The bZIP Transcription Factor MeaB Mediates Nitrogen Metabolite Repression at Specific Loci ^,

Abstract

In Fusarium fujikuroi, bikaverin (BIK) biosynthesis is subject to repression by nitrogen. Unlike most genes subject to nitrogen metabolite repression, it has been shown that transcription of bik biosynthetic genes is not AreA dependent. Searching for additional transcription factors that may be involved in nitrogen regulation, we cloned and characterized the orthologue of Aspergillus nidulans meaB, which encodes a bZIP transcription factor. Two transcripts are derived from F. fujikuroi meaB: the large transcript (meaB^L) predominates under nitrogen-sufficient conditions and the smaller transcript (meaB^S) under nitrogen limitation, in an AreA-dependent manner. MeaB is specifically translocated to the nucleus under nitrogen-sufficient conditions in both F. fujikuroi and A. nidulans. Deletion of meaB resulted in partial upregulation of several nitrogen-regulated genes, but only in the ΔmeaB ΔareA double mutant were the bikaverin genes significantly upregulated in the presence of glutamine. These data demonstrate that MeaB and AreA coordinately mediate nitrogen metabolite repression and, importantly, that independently of AreA, MeaB can mediate nitrogen metabolite repression at specific loci in F. fujikuroi.

The fungal rice pathogen Fusarium fujikuroi is well known for its production of gibberellins (GAs), a group of diterpenoid plant hormones (29, 40), as well as the red polyketide pigment bikaverin (16, 41). The biosyntheses of both GAs and bikaverin are strongly repressed by high levels of nitrogen in the culture medium (14, 16, 40, 41). This repression is partially released by rapamycin, the inhibitor of the conserved TOR kinase (35). Recently we have shown that this repression is mediated at the level of transcription. However, while the GA biosynthetic genes are regulated by nitrogen in an AreA-dependent manner (21, 32), AreA is not essential for the expression of the bikaverin genes (16, 41). In the case of bik1 (formerly pks4), which encodes the bikaverin-specific polyketide synthase, transcript levels were found to be higher in an ΔareA mutant, indicating that AreA might act as a repressor (16). This led us to postulate that a distinct mechanism mediates nitrogen regulation of this secondary metabolite's biosynthesis (41).

In Aspergillus nidulans, AreA activity is modulated by regulated degradation of the areA transcript upon glutamine addition (23, 24, 25). Furthermore, in A. nidulans and Neurospora crassa, AreA (NIT2) activity is inhibited by binding to the negatively acting regulator NmrA (Nmr1) under nitrogen-sufficient conditions (3, 15). Consequently, disruption of nmrA leads to partial derepression of AreA-regulated genes. Although the F. fujikuroi nmr gene fully complemented the N. crassa nmr1 mutant, deletion of nmr results in only marginal derepression of AreA-regulated genes, such as the GA biosynthetic gene cluster (21, 32). However, overexpression of nmr leads to partial repression of AreA activity, as demonstrated by reduced chlorate sensitivity, and Nmr has been shown to interact with AreA and complement the N. crassa nmr1 mutant. These observations are consistent with Nmr having the same mode of action in F. fujikuroi, A. nidulans, and N. crassa, although its importance may vary.

The limited impact of nmr deletion on the expression of AreA target genes, particularly in F. fujikuroi, suggests that additional factors must inhibit AreA activity under nitrogen-sufficient conditions. In A. nidulans, the bZIP transcription factor MeaB affects expression of nitrogen-regulated genes (26). Mutations within meaB are highly pleiotropic and result in the inappropriate expression of several activities subject to nitrogen metabolite repression (26). Recently, it has been reported that under conditions of nitrogen sufficiency MeaB activates nmrA expression by binding to a conserved sequence in the nmrA promoter, thus indirectly modulating AreA activity (43). MeaB orthologues have not yet been characterized in any other fungus, although recently meaB was identified by transposon tagging as a virulence factor in the plant pathogen Fusarium oxysporum (17), and in F. fujikuroi meaB was identified by microarray analysis as a gene whose transcription is regulated in response to nitrogen availability (32).

