Identification of a Gene Involved in the Synthesis of a Dipeptidyl Peptidase IV Inhibitor in Aspergillus oryzae

Abstract

WYK-1 is a dipeptidyl peptidase IV inhibitor produced by Aspergillus oryzae strain AO-1. Because WYK-1 is an isoquinoline derivative consisting of three l-amino acids, we hypothesized that a nonribosomal peptide synthetase was involved in its biosynthesis. We identified 28 nonribosomal peptide synthetase genes in the sequenced genome of A. oryzae RIB40. These genes were also identified in AO-1. Among them, AO090001000009 (wykN) was specifically expressed under WYK-1-producing conditions in AO-1. Therefore, we constructed wykN gene disruptants of AO-1 after nonhomologous recombination was suppressed by RNA interference to promote homologous recombination. Our results demonstrated that the disruptants did not produce WYK-1. Furthermore, the expression patterns of 10 genes downstream of wykN were similar to the expression pattern of wykN under several conditions. Additionally, homology searches revealed that some of these genes were predicted to be involved in WYK-1 biosynthesis. Therefore, we propose that wykN and the 10 genes identified in this study constitute the WYK-1 biosynthetic gene cluster.

INTRODUCTION

Diabetes is a serious threat to human health, and to date, a number of antidiabetic compounds have been developed to combat this disease. Dipeptidyl-peptidase IV (DPP-IV) inhibitors are antihyperglycemic agents that inhibit the activity of the DPP-IV enzyme, which digests insulinotropic hormones such as glucagon-like peptide 1 (GLP-1) (5, 20). GLP-1 is secreted from intestinal L cells following nutritional stimulation and immediately inactivated by DPP-IV (40). Inactivation of GLP-1 induces diabetes via decreased insulin secretion (5). Recently, many studies in which DPP-IV was inhibited as a means of treating patients with type II diabetes have been published. Treatment of type II diabetes by DPP-IV inhibitors has attracted attention, because compared to other therapies, this treatment has few side effects, such as body weight gain, gastroenteropathy, and hypoglycemia (15, 26, 39). An animal study showed that DPP-IV inhibitors were more effective for the treatment of early-stage diabetes than late-stage diabetes; however, they were not effective for the advanced illness (21). Thus, compounds with DPP-IV-inhibitory activity could not only be useful for the treatment of patients with diabetes and have fewer side effects but also be potential preventive agents for patients with presymptomatic diabetes.

Previously, we reported that a specific strain of Aspergillus oryzae (AO-1), which was isolated from a fermented food in Japan, produced the DPP-IV inhibitor WYK-1 (7). The yellow koji mold A. oryzae has been traditionally used in Japan in fermented foods such as soy sauce, miso, and sake, and therefore, its safety for human consumption has long been verified. Moreover, the United States Food and Drug Administration (FDA) classified this fungus as Generally Recognized as Safe (GRAS). Under specific conditions, a strain of A. oryzae that was isolated from soil produces the DPP-IV inhibitor (22, 30). However, the fungal strain used in the food industry is more suitable for the production of functional foods containing the DPP-IV inhibitor. We confirmed that AO-1, among the many edible A. oryzae strains that we tested, was the only strain that produced the DPP-IV inhibitor (7). Elucidation of the WYK-1 biosynthesis mechanism will be a vital step in improving all aspects of its production.

WYK-1 is an isoquinoline derivative consisting of three l-amino acids (Fig. 1). Besides ribosome-directed protein synthesis, many fungal peptides are synthesized by nonribosomal peptide synthetases (NRPSs), independent of ribosomes (25, 27, 35). In this study, we identified a gene encoding an NRPS that was involved in WYK-1 biosynthesis. Additionally, we hypothesized that this NRPS gene clusters with other genes related to WYK-1 biosynthesis.

Fig 1.

The chemical structure of WYK-1 and its proposed biosynthesis pathway.

MATERIALS AND METHODS

Strains.

