae play a role in the signaling pathway that mediates adaptation to ambient pH by inducing the expression of alkaline genes and repressing acidic genes. We showed here that at pH 7.8 pacC was over-expressed and gsn was down-regulated in wild-type N. crassa coinciding with low glycogen accumulation. In the pacCKO strain the glycogen levels and gsn expression at alkaline pH were, respectively, similar to and higher than the wild-type strain at normal pH (5.8). These results characterize gsn as an acidic gene and suggest a regulatory role for PACC in gsn expression. The truncated recombinant protein, containing the DNA-binding domain specifically bound to a gsn DNA fragment containing the PacC motif. DNA-protein complexes were observed with extracts from cells grown at normal and alkaline pH and confirmed by ChIP-PCR analysis. The PACC present in these extracts showed equal molecular mass, indicating that the protein is already processed at normal pH, in contrast to A. nidulans. Together, these results show that the pH signaling pathway controls glycogen accumulation by regulating gsn expression and suggest the existence of a different mechanism for PACC activation in N. crassa.


Introduction
Neurospora crassa is a filamentous fungus that has been widely used as a model organism for studies on fundamental aspects of eukaryotic biology. The completion of its genome sequence [1] and, more recently, the availability of a set of mutant strains individually knocked-out in specific genes have enormously accelerated the investigation of many aspects of the biology of this organism. We have been using N. crassa to study how the metabolism of glycogen, a storage carbohydrate that functions as carbon and energy reserve, is regulated. Glycogen biosynthesis starts with the self-glucosylating initiator enzyme glycogenin (EC 2.4.1.186), termed GNN in N. crassa [2]. After an oligosaccharide is synthesized, the elongation step is catalyzed by glycogen synthase (EC 2.4.1.11) and a branching enzyme (EC 2.4.1.18) that mediate the formation of a-1,4and a-1,6-glycosidic bonds, respectively, characteristic of the glycogen molecule. Glycogen synthase is the rate-limiting enzyme in this process and is inhibited by phosphorylation; however, the phosphorylated form can be activated by the allosteric effector glucose-6-phosphate [3].
In N. crassa, glycogen synthase (GSN) has been characterized at gene and protein levels. Under normal growth conditions, gsn expression is maximal at the end of the exponential growth phase, which coincides with the highest glycogen content. However, under stress conditions, such as heat shock, gsn expression is down-regulated and the enzyme is less active. This may explain the decrease in glycogen levels observed in cells exposed to heat shock [4][5][6]. gsn expression and GSN phosphorylation are also regulated by the cAMP-protein kinase A (PKA) signaling pathway [7]. Thus, the regulation of glycogen metabolism in N. crassa involves different proteins whose activation may be influenced by different signaling pathways. Analysis in silico of the gsn 59-flanking region led to the identification of cis DNA-binding motifs specific for transcription factors acting via different signaling pathways [8]. These motifs include a recognition site for the A. nidulans zinc-finger transcription factor PacC, suggesting a possible role for pH in regulating glycogen synthesis.
PacC has been extensively studied in A. nidulans and plays an important role in the pH signaling pathway. This transcription factor mediates the cell adaptation to neutral-alkaline pH by activating genes that are preferentially expressed at alkaline pH and repressing those preferentially expressed at acidic pH [9]. In A. nidulans, PacC is activated by two successive proteolytic cleavage steps at the C-terminus, the first being pH-dependent and activated by the products of six pal genes while the second is proteasome-mediated and pH-independent [10,11]. The proteolysis lead to the active protein PacC 27 that contains a DNA-binding domain formed by three C 2 H 2 zinc fingers capable of binding to the core consensus sequence 59-GCCARG-39 present in the promoters of pH-regulated genes (reviewed in [12,13]).
The Rim101p, a PacC orthologue in S. cerevisiae, was initially identified as a positive regulator of meiosis and rim101 mutants are sensitive to Na + or Li + ions and grow poorly at low temperatures [14]. The yeast protein therefore appears to have a broader role than simply that of promoting alkaline pH-inducible responses [15]. While Rim101p is associated with the pH response, there are important differences between PacC and Rim101p. For example, the yeast protein requires only a single cleavage step to be activated [16] and, whereas PacC acts as a transcriptional activator under alkaline pH, Rim101p exerts its role as a repressor [15]. In addition, at least three pathways participate in the response to high pH stress in S. cerevisiae (reviewed in [17]). Thus, the molecular mechanisms involved in the pH response differ among organisms.
PacC orthologues have been identified in numerous filamentous fungi and the role of PacC as a mediator of pH regulation in fungal pathogenicity was first described in Candida albicans, in which Rim101p governs pH responses, dimorphism, and pathogenesis [18]. In N. crassa, the PACC pathway is involved in the glycosylation of Pi-repressible acid phosphatase [19] and in the transcription of the hsp70 gene [20].
In this report, we demonstrated that the pH signaling pathway regulates glycogen metabolism in N. crassa by a process in which PACC may play a central role. As an A. nidulans PacC motif was identified in the gsn promoter, and since N. crassa has the six A. nidulans pal gene orthologues, we investigated the influence of pH on glycogen accumulation and the regulation of gsn expression. Gel mobility shift assays showed that recombinant PACC recognized and bound specifically to the gsn promoter; this binding was confirmed in vivo by ChIP-PCR analysis. We also found that PACC proteolysis may involve a different way from that described for the A. nidulans protein.

