Improved α-Amylase Production by Dephosphorylation Mutation of CreD, an Arrestin-Like Protein Required for Glucose-Induced Endocytosis of Maltose Permease and Carbon Catabolite Derepression in Aspergillus oryzae

ABSTRACT

Aspergillus oryzae produces copious amount of amylolytic enzymes, and MalP, a major maltose permease, is required for the expression of amylase-encoding genes. The expression of these genes is strongly repressed by carbon catabolite repression (CCR) in the presence of glucose. MalP is transported from the plasma membrane to the vacuole by endocytosis, which requires the homolog of E6-AP carboxyl terminus ubiquitin ligase HulA, an ortholog of yeast Rsp5. In yeast, arrestin-like proteins mediate endocytosis as adaptors of Rsp5 and transporters. In the present study, we examined the involvement of CreD, an arrestin-like protein, in glucose-induced MalP endocytosis and CCR of amylase-encoding genes. Deletion of creD inhibited the glucose-induced endocytosis of MalP, and CreD showed physical interaction with HulA. Phosphorylation of CreD was detected by Western blotting, and two serine residues were determined as the putative phosphorylation sites. However, the phosphorylation state of the serine residues did not regulate MalP endocytosis and its interaction with HulA. Although α-amylase production was significantly repressed by creD deletion, both phosphorylation and dephosphorylation mimics of CreD had a negligible effect on α-amylase activity. Interestingly, dephosphorylation of CreD was required for CCR relief of amylase genes that was triggered by disruption of the deubiquitinating enzyme-encoding gene creB. The α-amylase activity of the creB mutant was 1.6-fold higher than that of the wild type, and the dephosphorylation mimic of CreD further improved the α-amylase activity by 2.6-fold. These results indicate that a combination of the dephosphorylation mutation of CreD and creB disruption increased the production of amylolytic enzymes in A. oryzae.

IMPORTANCE In eukaryotes, glucose induces carbon catabolite repression (CCR) and proteolytic degradation of plasma membrane transporters via endocytosis. Glucose-induced endocytosis of transporters is mediated by their ubiquitination, and arrestin-like proteins act as adaptors of transporters and ubiquitin ligases. In this study, we showed that CreD, an arrestin-like protein, is involved in glucose-induced endocytosis of maltose permease and carbon catabolite derepression of amylase gene expression in Aspergillus oryzae. Dephosphorylation of CreD was required for CCR relief triggered by the disruption of creB, which encodes a deubiquitinating enzyme; a combination of the phosphorylation-defective mutation of CreD and creB disruption dramatically improved α-amylase production. This study shows the dual function of an arrestin-like protein and provides a novel approach for improving the production of amylolytic enzymes in A. oryzae.

INTRODUCTION

The amount of plasma membrane proteins involved in nutrient uptake and signal sensing is stringently controlled by eukaryotic cells for adaptation to environmental changes. In Saccharomyces cerevisiae, transporters of amino acids, metals, and sugars are sorted from the plasma membrane to the vacuole for degradation in response to a change in nutrients. This sorting is mediated by endocytosis, and ubiquitination of transporters by Rsp5, the homolog of E6-AP carboxyl terminus-type ubiquitin ligase, is required for this endocytosis (1).

A filamentous fungus, Aspergillus oryzae, is used for the production of traditional Japanese fermented foods such as sake, soy sauce, and miso (soybean paste) (2). A. oryzae secretes copious amounts of amylolytic enzymes, and maltose utilization is important for induction of amylase gene expression (3). In previous studies, we found that a major A. oryzae maltose permease (MalP) of the maltose utilization (MAL) cluster is required for α-amylase production (4, 5). In the presence of glucose, the expression of the amylase-encoding genes and malP was strongly repressed by carbon catabolite repression (CCR), which is mediated by the transcription factor CreA (4, 6). Although MalP was localized at the plasma membrane in the presence of maltose, glucose and other CCR-inducing sugars triggered rapid trafficking of MalP from the plasma membrane to the vacuole (5). This internalization depended on endocytosis, which was inhibited by repression of the ubiquitin ligase HulA (ortholog of S. cerevisiae Rsp5) (5). These results suggested that ubiquitination of MalP by HulA is required for MalP endocytosis. Rsp5 interacts with the proline-rich (PY) motifs of target proteins via its WW domains; however, most yeast transporters and MalP possess no PY motif. This implies the presence of an adaptor protein for mediating the interaction of HulA with MalP.

In S. cerevisiae, glucose induces the Rsp5-mediated endocytosis of various transporters, such as the maltose transporter (Mal61), high-affinity glucose transporter (Hxt6), and monocarboxylic acid transporter (Jen1) (7–11). Eleven arrestin-related trafficking (ART) proteins have been identified as adaptor proteins of Rsp5 and membrane proteins (12, 13), and Rod1/Art4, an ART protein, is an adaptor of Jen1 and Rsp5 (11). The phosphorylation state of Rod1/Art4 is controlled by the Snf1 kinase and PP1 phosphatase, and glucose-induced dephosphorylation of Rod1/Art4 was required for Jen1 endocytosis (11).

In filamentous fungi, 10 putative ART proteins were found in the Aspergillus nidulans genome, and a single ART protein, ArtA, was involved in the ammonium- or substrate-induced endocytosis of transporters of uric acid—xanthine (UapA), l-proline (PrnB), and purines (AzgA) (14). However, the other nine putative ART proteins are not required for UapA endocytosis, and their functions have not been investigated, except for PalF, which is involved in signaling to alkaline pH (15–17). Interestingly, CreD, one of the putative ART proteins, was originally identified as a repressor of carbon catabolite derepression in the creB mutant, which encodes a ubiquitin-processing protease (18, 19). In addition, the amino acid sequence of CreD is 55% similar to that of Rod1/Art4, and interaction between CreD and WW domains of HulA in bacterial two-hybrid assay has been reported (20). Therefore, we hypothesized that CreD serves as an adaptor protein of MalP and HulA for glucose-induced endocytosis in A. oryzae.