A. oryzae RIB40 was obtained from the National Research Institute of Brewing (Higashi-Hiroshima, Japan). A. oryzae AO-1 was isolated from commercially available koji used for the production of amazake (Masuyamiso, Hiroshima, Japan). Koji extract was prepared by incubating koji (Bio'c Co., Ltd., Aichi, Japan) with two parts water at 55°C for 16 h. A. oryzae strains were grown in koji extract medium (dilution of koji extract under conditions of 2 to 5 degrees Baume [specific gravity units]), Sabouraud medium (4% [wt/vol] glucose and 1% [wt/vol] peptone), or Czapek-Dox medium (0.6% [wt/vol] NaNO₃, 0.052% [wt/vol] KCl, 0.152% [wt/vol] KH₂PO₄, 2 mM MgSO₄ · 7H₂O, 1% [wt/vol] glucose, and 0.001 volume of trace elements solution (TLS; 0.1% [wt/vol] FeSO₄ · 7H₂O, 0.88% [wt/vol] ZnSO₄ · 7H₂O, 0.04% [wt/vol] CuSO₄ · 5H₂O, 0.01% [wt/vol] Na₂B₄O₇ · 10H₂O, and 0.005% [wt/vol] (NH₄)₆Mo₇O₂₄ · 4H₂O)) with or without 1.5% [wt/vol] agar.

High-performance liquid chromatography (HPLC) analysis of WYK-1.

A. oryzae was inoculated in koji extract or Sabouraud medium at a final concentration of 1 × 10⁵ conidia/ml and then incubated on a gyratory shaker (100 rpm) at 30°C for 72 h. The culture was centrifuged at 1,710 × g for 10 min. After the supernatant was filtered through a 0.45-μm membrane filter (DISMIC-13cp; Advantec, Tokyo, Japan), the filtrate was analyzed using HPLC. A reverse-phase C₁₈ column (TSK-gel ODS-80Tm column, 4.6 by 150 mm; TOSOH, Tokyo, Japan) was used at 40°C with a 10%-to-90% (vol/vol) methanol gradient at a flow rate of 1.0 ml/min, and detection was carried out at 210 nm with an SPD-M10Avp photodiode array detector (Shimadzu, Kyoto, Japan).

Reverse-transcription PCR (RT-PCR).

A. oryzae was inoculated in koji extract medium or Sabouraud medium at a final concentration of 1 × 10⁵ conidia/ml. The culture was then incubated at 30°C on a gyratory shaker (100 rpm) for 72 h in a 500-ml baffled Erlenmeyer flask. Subsequently, the mycelia were harvested using a Miracloth (Calbiochem, La Jolla, CA), and the total RNA was extracted using Isogen (Nippon Gene Co., Ltd., Tokyo, Japan) and an RNeasy plant minikit (Qiagen K.K., Tokyo, Japan) according to the manufacturer's instructions. RT-PCR was performed using a PrimeScript RT-PCR kit (TaKaRa Bio Inc., Shiga, Japan). PCR primers used in this work were designed on the basis of the database of genomes analyzed at the National Institute of Technology and Evaluation (NITE) (DOGAN: http://www.bio.nite.go.jp/dogan/project/view/AO) and are described in Table S1 in the supplemental material.

A. oryzae transformation.

Mycelia from 100-ml cultures were harvested using a sterilized glass filter and washed with sterile water. The mycelia were added to 10 ml of 0.3 mg/ml yatalase (TaKaRa Bio Inc., Shiga, Japan), which is a cell wall-digesting enzyme, in 10 mM sodium phosphate buffer (pH 6.5) with 0.8 M NaCl and digested at 30°C for 2 h. After filtration with Miracloth, the protoplasts were washed with 0.8 M NaCl and suspended (2 × 10⁸/ml) in solution I (0.8 M NaCl, 10 mM CaCl₂, 10 mM Tris-HCl [pH 8.0]). The protoplast suspension (0.2 ml) and 20 μl of plasmid DNA (20 μg) were incubated on ice for 30 min. Thereafter, 1 ml of solution II (40% polyethylene glycol [PEG] 4000, 50 mM CaCl₂, and 50 mM Tris-HCl [pH 8.0]) was added, and the suspension was incubated for 15 min at room temperature. After 8.5 ml of solution I was added, the suspension was centrifuged at 3,000 rpm for 5 min. The pelleted protoplasts were suspended in 0.2 ml of solution I and mixed with 5 ml of Czapek-Dox selection soft agar medium maintained at 50°C. The soft agar suspension was poured onto a Czapek-Dox agar plate and incubated at 30°C for 5 to 7 days.