Knockout of the N. crassa pacC Gene and Growth Phenotype Analysis
To investigate the role of PACC in the regulation of glycogen metabolism we generated a null deleted strain by replacing the entire pacC ORF (NCU00090, 1,866 bp) with the bar gene in a mus-52 KO background strain ( Figure 1A). The gene replacement and correct integration were confirmed by PCR analysis of the transformants ( Figure 1B). However, we were unable to eliminate the mus-52 mutation after crossing the transformants with a wildtype strain due to the close linkage between the mus-52 and pacC genes. The morphological characteristics of the pacC KO strain were analyzed by growth on VM agar plates and in race tubes. Since there were no apparent differences in the phenotypes of the mus-51 KO (used as the wild-type in this work) and mus-52 KO strains [21] we concluded that the growth defects observed in the pacC KO strain were caused specifically by the absence of the transcription factor. Linear growth analysis in race tubes showed that both strains grew better at acid (4.2) and normal (5.8) pH than at alkaline pH and that the pacC KO strain was unable to grow at alkaline pH (7.8) ( Figure 1C). We have also observed an accumulation of brown pigments after five days under normal growth conditions in flask ( Figure 1D), characterized as melanin and described in other transcription factor mutant strains [22].
Radial growth was analyzed in plates after 24 h of growth in VM medium and VM medium supplemented with yeast extract (Figures 1E and 2). The pacC KO strain displayed similar growth to the wild-type strain in both media at acid pH and slightly reduced growth at normal pH; the pacC KO strain did not grow at alkaline pH. The strains showed enhanced hyphal branching and shorter aerial hyphae in both media at alkaline pH ( Figure 2). However, the pacC KO strain displayed severe growth defects at alkaline pH, including a reduction in the extent and production of aerial hyphae. Together, these results show that PACC is required for growth under alkaline conditions.

Glycogen Levels and Gene Expression Under pH Stress
The gsn gene is regulated at transcriptional and post-translational levels. We have previously reported that gene expression and glycogen levels were altered under stress conditions such as heat shock and carbon starvation [5,6], both of which affect glycogen metabolism in N. crassa. The identification of a DNA motif for the A. nidulans pH signaling regulator PacC in the gsn promoter [8] prompted us to investigate whether this gene could be regulated by pH stress. Cells from both strains were germinated at pH 5.8 and then subjected to acid and alkaline pH stress after which samples were collected for glycogen quantification and RNA extraction. The glycogen levels in the pacC KO strain were higher than in the wild-type strain at normal growth pH (pH 5.8) ( Figure 3B), indicating that PACC is important for maintaining normal glycogen levels during vegetative growth. However, under acid stress (pH 4.2), both strains accumulated similar amounts of glycogen ( Figure 3A and 3B), suggesting that PACC does not affect glycogen accumulation under this condition in agreement with the growth results shown in Figure 1C. At alkaline pH, there was a decrease in the glycogen levels of both strains compared to the levels at pH 5.8. The fact that the glycogen level in the pacC KO strain after 30 min of alkaline stress was similar to that of the nonstressed wild-type strain ( Figure 3B, time 0) emphasizes the role of PACC as a major regulator of glycogen accumulation under alkaline pH stress. The glycogen levels were also quantified in wild-type cells that were returned to normal pH after being subjected to acid and alkaline pH stress (RE, recuperation), however they did not recover the glycogen levels seen before stress ( Figure 3A) in both conditions.
To investigate whether the effect of extracellular pH on glycogen accumulation resulted from changes in gsn and gpn (named here as the gene encoding the enzyme glycogen phosphorylase) expression we analyzed their mRNA levels during pH stress and recuperation at pH 5.8 in a wild-type strain. During acidic stress, the gsn transcript levels were kept constant while the gpn transcript decreased during 60 min of incubation ( Figure 4A) and the levels were not recovered after returning the cells to normal growth pH ( Figure 4A, RE). When cells were transferred to alkaline medium, gsn and gpn transcript levels decreased markedly during 60 min of stress, however only gsn levels were recovered to normal levels after 60 min of incubation at pH 5.8 ( Figure 4A). These results indicate that changes in extracellular pH control gsn expression. We also analyzed the pacC mRNA levels under pH stress and recuperation at pH 5.8 in the wild-type strain ( Figure 4A). The pacC transcript size was ,3 kb. Whereas pacC expression was low at acid pH, at alkaline pH there was a marked increase in expression compared to that at physiological pH (5.8), mainly during the first 15 min of stress. Basal expression was achieved when the mycelia were transferred back to pH 5.8 (RE). Expression of gsn and gpn genes was analyzed in the pacC KO strain under acid and alkaline pH stress and compared to the wild-type strain. The gsn mRNA levels were higher in the pacC KO strain than in the wild-type strain at all pH values ( Figure 4B), suggesting a repressive role for PACC in gsn expression. This result explains, at least partially, the glycogen levels shown in Figure 3B. There was very low expression for gpn gene in the mutant strain in all pH suggesting that PACC also controls gpn expression under normal pH. Together, the results shown in Figures 3B and 4 suggest that alkaline pH induces pacC expression and a concomitant decrease in glycogen levels by down-regulating the gsn gene. These findings characterize gsn as an acid-specific gene.