In this study, we investigated the involvement of CreD in glucose-induced endocytosis of MalP and CCR relief of amylase-encoding genes. In addition, we identified two serine residues as putative phosphorylation sites of CreD and further examined the involvement of CreD phosphorylation/dephosphorylation in the regulation of MalP endocytosis and CCR relief of amylase-encoding genes.

RESULTS

Involvement of CreD in the glucose-induced endocytosis of MalP.

A. oryzae creD (FungiDB accession no. AO090003000144) is predicted to encode a protein of 603 amino acids with 79% amino acid identity to A. nidulans CreD in the Aspergillus genome database (http://www.aspgd.org/). A. oryzae CreD contains N- and C-terminal arrestin-like domains at its N terminus and four PY (one PPxY and three PxY) motifs in the C-terminal region (Fig. 1A). To examine the involvement of CreD in MalP endocytosis, we constructed a creD deletion mutant with the ΔligD::ptrA mutant (21) as the host strain (see Fig. S1 in the supplemental material). The creD deletion mutant showed an approximately 20% smaller colony size than the wild-type strain on minimal agar medium (Fig. 1B). A plasmid expressing GFP-MalP (5) was introduced into the creD deletion mutant and wild-type (ΔligD::ptrA) strains to observe the subcellular localization of MalP. Green fluorescent protein (GFP)-MalP expression was induced by incubating the hyphae in maltose-containing medium, following which the hyphae were transferred to glucose-containing medium and subsequently examined by fluorescence microscopy. The majority of the GFP fluorescence disappeared from the plasma membrane of the wild-type strain after 30 min of incubation in the glucose-containing medium; however, strong GFP fluorescence was observed in the intracellular compartments that were presumed to be vacuoles (Fig. 1C) in accordance with our previous study (5). In contrast, GFP fluorescence was observed at the plasma membrane of the ΔcreD mutant even after 120 min of incubation in the glucose-containing medium (Fig. 1C). To examine the proteolytic degradation of GFP-MalP, intracellular proteins enriched at the membrane were analyzed by Western blot analysis with an anti-GFP antibody. In addition to a signal at approximately 90 kDa corresponding to a molecular mass of intact GFP-MalP, two signals with lower molecular masses emerged in the wild-type strain after the addition of glucose (Fig. 1D). The size of the lower-molecular-weight signal was in close agreement with the molecular mass of free GFP (27 kDa), which is known to be highly resistant to vacuolar proteases. The second signal at approximately 37 kDa was presumed to be an intermediate degradation product of GFP-MalP. In contrast, these two signals were not observed in the ΔcreD mutant (Fig. 1D). These results indicated that CreD plays an important role in glucose-induced degradation of MalP through endocytosis.

FIG 1.

Effect of creD deletion on GFP-MalP endocytosis and degradation after glucose addition. (A) Schematic representation of the primary structure of A. oryzae CreD. The single PPxY and three PxY motifs are represented as black and white bars, respectively. (B) Growth of the creD deletion mutant. Approximately 1 × 10⁴ conidiospores of the wild type (ΔligD::ptrA) and the ΔcreD mutant were grown on a minimal agar plate containing 1% glucose as the sole carbon source at 30°C for 4 days. (C) Subcellular localization analysis of GFP-MalP. Hyphae of GFP-MalP-expressing strains were grown in MM containing 1% glycerol as the sole carbon source for 20 h and transferred to fresh MM containing 1% maltose. After incubation for 3 h, hyphae were immersed in fresh MM containing 0.5% glucose as the sole carbon source and GFP fluorescence was observed at the time points indicated. DIC, differential interference contrast. (D) Western blot analysis of GFP-MalP with anti-GFP antibody. Mycelia grown for 24 h in MM plus 0.1% polypeptone containing 1% glycerol were transferred to fresh MM containing 1% maltose. After incubation for 90 min, glucose was added to the medium at a final concentration of 0.5%. The intracellular proteins were extracted from the mycelia harvested at the time points indicated after glucose addition and used for Western blot analysis. The arrow indicates an intermediate degradation product of GFP-MalP.

Involvement of CreD in carbon catabolite derepression.

A. nidulans CreD was originally identified as a repressor of carbon catabolite derepression in the mutant lacking creB, the gene encoding a ubiquitin protease (18, 19). This suggested that CreD is involved in carbon catabolite derepression. We measured the α-amylase activity of the wild type and the ΔcreD mutant after cultivation in liquid medium containing maltose with or without glucose to investigate the role of CreD in carbon catabolite derepression (Fig. 2). The ΔcreD mutant showed 60% lower α-amylase activity than the wild-type strain in maltose medium without glucose. In addition, the ΔcreD mutant showed minimal α-amylase activity in maltose medium containing 1% glucose. These results suggested that CreD plays a critical role in carbon catabolite derepression.

FIG 2.

α-Amylase activity of the ΔcreD mutant. The wild type (WT) (ΔligD::ptrA) and the ΔcreD mutant were grown in MM plus 0.1% polypeptone containing 1% maltose with or without 1% glucose at 30°C for 24 h. The α-amylase activity in the culture broth was divided by the mycelial dry weight. Error bars indicate the standard errors of three independent experiments. The P values were calculated with the unpaired Student t test. **, P < 0.01.

Identification of CreD phosphorylation sites.