Screening of niaD mutants from A. oryzae AO-1.

A previously described method was used to obtain niaD mutants (8). Fungal conidia (1 × 10⁶) were spread on the perchlorate agar plate medium (0.05% [wt/vol] KCl, 0.1% [wt/vol] KH₂PO₄, 5% [wt/vol] KClO₃, 1% [wt/vol] glucose, 0.001 volume of TLS, 2 mM MgSO₄, 0.145% [wt/vol] l-glutamate, and 3% [wt/vol] agar [pH 4.6]). After the plates were incubated at 30°C for several days, perchlorate-resistant colonies were transferred to Czapek-Dox plates containing one of the three different types of nitrogen sources, i.e., nitrate, nitrite, or hypoxanthine, and cultured at 30°C. Colonies that were able to grow on the nitrite and hypoxanthine media but not on the nitrate medium were used as niaD strains.

Construction of ku70 RNAi strain.

A plasmid vector (pUNA-K70i) for suppression of ku70 by RNAi was constructed as follows. First, the gateway conversion cassette A was amplified with primer set GR1 or GR2 from Invitrogen (Carlsbad, CA). Then it was subcloned into pGEM-T Easy (Promega, Tokyo, Japan) to construct pTE-R1-1 and pTE-R2-1, which included the attR1 or attR2 sequences at both cassette ends. Second, an RNA interference (RNAi) spacer fragment (104 bp) was amplified with primers MtA1 and MtA2 by using the RIB40 genome as the template, and the product was subcloned into pCR-Blunt (Invitrogen) to construct pCR-Mtl2. The NotI-EcoRV and MluI-SpeI fragments of pTE-R2-1 were inserted into the NotI-EcoRV and MluI-SpeI sites of pCR-Mtl2 to construct pMtli-R2 containing the RNAi spacer flanked by attR2 on both sides. Next, the EcoRV fragment, which had attR1 sequences at both ends, of pTE-R1-1 was inserted into the SmaI site of pUNA containing the amyB promoter and niaD to construct the destination vector pUNAgateR11 (38). The ku70 gene was amplified with the primer set K71 and K72 and the A. oryzae RIB40 genome as the template. The product was then subcloned into pENTR/D-TOPO (Invitrogen) to construct the entry vector pENTR-K70i. Finally, the entry vector pENTR-K70i, the RNAi spacer plasmid pMtli-R2, and the destination vector pUNAgateR11 were mixed and used in the LR clonase II reaction for the construction of pUNA-K70i. The ku70 RNAi cassette in pUNA-K70i was expressed under the control of the amyB promoter (see Fig. S1 in the supplemental material).

Disruption of the wykN gene.

An experimental outline of the protocol for disruption of the wykN gene is shown in Fig. S2 in the supplemental material. The promoter and middle open reading frame (ORF) regions of wykN were amplified by PCR with primers 009-1f (CCGTAACACATCAATGGGACTGCGGCCGCCGCCAACCATCGGTTAGG) and 009-1r (CGCAAAATGAGGCGAATAGG) and primers 009-2f (GGGAAGCCGAAAGGTGTAATC) and 009-2r (CCTAACCGATGGTTGGCGGCGGCCGCAGTCCCATTGATGTGTTACGG); the underlining indicates the NotI site. These DNA fragments were fused by PCR using the primers 009-1r and 009-2f. The resultant DNA fragment was ligated into SmaI-digested pPTRI (TaKaRa Bio Inc., Shiga, Japan) containing a pyrithiamine-resistant gene (ptrA) to construct pTRI-ΔwykN, which contained a cassette for disruption of the wykN gene. Thereafter, pTRI-ΔwykN was digested with NotI and used for transformation of A. oryzae AO-1:ku70i. Genomic DNA of pyrithiamine-resistant colonies was extracted according to Lee et al. (16). Genomic DNA samples were digested with EcoRI, separated by electrophoresis on a 0.8% [wt/vol] agarose gel, and transferred onto Hybond-N⁺ membranes (Amersham, Tokyo, Japan) according to Chomczynski (4). A digoxigenin (DIG)-labeled probe was constructed using a DIG PCR labeling kit (Roche Diagnostic GmbH, Mannheim, Germany), according to the manufacturer's instruction. The primer sequences for the probes are described in Table S1 in the supplemental material. Hybridization was performed using CDP-star (Roche Diagnostics GmbH, Mannheim, Germany) according to manufacturer's instructions.