Glycogen Levels and Gene Expression Under Combined pH and Heat Stress
As previously described, heat shock (45uC) down-regulates gsn expression, resulting in decreased glycogen accumulation compared to that observed at 30uC (the normal temperature of growth) [5,6]. We therefore examined the effect of a combination of heat and pH stress on glycogen accumulation and in the expression of gsn and pacC genes in the wild-type strain. For this, cells were germinated at pH 5.8 and transferred to media of different pH preheated to 45uC. Glycogen levels were strongly reduced during heat shock, regardless of the pH ( Figure 5A, black bars, HS). When heat stress was removed while maintaining the pH stress, normal glycogen levels were regained only in media with normal and acid pH ( Figure 5A, compare with the results shown in Figure 3A for pH 4.2). The cells were unable to recover normal glycogen levels at alkaline pH ( Figure 5A, pH 7.8) and they were the same as under heat stress.
The influence of pH and temperature stress on gsn expression was also analyzed, as an attempt to correlate with glycogen accumulation. The gsn transcript levels were reduced in cells subjected to heat shock, regardless the media pH ( Figure 5B). However, when cells were transferred back to normal temperature (30uC) and pH stress was maintained the gsn expression levels were recovered only at pH 4.2 and 5.8, but not at pH 7.8, in agreement with the results shown in Figure 5A. As with gsn, pacC expression was strongly repressed by heat stress at both pH values. In addition, gene expression was recovered after returning the cells to normal temperature (30uC) at all pH values ( Figure 5B). This interesting result suggests that the heat signaling pathway apparently elicits stronger cellular responses than the pH signaling pathway, at least for pacC expression. It is noteworthy the existence of six STRE (STress Response Elements) and four HSE (Heat Shock Elements) motifs in the pacC promoter, all of which are DNA elements known to be responsive to heat stress.

Recombinant PACC Binds to the gsn Promoter
To investigate whether the PacC DNA motif identified in the gsn promoter was, indeed, recognized by PACC, a gel mobility shift assay was performed using a truncated recombinant PACC and a 146 bp DNA fragment containing the PacC core sequence (59-CTTGGC-39) as probe. Recombinant PACC bound efficiently to the fragment corresponding to the PacC recognition motif in the gsn promoter ( Figure 6A, lanes 2 and 6). DNA-protein binding was reduced in the presence of unlabelled specific competitor ( Figure 6A, lanes 3 and 7). The specificity of the DNA-protein complexes was further confirmed by adding a 27 bp DNA oligonucleotide (oligo pacC) as a specific competitor ( Figure 6A, lanes 4 and 8) and by supershift experiments ( Figure 6A, lanes 11 to 13). An interesting result was the presence of two complexes exhibiting different molecular masses when the specific competitor was added to the reaction ( Figure 6A, lanes 3 and 7). We speculated that they may represent complexes with distinct conformational structures. Additional experiments will be necessary to clarify this. The DNA oligonucleotide completely abolished both complexes and the addition of anti-PACC antibody to the reaction increased the molecular mass of the DNA-protein complex, which shifted close to the gel origin ( Figure 6A, lanes 12 and 13).
The specificity of the DNA-protein complex was also confirmed by using a mutated pacC probe (mpacC) in which the 59-CTTGGC-39 core sequence was changed to 59-AAATTA-39 ( Figure 6B). There was a marked reduction in the formation of DNA-protein complex and a weaker complex was only visible after adding 2.0 mg of protein ( Figure 6B

Binding of Proteins from Crude Cellular Extract to the PacC Binding Site of the gsn Promoter
To investigate whether PACC binding to the gsn promoter was pH-dependent, gel shift assays were performed with crude cellular extracts from wild-type and pacC KO cells grown under normal (5.8) and alkaline (7.8) pH. The cellular extracts were fractionated by affinity chromatography, bound proteins were eluted with a KCl gradient and the fractions were analyzed for their ability to bind to the pacC probe previously used. Fractions from wild-type strain mycelium showing DNA-binding activities were identified in extracts from cells grown at both pH conditions (results not shown). One active chromatographic fraction from each pH was assayed for its binding to the PacC consensus site of the gsn promoter in the presence of both competitors. DNA-protein complexes were observed when proteins were incubated with the DNA fragment ( Figure 7A, lanes 2 and 7). When the same fragment was used as a specific competitor in 10-fold molar excess the complexes were shifted only when crude cellular extract prepared from mycelia grown at pH 5.8 was used as the protein source ( Figure 7A, lane 3). However, the specific competitor oligo pacC was able to shift the DNA-protein complexes when proteins from both extracts were analyzed ( Figure 7A, lanes 4-6, 10 and 11). The binding reaction was apparently stronger when cellular extract prepared from cells grown at alkaline pH was used as a protein source since the DNA-protein complex was shifted only at higher oligo pacC concentrations (20-and 30-fold molar excess). Fractions from pacC KO strain were also assayed for binding reactions and no DNA-protein complex was observed. The Figure 7A (lanes [13][14][15][16] shows the results with one chromatographic fraction from each pH and the competition with the oligo pacC. This result indicated the existence of an active PACC protein in the wild-type strain that specifically bound to the PacC motif even under normal growth pH (5.8). It is noteworthy that the DNAprotein complexes migrated close to the gel origin suggesting that they may include other proteins in addition of PACC. It is also interesting since low levels of pacC transcript could be detected in wild-type cells grown at pH 5.8 ( Figure 4A).