To investigate the posttranslational modification of CreD, creD was genetically modified to incorporate three FLAG tags (3×FLAG) at the carboxyl terminus of CreD (see Fig. S2 in the supplemental material). The cell lysate of the mycelium cultured in liquid medium containing glycerol as the sole carbon source was used for Western blot analysis. As shown in Fig. 3A, CreD-3×FLAG was detected as multiple signals, suggesting that a posttranslational modification(s) occurred in FLAG-tagged CreD. The disappearance of higher-molecular-weight signals after treatment of CreD-3×FLAG with alkaline phosphatase revealed that this modification was phosphorylation (Fig. 3A).

FIG 3.

Identification of CreD phosphorylation sites. (A) Alkaline phosphatase treatment of CreD-3×FLAG. The intracellular proteins were extracted from mycelia grown for 24 h in MM plus 0.1% polypeptone containing 1% glycerol as the sole carbon source and treated with calf intestine alkaline phosphatase (CIAP). Approximately 10 μg of intracellular proteins was used for Western blot analysis. (B) Amino acid sequences of a potential Snf1 recognition motif located within CreD. Ø denotes hydrophobic amino acids. (C) Western blot analysis of CreD-3×FLAG containing mutations in the potential Snf1 recognition motif. The cell lysate of the mycelium grown for 24 h in MM plus 0.1% polypeptone containing 1% glycerol as the sole carbon source was used for Western blot analysis. PgkA was used as a loading control. WT, wild type.

Yeast Rod1/Art4 is phosphorylated by the Snf1 kinase at Ser 447 (22). In A. oryzae CreD, two serine residues at positions 402 and 515 corresponded to the consensus sequence of the Snf1 kinase's substrate recognition site (Fig. 3B). Therefore, we replaced these two serine residues with alanine and assessed the phosphorylation status of tagged CreD. As shown in Fig. 3C, phosphorylation of CreD-3×FLAG was abolished in these double mutants, whereas single serine mutants retained the phosphorylated forms. This result suggested that CreD was phosphorylated at the serine residues at positions 402 and 515.

Interaction of CreD with HulA.

Similar to S. cerevisiae Rsp5, A. oryzae HulA possesses three tandem WW domains. To investigate the interaction of these domains with CreD, the WW domains were expressed in Escherichia coli as a fusion protein with glutathione S-transferase (GST) and subjected to a pulldown assay. As shown in Fig. 4A, CreD-3×FLAG was observed only in the eluted sample of the fusion protein (GST-WW), whereas no signal was detected in the eluted sample of GST alone. This clearly indicated that the tandem WW domains of HulA interacted with CreD-3×FLAG. To confirm the interaction between HulA and CreD in vivo, three-hemagglutinin (3×HA)-tagged HulA was coexpressed with CreD-3×FLAG in A. oryzae and the cell lysate was subjected to coimmunoprecipitation (Co-IP) analysis. We used the enoA promoter to express 3×HA-HulA, since the latter was difficult to detect clearly by Western blotting when expressed from its own promoter. Detection of CreD-3×FLAG in the elution samples obtained after the immunoprecipitation of 3×HA-HulA indicated that 3×HA-HulA interacted with CreD-3×FLAG (Fig. 4B). To investigate the involvement of phosphorylation and/or dephosphorylation of CreD in its interaction with HulA, both of the phosphorylation sites of CreD were replaced with either glutamic acid or alanine to mimic phosphorylation and dephosphorylation, respectively, of the serine residues. However, interaction between 3×HA-HulA and CreD-3×FLAG was not inhibited by either mutation (Fig. 4B). These results suggested that the phosphorylation state of CreD was not involved in the regulation of its interaction with HulA.

FIG 4.

Analysis of the physical interaction between HulA and CreD. (A) GST pulldown assay. GST alone and the fusion protein with the HulA WW domains (GST-WW) were purified from E. coli and incubated with the cell lysate of mycelia grown for 24 h in MM plus 0.1% polypeptone containing 1% glucose. The eluted sample from glutathione Sepharose was subjected to Western blot analysis and Coomassie brilliant blue (CBB) staining to detect purified GST proteins and bound CreD-3×FLAG, respectively. (B) Co-IP analysis of 3×HA-HulA and CreD-3×FLAG. The cell lysate of mycelia grown for 24 h in MM plus 0.1% polypeptone containing 1% glucose was incubated with anti-HA antibody beads. The cell lysate and eluted sample from the antibody beads were used for Western blot analysis. SA and SE indicate cell lysates from the creD mutants harboring double alanine (SA) and double glutamic acid (SE) substitutions at both phosphorylation sites.

The phosphorylation state of CreD is not involved in MalP endocytosis and CCR in the wild-type strain.

CreA-mediated CCR and rapid endocytosis of MalP were induced by glucose and mannose, but not by maltose and xylose (5). To examine the effects of these sugars on the phosphorylation state of CreD, the CreD-3×FLAG-expressing strain was incubated in minimal medium (MM) containing these sugars. Interestingly, dephosphorylation of CreD-3×FLAG was induced within 10 min after incubation in glucose- and mannose-containing media, whereas maltose and xylose did not induce any dephosphorylation (Fig. 5). This suggested that CreD-3×FLAG is rapidly dephosphorylated under the conditions that induce MalP endocytosis.

FIG 5.

Effects of sugars on the phosphorylation state of CreD. Mycelia grown for 24 h in MM plus 0.1% polypeptone containing 1% glycerol were transferred to fresh MM containing the sugars indicated at 1%. Mycelia were harvested at the time points indicated, and the cell lysate was subjected to Western blot analysis.