DNA sequence analysis of AO-1.

The genomic DNA was extracted according to Lee et al. (16), and wykN and its flanking regions were amplified by PCR. The telomere region was analyzed with a Genome Walker kit (Clontech Laboratories Inc., Palo Alto, CA). The WYK-1 cluster of AO-1 was analyzed using a combination of the Genome Walker kit and the GPS-1 genome priming system (New England BioLabs Inc., Beverly, MA) according to the manufacturer's instructions. DNA sequencing was performed using ABI PRISM model 3100 genetic analyzer (Life Technologies Japan, Tokyo, Japan).

Nucleotide sequence accession number.

The nucleotide sequences have been submitted to the DDBJ-EMBL-GenBank database under accession no. AB705455.

RESULTS

Screening of the gene involved in the biosynthesis of WYK-1.

WYK-1 is an isoquinoline derivative of a tripeptide consisting of three amino acids—l-tryptophan, l-phenylalanine or l-tyrosine, and l-leucine (Fig. 1). We focused on genes encoding NRPSs as candidates for WYK-1-biosynthetic genes because they synthesize many small bioactive peptides (10, 17, 27, 36). NRPSs are usually large proteins of more than 1,000 kDa. According to the PKS/NRPS analysis website (http://nrps.igs.umaryland.edu/nrps/) (2), a gene encoding an NRPS that can synthesize a peptide composed of three amino acids is generally approximately 9 kb long. The genomic sequence of A. oryzae RIB40 has been analyzed (18) and is available at DOGAN (http://www.bio.nite.go.jp/dogan/project/view/AO). DOGAN predicted that the A. oryzae RIB40 genome contains 28 genes encoding NRPSs. Among them, four genes were approximately 9 to 16 kb long (AO090001000009, AO090020000380, AO090038000390, and AO090103000355). Our PCR results confirmed that the A. oryzae AO-1 genome also contained these four genes (data not shown). Thereafter, we analyzed the expression levels of these AO-1 NRPS genes by RT-PCR under the WYK-1 production or nonproduction conditions. The results showed that only AO090001000009 was specifically expressed under WYK-1 production conditions, whereas the other three genes were nonspecifically expressed, regardless of WYK-1 production (Fig. 2). We designated AO090001000009 as wykN and analyzed the gene in detail.

Fig 2.

(A) Expression of putative NRPS genes analyzed by using RT-PCR under various culture conditions. Lane 1, RIB40 cultured in koji extract broth; lane 2, RIB40 cultured in Sabouraud broth; lane 3, AO-1 cultured in koji extract broth; lane 4, AO-1 cultured in Sabouraud broth. (B) Production of WYK-1 under various culture conditions.

Sequence analysis of wykN.

We cloned AO-1 wykN and its flanking regions by PCR and sequenced them. The ORF of AO-1 wykN consisted of 9,531 nucleotides (including a 72-bp intron), and this ORF encoded a protein composed of 3,152 amino acids. We compared the nucleotide and deduced amino acid sequences of wykN genes from AO-1 and RIB40. As shown in Fig. S3 in the supplemental material, AO-1 WykN has an additional nine amino acids at the C terminus because the stop codon in RIB40 is replaced with TGG (W) in AO-1. In total, AO-1 WykN contained 32 amino acids different from RIB40, including the C-terminal extension. According to the domain analysis by the PKS/NRPS analysis website, WykN was predicted to be a trimodular NRPS (see Fig. S4 in the supplemental material). This domain structure was reasonable for WYK-1 production, because one module can correspond to one amino acid, and WYK-1 contains three amino acids.