We also investigated whether the PACC present in the two cellular extracts had equal molecular mass since the same protein from A. nidulans grown at alkaline pH has a lower molecular mass than that observed at normal growth pH [10]. The presence of PACC in the cellular extracts was analyzed using polyclonal anti-PACC antibody. The protein was detected in both cellular extracts and, surprisingly, had the same molecular mass (,60 kD), which was similar to that of the intermediate PacC form from A. nidulans grown at alkaline pH (53 kD) ( Figure 7B). A single PACC form was identified in the cellular extract prepared up to 4 h after the cells were shifted to alkaline pH. There was little protein induction in extracts from 1 h after shifting. The processed PacC 27 protein described in A. nidulans grown at alkaline pH was not detected in N. crassa.
Chromatin immunoprecipitation (ChIP-PCR) experiments in the wild-type strain were performed to confirm that N. crassa PACC binds to the gsn promoter in vivo and to assess whether protein binding was pH-dependent. Figure 7C shows that endogenous PACC was associated with the gsn promoter region containing the PacC motif. In addition, protein association was observed with genomic DNA extracted from cells grown at normal pH (5.8) and from cells 2 h after shifting to alkaline pH. Together, these results indicate that N. crassa PACC specifically associates with the gsn promoter even at pH 5.8, the pH for normal growth, thus confirming that the protein is active at physiological pH.

Discussion
Extracellular pH has an important role in cell biology as it regulates gene expression and consequently influences a variety of cellular processes, such as growth, differentiation and develop- Figure 4. gsn, gpn and pacC gene expression during acid and alkaline pH stress. Cells from the wild-type and pacC KO strains were cultivated at pH 5.8 for 24 h and shifted to pH 4.2 and pH 7.8. Samples were collected and used to extract total RNA. Total RNA (15 mg) was separated by electrophoresis in a denaturing formaldehyde gel, transferred to nylon membrane and probed with a-32 P-radiolabeled 678 bp gsn cDNA, or 798 bp gpn cDNA or 639 bp pacC cDNA fragments (gel autoradiographies). The 28 S rRNA was used as a loading control after ethidium bromide staining. The results shown are the average of at least three independent experiments. (A) Analysis of the gsn, gpn and pacC genes in the wild-type strain at different times after pH shifting. After pH stress the remaining cultures were transferred back to physiological conditions (RE, recuperation, pH 5.8) and samples were collected. (B) Analysis of the gsn and gpn genes in the pacC KO strain compared to the wild-type strain at different times after pH shifting. 0, cell samples before pH shifting (control). doi:10.1371/journal.pone.0044258.g004 ment. The regulation of gene expression by pH has been widely investigated in microorganisms because of the great variations in ambient pH that they face in their natural environments. The ambient pH signaling pathway has been exhaustively studied in the filamentous fungus A. nidulans and in several yeasts, including S. cerevisiae. In A. nidulans, pH regulation is mediated by the action of a cascade of proteins (pal gene products) that leads to activation of the transcription factor PacC. This protein, which has a central role in the pH signaling pathway, is activated at alkaline pH after two proteolytic cleavage steps and acts as a transcriptional activator of alkaline-expressed genes (reviewed in [23]).
In S. cerevisiae, the pathway leading to activation of the Rim101p orthologue shares similarities with that of A. nidulans PacC.
However the mechanism of alkaline pH adaptation is not exactly the same since Rim101p requires only a single cleavage step to be activated and, whereas PacC acts either as an activator or repressor, Rim101p is a repressor of the genes encoding the transcriptional repressors Nrg1p and Smp1p at alkaline pH. Nrg1p is required for the repression of alkaline pH-inducible genes, including the gene encoding Ena1 Na + -ATPase, a protein essential for growth at alkaline pH [15]. Thus, Rim101p is indirectly required for S. cerevisiae adaptation to alkaline stress. In addition to Rim101p-mediated alkaline pH regulation, there are also Rim101p-independent mechanisms involved in pH regulation in S. cerevisiae. One is the calcineurin pathway, the activation of which affects the expression of a number of genes through dephosphorylation of the transcription factor Crz1 [24]. High pH stress triggers a transient rise in cytoplasmic calcium resulting in calcineurin activation and ENA1 induction through Crz1 [25]. More recently, Platara et al. [26] described another signaling pathway involving the protein kinase Snf1p that acts through the Mig2p repressor, i.e., a third regulatory pathway involved in the response to high pH stress.
The results described here indicate a role for N. crassa PACC as a regulator of glycogen metabolism through its ability to target GSN via a PacC binding site in the gsn promoter. The pacC KO strain showed impaired growth at alkaline pH, in agreement with the better growth of N. crassa at acidic (5.8) compared to alkaline (7.8) pH. Our findings suggest the existence of a different mechanism of PACC activation under alkaline pH stress when compared to A. nidulans and S. cerevisiae.