Next, we investigated the involvement of CreD dephosphorylation in the endocytosis of MalP by monitoring GFP-MalP fluorescence in the CreD-3×FLAG-S402/515A (SA) and CreD-3×FLAG-S402/515E (SE) mutant strains. In both strains, GFP fluorescence at the plasma membrane disappeared within 30 min after the addition of glucose (Fig. 6A). In addition, Western blot analysis showed that degradation of GFP-MalP was not repressed by either mutation (Fig. 6B). These results suggested that induction of endocytosis was not affected by the phosphorylation state of CreD. To investigate whether the phosphorylation or dephosphorylation of CreD is involved in carbon catabolite derepression, the α-amylase activities of the CreD-3×FLAG-SA and CreD-3×FLAG-SE strains were measured after cultivation in maltose-containing medium in the presence of 1% glucose (Fig. 6C). Although both strains showed slightly less α-amylase activity than the wild-type strain, these reductions were negligible compared to that observed upon creD deletion (Fig. 2). This suggested that the phosphorylation status of CreD does not regulate carbon catabolite derepression in the wild-type strain.

FIG 6.

Effects of CreD phosphorylation and dephosphorylation mimics on MalP endocytosis and α-amylase production. (A) Subcellular localization analysis of GFP-MalP in CreD phosphorylation site mutants (SA, double alanine mutation; SE, double glutamic acid mutation). Hyphal cells of CreD phosphorylation site mutants expressing GFP-MalP were immersed in glucose-containing medium and examined by confocal microscopy as described in the legend to Fig. 1C. DIC, differential interference contrast. (B) Western blot analysis of GFP-MalP in CreD phosphorylation site mutants. Intracellular proteins extracted from the mycelia after glucose addition were used for Western blot analysis as described in the legend to Fig. 1D. The arrowhead indicates an intermediate degradation product of GFP-MalP. (C) α-Amylase activity of CreD phosphorylation site mutants. α-Amylase activity was measured after cultivation in MM plus 0.1% polypeptone containing 1% maltose and 1% glucose at 30°C for 24 h. Error bars indicate the standard errors of three independent experiments. The P values were calculated by the unpaired Student t test. *, P < 0.05; ns, not significant.

Dephosphorylation of CreD is required for CCR relief in the ΔcreB mutant.

Disruption of A. oryzae creB resulted in relief from CCR and increased production of α-amylase (23, 24). We generated the creB and creD double-disruption mutant strain to investigate the involvement of creD in carbon catabolite derepression in the creB disruption mutant (see Fig. S1 in the supplemental material). The ΔcreB mutant formed a clear zone by starch degradation around the colony on a starch agar plate supplemented with glucose, which was completely abolished by creD deletion (Fig. 7A). Northern blot analysis showed that transcriptional repression of α-amylase-encoding genes (amyA, amyB, and amyC) (25) and MAL cluster genes (malP and malT) (4) by glucose was relieved in the ΔcreB mutant, and this repression was partially recovered in the ΔcreB ΔcreD mutant (Fig. 7B). These results indicated that CreD is required for CCR relief triggered by creB disruption. Interestingly, mutation of CreD phosphorylation sites to glutamic acid also repressed the clear zone formation of the ΔcreB mutant on starch agar medium supplemented with glucose (Fig. 7C). In contrast, dephosphorylation mutants of CreD showed larger clear zone formation in the ΔcreB background (Fig. 7C). These mutations had no apparent effect on the clear zone formation observed in creA disruption mutants (see Fig. S3 in the supplemental material). Next, we measured the α-amylase activity of CreD dephosphorylation mutants in a creB disruption background after cultivation in medium containing high concentrations of maltose. A previous study has shown that creB disruption increases α-amylase production under these culture conditions (24). We observed that the α-amylase activity of the creB disruptant was 1.6-fold higher than that of the wild-type strain, and the dephosphorylation mutation of CreD further improved the α-amylase activity to 2.6-fold, whereas this mutation had no apparent effect on α-amylase production in the wild-type strain (Fig. 7D). These results suggested that dephosphorylation of CreD is required for CCR relief stimulated by creB disruption and that the dephosphorylation mutation promoted α-amylase production in the creB disruption mutant.

FIG 7.

Involvement of CreD dephosphorylation in the CCR relief stimulated by creB disruption. (A) Clear zone formation by the creB and creD double-disruption mutant on starch agar medium containing glucose. Approximately 1 × 10⁴ conidiospores of each strain were grown on minimal agar medium containing 1% starch and 1% glucose at 30°C for 2 days. Triton X-100 at a final concentration of 0.25% was added to the medium to clearly distinguish the clear zones. (B) Northern blot analysis of α-amylase-encoding genes and MAL cluster genes. Mycelia grown for 24 h in MM plus 0.1% polypeptone containing 1% glycerol were transferred to fresh MM containing 1% maltose with or without 1% glucose supplementation. Total RNA was extracted from mycelia harvested at the time points indicated. Approximately 20 μg of total RNA was subjected to Northern blot analysis, and digoxigenin-labeled fragments of each gene were used as probes. The 18S rRNA was used as a loading control. (C) Clear zone formation by the creB disruption mutant harboring mutations in the CreD phosphorylation sites on starch agar medium containing glucose. All strains were grown on starch plus glucose agar medium as described for panel A. (D) α-Amylase activity of the ΔcreB mutant harboring the CreD dephosphorylation mutation. Approximately 2 × 10⁶ conidiospores of each strain were grown in yeast extract-peptone medium plus 5% maltose at 30°C for 48 h. Harvested mycelia were incubated in 100 mM phosphate buffer for 60 min to release the α-amylase bound to the cell wall. The α-amylase activities in the culture broth and phosphate buffer were measured, and the total activity was divided by the mycelial dry weight. Error bars indicate the standard errors of three independent experiments. The P values were calculated by the unpaired Student t test. **, P < 0.01; ns, not significant.

DISCUSSION

In a previous paper, we showed that MalP endocytosis was rapidly induced by CCR-inducing sugars, whereas repression of HulA expression inhibited this endocytosis (5). In this study, we demonstrated that HulA interacts with the arrestin-like protein CreD and that CreD is required for both MalP endocytosis and CCR relief. In addition, a dephosphorylation mutation of CreD in the background of creB disruption further promoted α-amylase production.