Because AO-1 and RIB40 wykN showed 99.5% identity, and RIB40 wykN was located on the left telomeric end of chromosome 2, AO-1 wykN may also be located at the same position on the chromosome. Therefore, we performed genome walking to determine the position of AO-1 wykN on chromosome 2. As shown in Fig. S5 in the supplemental material, AO-1 wykN is located approximately 21 kb from the left telomere of chromosome 2. Comparatively, RIB40 wykN is located approximately 16 kb from the left telomere of chromosome 2, thereby suggesting that AO-1 chromosome 2 has an additional 5.3-kb sequence between wykN and the telomere (see Fig. S5).

Gene disruption of AO-1 wykN.

To confirm that wykN is involved in WYK-1 synthesis, we created A. oryzae AO-1 wykN disruptants and analyzed the production of WYK-1. Initially, we tried to obtain the disruptants by transformation of A. oryzae AO-1 with a disruption cassette that contained both upstream and downstream sequences of wykN but lacked its ORF region, as shown in Fig. S2 in the supplemental material. However, we were unable to obtain any disruptants in several trials (0 disruptants in 14 transformants). This might be attributed to high frequency nonhomologous recombination in A. oryzae (14). Nonhomologous recombination is promoted through nonhomologous end joining (NHEJ) by binding of chromosomal double-strand-break ends (6, 28). It was reported that the disruption of ku70 greatly enhances nonhomologous recombination in A. oryzae (29). Therefore, we sought to improve the frequency of homologous recombination by downregulating ku70 using RNA interference (37). First, we screened niaD-deficient strains (niaD strains) from A. oryzae AO-1 and then transformed the AO-1 niaD strain with the ku70-silencing plasmid (pUNA-K70i) by using niaD as a marker. Finally, the resultant strain (AO-1 ku70i) with the silenced ku70 was transformed with the disruption cassette of wykN by using ptrA as a selective marker (see Fig. S2). Pyrithiamine-resistant transformants were selected and subjected to Southern blot analysis with two types of DNA probes. As shown in Fig. 3, wykN of strain 009-I was disrupted and replaced with ptrA. The frequency of successful disruptants was improved to approximately 33% (2 disruptants in 6 transformants) by using the RNA interference method. Subsequently, we cultured the wild-type strain (AO-1) and the wykN disruptant (009-I) under WYK-1 production conditions. Although WYK-1 was produced normally by the wild-type strain, it was not produced by the disruptant (Fig. 4). Furthermore, we confirmed that WYKN was not expressed in the disruptant even under WYK-1 production conditions (Fig. 3). Taking these results together, we concluded that the wykN gene of AO-1 was required for the production of WYK-1.

Fig 3.

Disruption of the wykN gene. (A) Structures of the disruption cassette and wild-type wykN gene. E, EcoRI sites. (B) Results of Southern blot analysis of the wykN gene. Ku, AO-1 ku70i; 009-I, AO-1 wykN disruptant; M, DIG molecular weight marker VII (Roche Diagnostics K. K., Tokyo, Japan). (C) Expression levels of the wykN gene determined by RT-PCR.

Fig 4.

HPLC chromatogram of culture fluid of the wykN disruptant. (A) AO-1:ku70i; (B) wykN gene disruptant of AO-1 (009-I); (C) WYK-1 standard. The arrow indicates WYK-1, which was eluted at 33 min.

The WYK-1 synthesis gene cluster.