PACC Regulates Glycogen Levels Under pH Alkaline Stress through Regulation of gsn Expression
We investigated the glycogen levels and gsn expression in response to extracellular pH changes. Our results provide evidence that both processes are under pH control in N. crassa. Cells of the wild-type strain grown at alkaline pH showed low intracellular glycogen accumulation and a reduction in gsn transcript levels. In contrast, the pacC KO strain accumulated similar amounts of glycogen when compared to the non-stressed wild-type strain and showed high gsn transcript levels at alkaline pH. These results are sufficient to characterize gsn as a pH-regulated gene, and that it is an acid-specific gene. Many genes have been reported to be regulated by ambient pH [27]. In the case of gsn, the finding that wild-type and pacC KO strains had similar amounts of glycogen at normal and alkaline pH, respectively, reinforced the criteria for a gene to be regulated by ambient pH [28]. The analysis of pacC gene expression in N. crassa showed that it was under pH control, as also described for A. nidulans [9] and other filamentous fungi [29][30][31]. This observation supports the fact that, at alkaline pH, the over-expression of pacC controls glycogen levels by regulating gsn, but not gpn, expression.
The patterns of glycogen accumulation and gene expression under combined pH and heat stress were very interesting. Whereas in the wild-type strain glycogen and gsn transcript levels returned to normal, this strain was unable to recover normal levels when exposed concomitantly to heat and alkaline stress. In addition, pacC was not induced under these conditions. The presence of STRE and HSE motifs in the pacC promoter suggests that this gene is regulated by heat shock. As with gsn downregulation under heat stress [5,6], pacC is another example of a gene in N. crassa having STRE and HSE binding sites in its 59flanking region being down-regulated at the transcriptional level by heat stress. The gsn promoter has one STRE motif, which was demonstrated to be involved in the gene transcription modulation when cells were exposed to heat shock (transferred from 30uC to 45uC) [6].
In S. cerevisiae the STREs are described as cis regulatory motif that mediates the transcriptional activation of genes through the action of the stress response regulator proteins Msn2p/4p [32]. Orthologous proteins have not been identified in the N. crassa genome database, suggesting that in this organism STRE may modulate gene expression by an alternative mechanism [33]. The activation of protein kinase A (PKA) leads to Msn2/4p phosphorylation and its cytoplasmic accumulation, thereby preventing the transcription of STRE-regulated genes [34]. Recently, Casado et al. [35] demonstrated that the exposure of S. cerevisiae to alkaline medium resulted in a decrease in cAMP that led to inhibition of the PKA pathway and the accumulation of Msn2/4p in the nucleus. These data indicate that the adaptative response to alkaline pH involves PKA-regulated Msn2/4p-mediated gene remodeling. Since N. crassa lacks Msn2/4p orthologues we PACC shows the same molecular mass at pH 5.8 and pH 7.8. Crude cellular extracts prepared from wild-type N. crassa submitted or not to alkaline pH stress were analyzed by Western blotting using a polyclonal anti-PACC antibody. The protein a-tubulin (theoretical molecular mass 50 kD) was used as a loading control. (C) Chromatin immunoprecipitation assay using the polyclonal anti-PACC antibody. Genomic DNA samples from wild-type N. crassa submitted or not to pH stress were immunoprecipitated with the anti-PACC antibody and subjected to PCR to amplify a 146 bp DNA fragment of the gsn promoter containing the pacC motif. A plasmid construction containing the entire sequence of the gsn gene, including its 59-and 39flanking regions, was used as a positive control. As a negative control, the immunoprecipitation reactions were performed without the anti-PACC antibody. L, 1 kb DNA ladder. doi:10.1371/journal.pone.0044258.g007 hypothesize that the down-regulation of pacC under heat shock and alkaline pH results from the activation of proteins that mediate gene transcription regulation under heat stress through STRE. Whether the PKA signaling pathway is involved in this process remains to be demonstrated. We have previously reported that a N. crassa mutant strain with an inactive PKA pathway showed hyper-accumulation of glycogen during growth when compared to the wild-type strain [7]. However, the gsn transcript level was not increased in this strain; rather, GSN was less phosphorylated in vitro and therefore more active, suggesting that GSN post-translational modification is likely to be the main mechanism controlling glycogen accumulation during vegetative growth when the PKA pathway is inactive.
PACC Binds to the gsn Promoter in vitro and in vivo and is Active at Normal and Alkaline pH DNA shift experiments showed that truncated PACC having the DNA-binding domain was able to bind to a DNA fragment from the gsn promoter containing the PacC binding site. The DNA-binding reaction was specific, based on three criteria: (1) binding was reduced when a DNA fragment mutated in the PacC core sequence was used, (2) a DNA oligonucleotide containing the PacC binding site competed in the binding reactions by reducing the DNA-protein complexes, and (3) anti-PACC antibody led to a DNA-protein supershift. Together, these findings revealed that the PacC binding site in the gsn promoter was the target for the binding of recombinant PACC. In our results, it was observed binding reduction when the mutated probe was used, instead of no binding, which was completely abolished in the presence of the DNA oligonucleotide as competitor. We suggested either that the core sequence neighborhoods are important for PACC recognition or the N. crassa PACC has an alternative binding site similar to what was described for Rim101p in C. albicans [36]. DNA-protein complexes were also observed when crude cellular extracts were used as protein sources. Interestingly, proteins from extracts prepared from wild-type cells grown at normal and alkaline pH were able to bind to the DNA fragment in vitro and in vivo, and the PACC present in these extracts had the same molecular mass (,60 kD). According to the model for pH-mediated regulation of gene expression in A. nidulans [23] full-length PacC (PacC 72 ) is synthesized as an inactive form and an alkaline signal is transduced by the pal signaling pathway, thereby activating PacC 72 to PacC 53 and PacC 27 by two successive proteolytic steps. The S. cerevisiae Rim101p is similar in size to the A. nidulans protein but they differ in that the yeast protein is proteolytically processed at the Cterminus by a single cleavage step and the smaller size form (similar to PacC 27 ) has not been detected [16]. The results described here suggest that N. crassa PACC is subject to only one proteolytic processing step since the molecular mass of the protein present in cells grown at alkaline pH was lower than the theoretical value of 67 kD. According to Peñas et al. [37], N. crassa PACC has a conserved protease box that may lead to a processed form of PACC with a molecular mass of ,55 kD, the size found in our experiments. The most intriguing result concerning PACC was that the forms detected before and after pH stress had the same molecular mass, which indicates that alkaline pH is not the signal for proteolytic processing. Rather, the protein is already processed during growth at pH 5.8.