S. cerevisiae Rod1/Art4 is phosphorylated by the Snf1 kinase at a serine residue (Ser-447) present in the typical Snf1 recognition motif (22) and is dephosphorylated by PP1 phosphatase (composed of Reg1 and catalytic subunit Glc7) (11). The two putative phosphorylation sites of CreD were identified in a Snf1 recognition motif (Fig. 2C). The A. oryzae genome possesses one Snf1 ortholog (SnfA; FungiDB accession no. AO090701000767), the catalytic domain of which shares high identity with that of yeast Snf1 (see Fig. S4 in the supplemental material). Hence, we generated an A. oryzae snfA deletion mutant (see Fig. S5 in the supplemental material), and 3×FLAG-tagged CreD was expressed in this mutant strain. Although no growth phenotype of an A. nidulans snfA disruption mutant has been reported (26), the A. oryzae ΔsnfA mutant showed drastically reduced conidium formation, suggesting that A. oryzae SnfA is an important functional kinase (see Fig. S5 in the supplemental material). Surprisingly, the phosphorylated form of CreD-3×FLAG was present in the ΔsnfA mutant (see Fig. S5 in the supplemental material). In addition, the genomes of Aspergillus fungi do not harbor highly conserved homologs of yeast Reg1 and Glc7. These findings suggest that the phosphorylation-dephosphorylation mechanism of CreD differs from that of yeast Rod1/Art4.

Glucose-induced Jen1 endocytosis was impaired in the S. cerevisiae reg1Δ mutant, and constitutive Jen1 degradation occurred upon snf1 deletion, suggesting that Rod1/Art4 dephosphorylation promotes Jen1 endocytosis (11). However, the replacement of CreD phosphorylation sites with alanine or glutamic acid had no apparent effect on MalP endocytosis (Fig. 6A and B). Consistent with our result, a recent report showed that mutation of Rod1/Art4 Ser447 to alanine or glutamic acid had no apparent effect on 2-deoxyglucose-induced endocytosis of hexose transporters Hxt1 and Hxt3 (27). Therefore, substitution mutation to glutamic acid may not be sufficient as a phosphorylation mimic for inhibition of endocytosis. Further analysis is required to elucidate the precise involvement of CreD phosphorylation in the endocytosis of MalP.

We demonstrated a physical interaction between CreD and HulA by GST pulldown assay and Co-IP analyses (Fig. 4A and B). Interestingly, one of the CreD phosphorylation sites (Ser-515) is present within the second PxY motif (Fig. 2B). However, replacement of the serine residues with alanine or glutamic acid had no apparent effect on the interaction of CreD with HulA (Fig. 4B). This suggests that the phosphorylation status of CreD is not involved in its interaction with HulA. Similarly, there is no relationship between the phosphorylation status of yeast Art1 and its interaction with Rsp5 (28). Since CreD possesses four PxY and PPxY motifs, identification of the PxY or PPxY motif(s) required for interaction with HulA is important for further understanding of the role of CreD in MalP endocytosis and CCR relief.

Dephosphorylation of CreD was required for CCR relief triggered by creB disruption (Fig. 7C), whereas it had no critical role in the wild-type strain (Fig. 6C). This suggests that dephosphorylation of CreD promotes the ubiquitination of target proteins, which is rapidly antagonized by CreB in the wild-type strain. CreA is a candidate target of CreB and CreD. However, recent studies suggested that CreA was not a direct target of CreB (24, 29). Therefore, analysis of CreA stability in creB and creD disruption mutant strains is necessary to understand the mechanism of CCR regulation in filamentous fungi. In yeast, phosphorylation of Art1 and Rim8/Art9 has been proposed to limit their association with the plasma membrane (28, 30). Analysis of the subcellular localization of CreB and CreD will also augment our understanding of the target of these proteins. In addition, studies on the involvement of CreB in MalP degradation are necessary to understand the function of CreB in A. oryzae.

Considering its advantages for maltose assimilation, resistance to glucose-induced endocytosis of MalP was expected to contribute to the improvement of amylolytic enzyme production. In fact, the yeast maltose transporter Mal21, which is resistant to glucose-induced endocytosis, showed higher maltose uptake activity than that of Mal61 (31). In addition, deletion of four arrestin-like proteins, including Rod1, increased the cellobiose fermentation of yeast expressing cellobiose transporters from the filamentous fungus Neurospora crassa (32). However, creD disruption had a negative effect on α-amylase production, whereas MalP was still retained at the plasma membrane in the presence of glucose (Fig. 2). On the contrary, the dephosphorylation mutation of CreD promoted the α-amylase production of the ΔcreB mutant, and the combination of both mutations improved α-amylase activity by 2.6-fold compared to that of the wild-type strain (Fig. 7D). This finding provides a novel approach to improving the production of secretory glycoside hydrolases in filamentous fungi. Identification of the region or amino acid residue(s) within MalP required for endocytosis is important for improving maltose uptake activity and increasing amylase production.

Most ART proteins recognize several membrane proteins for endocytosis stimulated by the same or different environmental signals. In addition to Jen1, yeast Rod1/Art4 also regulates glucose/2-deoxyglucose-induced endocytosis of hexose transporters (Hxt1, -2, and -6) and agonist-induced endocytosis of α-factor pheromone receptor Ste2 (9, 27, 33, 34). The CreD-independent endocytosis of UapA in A. nidulans (14) suggests that CreD selectively recognizes a membrane protein(s). Although other targets of glucose-induced endocytosis have not been reported in filamentous fungi, transporters or transceptors of sugars, such as xylose and cellobiose, were considered as candidate targets of CreD. However, further studies to identify targets of CreD are required.