We speculated that wykN clusters with other genes related to WYK-1 biosynthesis. Furthermore, expression of genes in a cluster is coordinately regulated in many cases. Therefore, we performed expression analyses of 21 genes around AO-1 wykN under WYK-1 production and nonproduction conditions. RT-PCR analyses revealed that AO-1 wykN and the 10 downstream genes (from AO090001000010 to AO090001000019) were coordinately expressed under WYK-1 production conditions (Fig. 5). These genes were sequenced and compared with those of A. oryzae RIB40. The similarity of the WYK-1 cluster genes between AO-1 and RIB40 was 97.0% to 100.0% (Table 1). However, it was not clear whether these sequence differences could explain the difference in WYK-1 productivity between the two strains. We also performed a BLAST search of these genes to deduce their functions (Table 1). wykA, wykB, wykC, and wykD were similar to genes encoding hydroxylases that modify hydroxyl groups. wykE and wykF were similar to genes encoding oligopeptide transporters, which excrete or take up oligopeptides from or into cells. wykG was similar to genes encoding oxidase, which can form isoquinoline skeletons. wykH was similar to genes encoding N-methyltransferases, which can add methyl groups to nitrogen in the isoquinoline skeleton. wykI was similar to genes encoding O-methyltransferases, which can add methyl groups to oxygen in the isoquinoline skeleton (Fig. 1). wykR was similar to genes encoding transcriptional factors. According to the Pfam database (http://pfam.sanger.ac.uk/), wykR has a ZnII-Cys6 domain (E-value, 0.28), which was identified as a transcriptional activator specific to fungi (13, 31). Because these genes are suggested to be involved in WYK-1 synthesis, we hypothesized that they form the WYK-1 synthesis gene cluster.

Fig 5.

Structure of the WYK-1 gene cluster and expression levels of genes analyzed by RT-PCR. Numbers on the left of the panels correspond to gene numbers in the gene cluster at the top of the figure. A, predicted ORF on the AO-1 telomere region of the left arm of chromosome 2. AO090001000489 (laeA) is a regulator of secondary metabolism. β-Tubulin was used as a positive control. Lane 1, RIB40 cultured in koji extract broth; lane 2, RIB40 cultured in Sabouraud broth; lane 3, AO-1 cultured in koji extract broth; lane 4, AO-1 cultured in Sabouraud broth.

Table 1.

Results of a BLAST search of the WYK-1 gene cluster

ORF	Gene	Similarity (%)a	Best hit for gene with known function	Putative function	E-value
AO090001000009	wykN	99.3	Aspergillus fumigatus	Nonribosomal peptide synthetase	0.0
AO090001000010	wykA	99.5	Aspergillus fumigatus	Phenol hydroxylase	e−150
AO090001000011	wykE	99.6	Microsporum canis	Oligopeptide transporter	e−112
AO090001000012	wykG	99.2	Coccidioides and relatives	Oxidase	1e−42
AO090001000013	wykH	100.0	Aspergillus flavus	N-Methyltransferase	e−141
AO090001000014	wykI	97.4	Aspergillus flavus	O-Methyltransferase	e−107
AO090001000015	wykB	98.7	Aspergillus fumigatus	FAD-dependent oxidoreductase	0.0
AO090001000016	wykF	99.8	Penicillium marneffei	Oligopeptide transporter	0.0
AO090001000017	wykC	99.4	Talaromyces stipitatus	Iron/ascorbate dependent oxidoreductase	1e−67
AO090001000018	wykR	99.3	Aspergillus fumigatus	Cys6 finger domain protein	1e−27
AO090001000019	wykD	97.0	Aspergillus fumigatus	Haloacid dehalogenase, type II	1e−75

^aSimilarity between AO-1 and RIB40.

DISCUSSION

The DPP-IV inhibitor WYK-1 is produced by A. oryzae AO-1, which is edible and is suitable for production in functional foods for the prevention of diabetes. Identification of the WYK-1 biosynthesis pathway is important for improving WYK-1 production in this fungus. Therefore, we cloned the gene involved in WYK-1 synthesis. Because WYK-1 is a tripeptide derivative, we focused on NRPS genes, which produce many biologically active peptides in filamentous fungi (1, 19). We selected four genes with appropriate sizes for the synthesis of tripeptides. Expression analysis revealed that only AO-1 AO090001000009 (wykN) was induced under WYK-1 production conditions (Fig. 2). Furthermore, the wykN gene disruptant of A. oryzae AO-1 did not produce WYK-1 (Fig. 4). Taking these results together, we conclude that wykN is involved in WYK-1 synthesis in A. oryzae AO-1.