The finding that N. crassa has orthologues to all six Pal proteins, including PalH which, together with PalI, function as membrane pH sensors in A. nidulans [23], is particularly interesting. According to the model proposed for A. nidulans, the pH signal is transduced to downstream components by PalF phosphorylation and ubiquitination, leading to endocytosis of the PalF/PalH complex [38]. By overexpressing PalF covalently attached to ubiquitin in a null palH background Hervás-Aguilar et al. [39] have shown that PalF ubiquitination activates the signaling pathway under acidic conditions. The authors concluded that ubiquitination of PalF commands the cascade of intracellular events that mediate fungal adaptation to environmental pH.
Our results clearly demonstrated that in N. crassa ambient pH controls glycogen accumulation by regulating gsn expression, i.e., by regulating glycogen synthesis and not degradation. Although the data point to a central role for PACC many questions remain unanswered. One major concern is the fact that PACC shows the same molecular mass before and after alkaline pH stress, indicating that if there is any processing it must occur independently of ambient pH. Another question relates to the in vivo results showing that endogenous PACC binds to the gsn promoter even at pH 5.8. If there is protein association at both pH values, how are the differences in gsn expression and glycogen content to be explained? Since N. crassa has all six Pal protein orthologues we might speculate that PalH may function as a pH sensor, as in A. nidulans, and that all six Pal homologues are active components of the pH signaling pathway. Finally, is PACC the only regulator of glycogen accumulation and gsn expression at alkaline pH? Other molecular mechanisms are known to be involved in adaptation to environmental pH in S. cerevisiae and C. albicans [40]. The results described here reveal interesting differences between N. crassa and A. nidulans with regard to the pH signaling pathway that deserve to be investigated.
Conidia from 10 day cultures of wild-type and pacC KO strains were suspended in sterile water and counted. For radial growth analysis, 10 7 conidia/mL were inoculated onto plates containing solid VM medium and VM medium supplemented with yeast extract at pH 4.2, 5.8 and 7.8 for 24 h. Images of colony morphology were captured after 24 h using an AxioCam ICc3 coupled to the stereoscope trinocular Discovery V8 (Zeiss) at 80 x magnification. For linear growth, conidia were inoculated in race tubes containing VM medium at pH 4.2, 5.8 and 7.8 and incubated at room temperature. Measurements were made every 24 h.
For pH stress experiments, 10 9 conidia/mL were first germinated in 4 L of VM medium (pH 5.8) at 30uC, 250 rpm, for 24 h. After this period, the culture was filtered and the mycelia divided into three samples. One was frozen in liquid nitrogen and stored at 280uC for further processing (control sample, not submitted to stress) while the remaining two samples were transferred into 1 L of fresh VM medium containing 0.5% sucrose at pH 4.2 (for acid stress) and 7.8 (for alkaline stress). Samples (125 mL) from mycelia submitted to both pH conditions were harvested after different periods of incubation. For the experiments combining pH and heat stress, mycelia obtained from the wild-type strain after 24 h were filtered and divided into three samples that were transferred into 1 L of fresh VM medium containing 0.5% sucrose at pH 4.2, 5.8 and 7.8 preheated to 45uC. Samples (125 mL) were harvested after 30 min incubation and processed as before.
In both stress experiment samples from the wild-type strain were subjected to the recuperation conditions. The remaining mycelia from the pH stress experiments were filtered and transferred back into 400 mL of fresh VM medium at pH 5.8 and 30uC. Samples were collected after 30, 60 and 120 min of recuperation. For experiments combining pH and heat stress the remaining mycelia were transferred back into 400 mL of fresh VM medium pre-heated at 30uC at the three different pH. Cells from 120 mL samples were harvested after 30, 60 and 120 min of recuperation and processed. The mycelial samples were used for glycogen quantification and RNA extraction.

Generation of a pacC KO Mutant by Gene Replacement
A strain bearing a deletion in the pacC gene was generated as described by Ninomiya et al. [21]. Briefly, three PCR fragments were produced using the following pairs of primers: fragment 1, primers 0090-F1 and 0090-R2, fragment 2, 0090-F2 and 0090-R3, and fragment 3, primers 0090-F3 and 0090-R1 (Table 1). Fragments 1 and 3 contained a 1.5 kb upstream and downstream of the ORF NCU00090 (pacC, 1,866 bp), respectively, with an additional 20 bp overlapping sequence from the bar gene. Both fragments were amplified from wild-type strain genomic DNA. Fragment 2 contained the bar ORF (1 kb) plus overlapping sequences from the pacC gene at both ends and was amplified from the pIV9A-1 plasmid [6]. All three fragments were purified using a Qiagen gel purification kit, and 100 ng of each fragment was mixed and used to amplify a single fragment (4 kb) with the primers 0090-F1 and 0090-R1. All PCRs were done using Phusion High Fidelity Taq polymerase (NEB). The final PCR fragment was purified and used to transform 10 day conidia from the strain FGSC#9568. Transformants were selected by plating in medium containing Basta (200 mg/mL). Correct integration of the fragment into the genome was detected by PCR using primers 0090-F1 and 0090-R4 that annealed downstream from the amplified fragment. One transformant containing the right integration was backcrossed with FGSC#2490 to generate a homokaryotic strain.