In conclusion, this study revealed that CreD is involved in both glucose-induced endocytosis and carbon catabolite derepression. In addition, this study combines the dephosphorylation mutation of CreD with creB disruption as a novel approach to improving the production of secretory glycoside hydrolases in filamentous fungi. Although filamentous fungi secrete a large amount of useful hydrolytic enzymes, such as amylases, cellulases, hemicellulases, and xylanases, CCR limits their production (24, 35). We anticipate that further understanding of CreD function may provide important information regarding the efficiency of hydrolytic enzyme production by filamentous fungi.

MATERIALS AND METHODS

Strains and media.

An A. oryzae ΔligD::ptrA mutant (ΔligD::ptrA niaD⁻ sC⁻) (21) was used as the recipient strain for creD (FungiDB accession no. AO090003000144) deletion and for tagging of creD with 3×FLAG at the genomic locus. The ΔcreB mutant (ΔcreB::pyrG niaD⁻ sC⁻) (24) was used as the recipient for double disruption with creD. The ΔligD::loxP pyrG-deficient strain (ΔligD::loxP niaD⁻ sC⁻ pyrG⁻), which was derived from a ΔligD::loxP mutant (36), was used as the recipient strain for the deletion of snfA. The ΔcreA pyrG⁻ mutant (ΔcreA::sC niaD⁻ pyrG⁻) (24) was used to mutate the phosphorylation sites of CreD. The ΔligD::loxP pyrG-niaD mutant (ΔligD::loxP pyrG::niaD sC⁻) (24) was used as the corresponding wild-type strain for the ΔcreB, ΔcreB ΔcreD, and ΔsnfA mutants. The A. oryzae strains used in this study are shown in Table 1. E. coli DH5α was used for the construction and propagation of plasmids. S. cerevisiae BY4741 was used to construct the plasmid DNA for creD deletion. E. coli BL21(DE3) was used to express the WW domains of HulA (accession no. AO090012000923) as a fusion protein with GST. Czapek-Dox medium [which contained 0.6% NaNO₃; 0.05% KCl; 0.2% KH₂PO₄; 0.05% MgSO₄; trace amounts of FeSO₄, ZnSO₄, CuSO₄, MnSO₄, Na₂B₄O₇, and (NH₄)₆Mo₇O₂₄; and 1% glucose] was used as the standard MM for A. oryzae cultivation. For cultivation of the niaD-deficient strain, NaNO₃ was replaced with 0.5% (NH₄)₂SO₄. Methionine was supplemented at a final concentration of 0.0003% (0.02 mM) for cultivation of the sC-deficient strain.

TABLE 1.

A. oryzae strains used in this study

A. oryzae strain	Origin	Genotype	Reference
ΔligD::ptrA mutant	NS4	niaD− sC− ΔligD::ptrA	21
ΔcreD mutant	ΔligD::ptrA mutant	niaD− ΔligD::ptrA ΔcreD::sC	This study
ΔligD::ptrA GFP-MalP mutant	ΔligD::ptrA mutant	niaD::malP(p)-gfp-malP sC− ΔligD::ptrA	This study
ΔcreD GFP-MalP mutant	ΔcreD mutant	niaD::malP(p)-gfp-malP ΔligD::ptrA ΔcreD::sC	This study
CreD-3×FLAG mutant	ΔligD::ptrA mutant	niaD− CreD-3×FLAG::sC	This study
CreD-3×FLAG-S402A mutant	ΔligD::ptrA mutant	niaD− CreD-3×FLAG-S402A::sC	This study
CreD-3×FLAG-S515A mutant	ΔligD::ptrA mutant	niaD− CreD-3×FLAG-S515A::sC	This study
CreD-3×FLAG-SA mutant	ΔligD::ptrA mutant	niaD− CreD-3×FLAG-S402,515A::sC	This study
CreD-3×FLAG-SA GFP-MalP mutant	CreD-3×FLAG-SA mutant	niaD::malP(p)-gfp-malP CreD-3×FLAG-S402/515A::sC	This study
CreD-3×FLAG-SE mutant	ΔligD::ptrA mutant	niaD− CreD-3×FLAG-S402,515E::sC	This study
CreD-3×FLAG-SE GFP-MalP mutant	CreD-3×FLAG-SE mutant	niaD::malP(p)-gfp-malP CreD-3×FLAG-S402/515E::sC	This study
CreD-3×FLAG 3×HA-HulA mutant	CreD-3×FLAG mutant	niaD::enoA(p)-3×HA-hulA CreD-3×FLAG::sC	This study
CreD-3×FLAG-SA 3×HA-HulA mutant	CreD-3×FLAG-SA mutant	niaD::enoA(p)-3×HA-hulA CreD-3×FLAG-S402/515A::sC	This study
CreD-3×FLAG-SE 3×HA-HulA mutant	CreD-3×FLAG-SE mutant	niaD::enoA(p)-3×HA-hulA CreD-3×FLAG-S402/515E::sC	This study
ΔcreB mutant	ΔligD::loxP pyrG− mutant	niaD− sC− ΔcreB::pyrG	24
ΔcreB ΔcreD mutant	ΔcreB mutant	niaD− ΔcreD::sC ΔcreB::pyrG	This study
ΔligD::loxP pyrG::niaD mutant	ΔligD::loxP pyrG− mutant	sC− niaD::pyrG	24
ΔcreB CreD-3×FLAG-SE mutant	ΔcreB mutant	niaD− ΔcreB::pyrG CreD-3×FLAG-S402/515E::sC	This study
ΔcreB CreD-3×FLAG-SA mutant	ΔcreB mutant	niaD− ΔcreB::pyrG CreD-3×FLAG-S402/515A::sC	This study
CreD-3×FLAGpyrG mutant	ΔligD::loxP pyrG− mutant	niaD− sC− CreD-3×FLAG::pyrG	This study
ΔcreA CreD-3×FLAG mutant	ΔcreA pyrG− mutant	niaD− ΔcreA::sC CreD-3×FLAG::pyrG	This study
ΔcreA CreD-3×FLAG-SA mutant	ΔcreA pyrG− mutant	niaD− ΔcreA::sC CreD-3×FLAG-S402/515A::pyrG	This study
ΔcreA CreD-3×FLAG-SE mutant	ΔcreA pyrG− mutant	niaD− ΔcreA::sC CreD-3×FLAG-S402/515E::pyrG	This study
ΔsnfA mutant	ΔligD::loxP pyrG− mutant	niaD− sC− ΔsnfA::pyrG	This study
ΔsnfA CreD-3×FLAG mutant	ΔsnfA mutant	niaD− ΔsnfA::pyrG CreD-3×FLAG::sC	This study