Although WYK-1 has a tripeptide structure, it also contains some modifications, such as methylation and hydroxylation. Indeed, many studies have shown that genes encoding NRPSs, relevant modification enzymes, and their transcription factors form gene clusters (11, 34). DNA sequence analysis of neighboring regions of wykN identified many candidate genes encoding modification enzymes and transcription factors. Among them, expression of wykN and the 10 downstream genes were coordinately induced under the WYK-1 synthesis condition (Fig. 5). Therefore, we called these 11 genes (from AO090001000009 to AO090001000019) the WYK-1 gene cluster and proposed a WYK-1 biosynthesis pathway, as shown in Fig. 1. In this pathway, the wykN product synthesizes the tripeptide, followed by methylation by the wykI product. After cyclization by the wykG product to form an isoquinoline ring, methylation by the wykH product and hydroxylation by the wykA product leads to the formation of WYK-1. It will be important to elucidate the enzymatic functions of these gene products in future experiments.

Among the WYK-1 cluster genes, we were especially interested in the wykR gene. Secondary metabolite gene clusters in fungi often encode transcription activators that stimulate expression of entire cluster genes (11, 34). The wykR gene encodes a putative ZnII-Cys6 type transcription factor that shares homology with other fungus-specific transcription factors (13, 31); therefore, this transcription factor is a good candidate for the transcription regulator of the WYK-1 cluster. Expression of wykR was induced under the WYK-1 synthesis condition (Fig. 5), and we are currently analyzing the expression and function of the wykR gene. If WykR is the transcription factor for the entire WYK-1 cluster genes, we may generate WYK-1-hyperproducing strains by overexpression of wykR. Conversely, the production of many secondary metabolites is regulated by LaeA in Aspergillus (3, 9, 24). The A. oryzae RIB40 genome also contains the laeA gene (AO090003000489), which is located at a different region of chromosome 2 according to DOGAN. Based on this laeA sequence, we designed primers for RT-PCR to confirm expression in A. oryzae AO-1. However, the laeA gene was expressed at the same level in AO-1 regardless of the WYK-1 synthesis conditions (Fig. 5). Further studies are required to confirm the effect of the LaeA regulator on WYK-1 production.

Although WYK-1 is not produced by A. oryzae RIB40, it contains not only the wykN gene but also the entire WYK-1 gene cluster. Indeed, the DNA sequence of the AO-1 wykN gene is very similar to that of RIB40 wykN (99.5% identical). The AO-1 WykN protein contains 23 amino acid difference relative to the RIB40 protein. The AO-1 WykN protein also contains nine additional amino acids at the C terminus which were not found in RIB40. However, it is not clear that these differences are responsible for the productivity difference. In addition, the expression levels of wykN and other cluster genes in RIB40 were significantly lower than those in AO-1 were, even under WYK-1 production conditions (Fig. 5). A similar situation has been observed in RIB40 with respect to production of aflatoxin, which is a secondary metabolite toxin in filamentous fungi that is not expressed in RIB40, even though it contains all the aflatoxin-biosynthetic genes (33). Aflatoxin is not produced in RIB40 because of a mutant transcription cofactor that cannot stimulate the expression of the cluster genes (12). Therefore, we hypothesized that the reduced activity of a transcriptional activator that stimulates the expression of the WYK-1 cluster genes resulted in the reduced expression of the WYK-1 cluster genes, thereby giving rise to minimal production of WYK-1 in RIB40. An alternative hypothesis is that the position of the WYK-1 cluster on the RIB40 chromosome results in the reduced expression of this gene cluster. AO-1 has an additional 5.3-kb sequence between the WYK-1 cluster and the telomere, whereas the RIB40 cluster is closer to the telomere (see Fig. S5 in the supplemental material). Indeed, the expression of genes located proximal to the telomere has been reported to be silenced because of structural changes in chromosomes (23). Tokuoka et al. reported that the terminal length difference between chromosomes affects expression of secondary-metabolite-encoding genes in A. oryzae (32). We are testing these hypotheses to elucidate the WYK-1 nonproduction mechanism in RIB40.