pacC cDNA Cloning and Production and Purification of the Recombinant Protein
The pacC gene encodes a 621 amino acid protein with a theoretical molecular mass of 67 kD. A pacC cDNA fragment (639 bp) encoding the N-terminal region containing the zinc finger C 2 H 2 DNA-binding domain was amplified from the cDNA plasmid library pYADE5 [43] with the oligonucleotides 90-F and 90-R2 (Table 1), as previously described [8]. For expression of the His-DPACC recombinant protein the E. coli BL21(DE3) pLysS strain harboring the pET-DPACC plasmid was used. A truncated protein encompassing amino acids 1 to 213 fused to an N-terminal His 6 -tag was produced. Cells were grown at 37uC in 1 L of LB medium to an OD 600 of 0.7 and induced with IPTG (final concentration 0.4 mM) for 4 h at 37uC. The cells were harvested by centrifugation, suspended in buffer A (50 mM Tris-HCl, pH 8.0, 300 mM NaCl, 20 mM imidazole, 10% v/v glycerol, 1 mM DTT, 0.5 mM PMSF, 25 mM benzamidine and 5 mg/mL each of antipain, leupeptin and pepstatin A) containing 0.5% Triton X-100 and 0.5% Tween 20 and lysed by sonication (5 cycles of 30 s sonication and 30 s on ice). After centrifugation, the supernatant was subjected to affinity chromatography on a HisTrap HP column (GE Healthcare) using an Ä KTA Prime purification system. The recombinant protein was eluted with a linear gradient of imidazole (20-500 mM) in buffer A and dialyzed two times against 1 L of dialysis buffer (10 mM Tris-HCl, pH 7.9, 100 mM KCl, 10% v/v glycerol, 1 mM EDTA and 0.5 mM DTT). The purified protein was analyzed by SDS-PAGE followed by Coomassie Brilliant Blue staining [44] and was quantified by the Hartree [45] method using BSA as standard. The purified His-DPacC protein was used to raise antibodies in rabbits.

Glycogen Quantification and Gene Expression Assays
Mycelial pads were ground to a fine powder in a pre-chilled mortar in liquid nitrogen and extracted with lysis buffer (50 mM Tris-HCl, pH 7.6, 100 mM NaF, 1 mM EDTA, 1 mM PMSF, 0.1 mM TCLK, 1 mM benzamidine and 1 mg/mL each of pepstatin and aprotinin). Cell extracts were clarified by centrifugation (3,0006g, 10 min, 4uC) and the supernatants were used for glycogen and protein quantification. Glycogen was extracted by digestion with amyloglucosidase (30 mg/mL) and a-amylase (10 mg/mL) after ethanol precipitation, as previously described [7]. Free glucose was measured with a glucose oxidase kit and the glycogen content was normalized to the total protein concentration. Total protein was quantified by the Hartree method [45].
For gene expression analysis by Northern blotting, total RNA was extracted with LiCl [46] from mycelial samples subjected to pH and combined pH and temperature stress. Total RNA (15 mg) was electrophoresed on a 1.2% agarose-formaldehyde denaturing gel [47] at 65 V for 5 h and then transferred to neutral nylon membranes (Hybond-N, GE Healthcare) in 2 x SSC. The blots were probed either with a 678 bp gsn cDNA, or 798 bp gpn cDNA, or 639 bp pacC cDNA fragments (10 6 -10 8 cpm) radiolabelled with a-32 P-dATP (3,000 mCi/mmol) by random priming (NEBlot kit, Biolabs) in 5-10 mL of ULTRAhyb hybridization solution (Ambion) at 42uC overnight. After hybridization, the blot was washed twice in 2 x SSC containing 0.1% SDS for 20 min, and twice in 0.1 x SSC containing 0.1% SDS for 20 min followed by exposure to an X-ray film.
The pacC gene was also analyzed at protein level by Western blot. Total proteins from cells of the wild-type strain cultured at pH 5.8 for 24 h (control) and that subjected to alkaline pH stress (pH 7.8) for 1, 2, 3 and 4 h were prepared as described for glycogen extraction using glass beads. Proteins (100 mg) were separated by 12% SDS-PAGE gels [44] and electrophoretically transferred to a nitrocellulose membrane. Immunoblotting was performed with polyclonal anti-PACC antibody and band intensities were normalized to parallel blots probed with anti-atubulin antibody (Sigma). Blots were subsequently probed with HRP-conjugated secondary antibodies (Sigma) and developed with luminol reagent.