Plasmid construction.

The plasmid for creD deletion that uses the ATP sulfurylase gene (sC) as a selectable marker was constructed with the yeast homologous recombination system (37, 38) as follows. Approximately 1 kb each of the upstream and downstream regions of creD was amplified by PCR with primer sets creD5OF-creD5OR and creD3OF-creD3OR, respectively. The sC fragment of A. nidulans was obtained by SmaI/PstI digestion of plasmid pUSC (39). These three fragments were introduced into yeast cells with the BamHI/EcoRI-digested pYES2 vector (Life Technologies, Carlsbad, CA, USA), and the assembled plasmid was named pΔcreD::sC.

The plasmid used to introduce the 3×FLAG tag at the C terminus of CreD was constructed as follows. The 3′ region of creD containing a partial 3×FLAG-encoding sequence was amplified by PCR with primers creD-sen and creD3FLAG-anti2 with A. oryzae genomic DNA as a template. The creD terminator region containing a partial 3×FLAG-encoding sequence was also amplified by PCR with primers creD-Ter-sen-3F and creD-Ter-anti on A. oryzae genomic DNA. These two amplified fragments were then cloned into SmaI-digested pUSC with the In-Fusion HD Cloning kit (TaKaRa Bio Inc., Shiga, Japan) in accordance with the manufacturer's instructions. The resultant plasmid containing the partial creD gene fused to the 3×FLAG sequence was designated pUSC-creD-3×FLAG. The introduction of mutations into this plasmid at the codons corresponding to serine residues at positions 402 and 515 was performed by the overlap extension PCR method with the primers shown in Table 2. For introduction into the ΔcreA mutant, the existing selection marker was replaced with a DNA fragment of A. nidulans containing the orotidine-5′-decarboxylase gene (pyrG), which was amplified by PCR with plasmid pUC/pyrG (24) as a template and the AnpyrGsen and AnpyrGantiSphI primers. The resultant PCR product was digested with BglII/SphI and ligated to BamHI/SphI-digested pUSC-creD-3×FLAG, yielding pUpyrG-creD-3×FLAG.

TABLE 2.

Nucleotide sequences of the primers used in this study

^aRestriction sites are underlined.

The plasmid used to express the WW domains of HulA as a fusion protein with GST (GST-WW) was constructed as follows. The fragment containing the three tandem WW domains of HulA was amplified by PCR with primers HulAWWdomainsenEcoRI and HulAWWdomainantiXhoI and A. oryzae genomic DNA. The resultant fragment was digested with EcoRI/XhoI and inserted into pGEX-4T-1 (GE Healthcare UK Ltd., Buckinghamshire, UK).

The plasmid used to express 3×HA-tagged HulA was constructed as follows. The 3×HA-fused hulA fragment was generated by fusion PCR. The 3×HA fragment was amplified by PCR with plasmid pYM22 (40), purchased from Euroscarf (Frankfurt, Germany), as a template with primers N-3HA-sen-op and N-3HA-anti. The hulA fragment was amplified by PCR with primers 3HA-hul-sen and 3HA-hul-anti and A. oryzae genomic DNA. These two PCR fragments were then mixed, and a second round of PCR with primers N-3HA-sen-op and 3HA-hul-anti was performed. The resultant fragment was cloned into NotI-digested pNE (3) harboring the enolase gene (enoA) promoter and niaD as a fungal selectable marker with the In-Fusion HD cloning kit. The resultant plasmid was designated pNE-3×HAHulA.

The DNA fragment for snfA (accession no. AO090701000767) deletion was constructed by fusion PCR as follows. DNA fragments upstream and downstream of the snfA-coding region were amplified by PCR with A. oryzae genomic DNA as a template and primer sets snfAupsen-AnpyrG3snfAupanti and AnpyrG5snfAdownsen-snfAdownanti, respectively. A DNA fragment containing A. nidulans pyrG was amplified by PCR with plasmid pUC/pyrG as a template and the primer set AnpyrGsen-AnpyrGantiPstI. These three PCR fragments were then mixed, and a second round of PCR with primers snfAupsen and snfAdownanti was performed. The nucleotide sequences of all of the primers used in this study are shown in Table 2.

A. oryzae transformation and Southern blot analysis.

SphI/KpnI-digested pΔcreD::sC, StuI-digested pUSC-creD-3×FLAG and its variant plasmids, EcoRV-digested pUpyrG-creD-3×FLAG and its variant plasmids, HpaI-digested pNE-3×HAHulA, and a PCR fragment for snfA deletion were introduced into A. oryzae by the protoplast-polyethylene glycol method described by Gomi et al. (41). Deletions of creD and snfA were confirmed by Southern blotting as described by Tanaka et al. (42).

Fluorescence microscopy.

Fluorescence imaging of GFP-MalP with a confocal fluorescence microscope was performed as described previously (5).

Extraction of GFP-MalP.

GFP-MalP-expressing strains were grown in liquid MM plus 0.1% polypeptone containing 1% glycerol as the carbon source for 24 h and then transferred to fresh MM containing 1% maltose. After shaking for 1.5 h at 30°C to induce GFP-MalP expression, glucose was added to the medium at a final concentration of 0.5%. The mycelia were harvested by filtration through Miracloth (Merck Millipore, Billerica, MA, USA) at equal intervals, washed with distilled water, and frozen in liquid nitrogen. GFP-MalP was extracted by modifying the methods of Becuwe et al. (11) and Gournas et al. (43). In brief, mycelia were ground to a fine powder in liquid nitrogen with a mortar and pestle and the powdered mycelium was suspended in protein extraction buffer (25 mM Tris-HCl [pH 8.0], 100 mM NaCl, 2 mM phenylmethylsulfonyl fluoride [PMSF], 15 μM pepstatin A, complete EDTA-free protease inhibitor [Roche, Indianapolis, IN, USA], 5 mM EDTA). The debris of the crushed mycelium was removed by centrifugation at 1,000 × g for 3 min at 4°C, and the supernatant was further centrifuged for 45 min at 20,400 × g and 4°C. The resultant pellet was suspended in 300 μl of protein extraction buffer containing 0.1% Triton X-100 and incubated on ice for 30 min. For protein precipitation, trichloroacetic acid (TCA) at a final concentration of 10% was added; this was followed by further incubation on ice for 15 min. After centrifugation at 20,400 × g for 15 min at 4°C, the pellet was washed twice with 80% chilled acetone. The pellet was suspended in 50 μl of MURB (50 mM sodium phosphate, 25 mM MES-NaOH, 3% SDS, 3 M urea) and incubated at 37°C for 20 min. A 10-μl sample was then subjected to sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) with a 10% acrylamide gel.

Extraction of CreD-3xFLAG protein.

The mycelium harvested with Miracloth was ground to a fine powder in liquid nitrogen with a mortar and pestle. The powdered mycelium was suspended in protein extraction buffer (25 mM Tris-HCl [pH 8.0], 0.1% Triton X-100, 100 mM NaCl, 2 mM PMSF, 15 μM pepstatin A, complete EDTA-free protease inhibitor, 5 mM EDTA) and incubated on ice for 15 min. After centrifugation at 20,400 × g for 10 min and 4°C, the supernatant was collected, added to an equal volume of 2× Laemmli sample buffer (44), and then boiled for 3 min. The protein concentration in the supernatant was measured with a Coomassie protein assay kit and a microplate photometer Multiskan FC (Thermo Fisher Scientific Inc., Waltham, MA, USA). Approximately 10 μg of protein was subjected to SDS-PAGE with an 8% acrylamide gel.

Western blot analysis.

After SDS-PAGE, the proteins were transferred to an Immobilon P polyvinylidene difluoride membrane (Merck Millipore) with Towbin buffer (40 mM Tris, 38 mM glycine, 20% methanol). For highly efficient transfer of membrane proteins such as GFP-MalP, modified Towbin buffer (40 mM Tris, 38 mM glycine, 10% methanol, 0.05% SDS) was used. Anti-GFP (mFX75), anti-DYKDDDDK, and anti-HA antibodies (Wako Pure Chemical Industries Ltd., Osaka, Japan) were used to detect the tagged proteins in accordance with the manufacturer's instructions. 3-Phosphoglycerate kinase (PgkA; accession no. AO090038000395) was used as a loading control with a custom anti-PgkA peptide antibody (Sigma-Aldrich Japan, Tokyo, Japan).

Alkaline phosphatase treatment.

Approximately 100 μg of intracellular proteins concentrated by TCA precipitation was treated with calf intestine alkaline phosphatase (TaKaRa Bio Inc.) as described by Noguchi et al. (45).

GST pulldown assay.

E. coli BL21(DE3) harboring a plasmid for the expression of GST alone or GST-WW was cultured in Luria-Bertani (LB) medium plus ampicillin (Amp) overnight and then transferred to fresh LB medium plus Amp. Isopropyl-β-D-thiogalactopyranoside was added to the medium at a final concentration of 0.25 mM when the optical density at 600 nm of the culture broth reached 0.4 to 0.6. After 2 h of incubation at 23°C, soluble proteins were extracted with the BugBuster protein extraction reagent (Merck Millipore) in accordance with the manufacturer's instructions. The GST-WW domains were purified with glutathione Sepharose 4 Fast Flow (GE Healthcare UK Ltd.) and mixed with the cell lysate of the CreD-3×FLAG-expressing strain grown in liquid MM plus 0.1% polypeptone containing 1% glucose as the carbon source for 24 h. After 2 h of incubation at 4°C and washing with phosphate-buffered saline (PBS; 140 mM NaCl, 2.7 mM KCl, 10 mM Na₂HPO₄, 1.8 mM KH₂PO₄, pH 7.5), GST, GST-WW, and the bound proteins were eluted by boiling in 2× Laemmli sample buffer and subjected to Western blot analysis.

Co-IP analysis.

The cell lysate prepared by the CreD-3×FLAG extraction method was incubated with anti-HA antibody beads (Wako Pure Chemical Industries Ltd.) for 2 h at 4°C. After the beads were washed with PBS, 3×HA-HulA and the bound proteins were eluted by boiling in 2× Laemmli sample buffer and subjected to Western blot analysis.

Total RNA extraction and Northern blot analysis.

Total RNA extraction and Northern blot analysis were performed as previously described (42).

Measurement of α-amylase activity.

The α-amylase activity in the culture broth was measured by iodine-starch reaction as described previously (24, 46). One unit of α-amylase activity was defined as the amount of enzyme required for starch digestion until the percent transmission of the iodine-starch solution was 66% in 30 min under the reaction conditions used.