In some experiments, 35 mg of protein from chromatographic fractions (obtained as described below) was used. A radiolabeled DNA probe (,10 4 cpm) was added and the reactions were incubated at room temperature for 20 min. Free probe was separated from DNA-protein complexes by electrophoresis on a native 5% polyacrylamide gel in 0.5 x?TBE buffer at 300 V, 10 mA and 10uC. After electrophoresis, the gel was dried and autoradiographed. For competition assays, an excess of specific DNA competitor was added to the binding reactions 10 min prior to incubation with the radiolabeled probe.
For the supershift experiments, 5 mg of recombinant His-DPACC protein was incubated with an excess of polyclonal anti-PACC antibody (title 1:500) for 10 min on ice prior to the addition of radiolabeled pacC probe and further 20 min incubation at room temperature.

Preparation and Fractionation of Crude Cellular Extract for EMSA
Cells from the wild-type and pacC KO strains exposed to alkaline stress (pH 7.8, 60 min) or not (pH 5.8, control) were used to prepare crude cellular extracts. About 10 mg of frozen mycelia were ground to a fine powder in liquid nitrogen in a pre-chilled mortar, homogenized in 20 mL of lysis buffer (15 mM HEPES-KOH, pH 7.9, 10% v/v glycerol, 500 mM KCl, 5 mM MgCl 2 , 0.5 mM EDTA, 1 mM DTT, 0.5 mM PMSF, 25 mM benzamidine, 50 mM NaF and 10 mg/mL each of antipain and pepstatin A) and stirred with glass beads in eight cycles that consisted of 30 s of stirring and 30 s on ice. Crude cellular extract was recovered after centrifugation (3,200 6g, 2 min, 4uC), dialyzed against buffer D (15 mM HEPES-KOH, pH 7.9, 15% v/v glycerol, 100 mM KCl, 1 mM EDTA) at 4uC for 2 h, and cleared by centrifugation (20,000 6 g, 20 min, 4uC) before loading onto a Heparin-Sepharose FF column (GE Healthcare). The proteins were eluted with a 0.1-1.5 mM KCl linear gradient and the protein fractions were dialyzed against buffer D containing 0.5 mM PMSF, 25 mM benzamidine and 50 mM NaF and then frozen in liquid nitrogen and storage at 280uC. Total protein was quantified by the Hartree method [45]. Proteins from the chromatographic fractions (35 mg of total protein) were analyzed for their ability to bind to the DNA fragment of the gsn promoter containing the pacC binding site.

DNA Probe and Competitors for EMSA
A putative cis PacC motif (59-GCCAAG-39) was identified in the gsn gene 5'-flanking region (starting at nucleotide 21807) by analysis in silico using the MatInspector tool (www.genomatix.de). To produce the pacC probe, a 146 bp DNA fragment containing the PacC motif was amplified from the pIV9A-1 plasmid (GenBank#AF417205) using oligonucleotides PacC-F and GSN-RP3 (Table 1) in the presence of a-32 P-dATP (3,000 Ci/mmol) and purified on 2% low-melting point agarose gel. A mutant probe (mpacC) was prepared by changing the element core sequence from 59-CTTGGC-39 to 59-AAATTA-39 by site-directed mutagenesis in a two-step PCR [48]. The oligonucleotide pairs mPacC-F/ GSN-RP2 and pGSN-F/mPacC-R (Table 1) were used in the first reaction to amplify two fragments, and the oligonucleotide pair pGSN-F/GNS-RP2 was used in a second reaction to amplify the whole DNA fragment containing the mutation. The 311 bp fragment was subcloned into the vector pMOS-Blue (GE Healthcare) and the desired mutation was confirmed by DNA sequencing. For EMSA, the mutant probe (mpacC) was amplified by PCR in the presence of a-32 P-dATP using the oligonucleotides PacC-F and GSN-RP3 and then purified.
The unlabeled 146 bp pacC probe was used as a specific DNA competitor. A 27 bp DNA oligonucleotide was also used as a competitor after annealing the complementary oligonucleotides OligoPacC-F and OligoPacC-R ( Table 1). The specific competitors were quantified by measuring the absorbance at 260 nm and then added to the binding reaction in a 15-to 40-fold molar excess for the specific competitor and a 1-to 15-fold molar excess for the DNA oligonucleotide.  ChIP-PCR Analysis ChIP assays were performed according to Tamaru et al. [49], with modifications. Briefly, the wild-type strain was grown in liquid VM medium at 30uC, 250 rpm, for 24 h, and subjected to pH stress as previously described. Mycelial samples (125 mL aliquots) were subsequently transferred into 500 mL flasks and the chromatin was fixed by adding formaldehyde to 1% final concentration followed by incubation for 30 min at 30uC and 250 rpm. Formaldehyde was quenched by adding 125 mM glycine to each sample and then incubating at 30uC, 250 rpm, for 10 min. Chromatin prepared from each sample was precleared with normal rabbit IgG (Upstate) and protein A agarose (Sigma) 50% slurry pre-blocked with sonicated salmon-sperm DNA and then immunoprecipitated with polyclonal anti-PACC antibody and protein A-agarose. As a negative control, a reaction without anti-PACC antibody was done. The 146 bp DNA fragment associated with the PACC protein was amplified by PCR using the oligonucleotides PacC-F and GSN-RP3 and the reaction products were analyzed on a 2% agarose gel. A plasmid containing the entire sequence of the gsn gene, including its 59-and 39-flanking regions, was used as a positive control for PCR.