In this paper, we describe the cloning and deletion of F. fujikuroi meaB and show that it complements the ΔmeaB mutation in A. nidulans. F. fujikuroi meaB produces two distinct transcripts which are differentially regulated by nitrogen availability. Deletion of the gene resulted in marginally increased transcription of nitrogen-regulated genes under nitrogen-limiting conditions (e.g., the GA and bikaverin biosynthesis genes) but did not overcome repression by glutamine. However, derepression of the bikaverin biosynthesis genes was observed in the ΔmeaB ΔareA double mutant. To better understand and compare the functions of MeaB in A. nidulans and F. fujikuroi, we performed green fluorescent protein (GFP) fusion experiments and showed that MeaB is reversibly translocated to the nucleus under nitrogen-sufficient conditions in both species. In contrast to previously published data for A. nidulans (43), in F. fujikuroi MeaB does not significantly regulate nmr expression, with both meaB and nmr being regulated by AreA. We investigated this apparent disparity and found that nmrA expression in A. nidulans is also not meaB dependent.

MATERIALS AND METHODS

F. fujikuroi strains and culture conditions.

We used the wild-type strain IMI58289 (Commonwealth Mycological Institute, Kew, United Kingdom), the areA deletion strain ΔareA-T19 (39) and nmr deletion (Δnmr-T20) and overexpressing (glnA_prom::nmr) mutants (21). For all cultivations, the F. fujikuroi strains were first precultivated for 72 h in 300-ml Erlenmeyer flasks with 100 ml of Darken medium (9) with 2.0 g liter⁻¹ glutamine in place of (NH₄)₂SO₄ on a rotary shaker. A 500-μl aliquot of this culture was used as inoculum for cultivations in ICI and complete medium (CM). For DNA isolation and work with protoplasts, F. fujikuroi strains were incubated in 100 ml CM (27) at 28°C on a rotary shaker at 200 rpm for 3 days or 18 h, respectively. For RNA isolation in experiments with the ΔareA mutants, the fungal strains were grown first in synthetic ICI medium (Imperial Chemical Industries Ltd., United Kingdom) (13) with 10 mM glutamine for 3 days, and the washed mycelia were then transferred to ICI medium without nitrogen for 2 h to induce starvation conditions. After this time, 100 mM glutamine or 100 mM (NH₄)₂SO₄ was added to one-half of the culture flasks, and the other half was maintained under starvation conditions for two more hours. For expression studies with the wild-type strain, the mycelia were grown in 10% ICI medium with NH₄NO₃ for 5 days, washed, and transferred into ICI medium with or without nitrogen sources as indicated below. Mycelia were harvested after 2 h or at the times indicated.

Bacterial strains and plasmid constructs.

Escherichia coli strain Top10 (Invitrogen, Groningen, Netherlands) was used for plasmid propagation. The coding region of meaB was amplified using primers meaB-F1 and meaB-R1 and cloned into the PCR2.1 TOPO vector (Invitrogen) to produce plasmid pmeaB. pUC19 was used to clone genomic DNA fragments from phage carrying parts or all of F. fujikuroi meaB. For construction of the pΔmeaB replacement vector, the 0.8-kb SacI/XbaI fragment from the 5′-noncoding region and a 1.0-kb HindIII/SalI fragment from the 3′-noncoding region were cloned into the plasmid pNR1 (19) carrying the nourseothricin resistance cassette. The SacI/SalI fragment of the resulting replacement vector, pΔmeaB, carrying both flanks and the nourseothricin resistance cassette, was used for gene replacement experiments in the wild-type, ΔareA, and Δnmr genetic backgrounds. For overexpression of meaB, the gene was amplified with primers MeaB_OE/Hind/MeaB_OE/SspI, and the fragment was cloned into the HindIII/SspI-restricted areA-overexpressing vector pglnA_prom::areA carrying the promoter of the F. fujikuroi glnA gene (34; S. Huber and B. Tudzynski, unpublished data). This procedure exchanged the areA gene for the meaB gene. For complementation of the ΔmeaB mutant, the full-length cDNA clones of meaB^L and meaB^S were amplified by PCR using primer pairs meaB-GrTr-Xba/meaB-GrTr-Hind and meaB-klTr-GFP-Xba/meaB-klTr-GFP-Bam, respectively. The XbaI/HindIII fragment of meaB^L was cloned into vector pUCH2-8 (1), while the XbaI/BamHI fragment of meaB^S was cloned into vector pWMS (B. Schönig, unpublished data), with both vectors carrying the hygromycin resistance cassette. The pWMS vector was generated by transferring the ccg1-GFP fusion cassette from pMF272 (12) into pOli-HP (30) containing the hygromycin resistance cassette. The resulting vectors were named pmeaB^L-cDNA and pmeaB^S-GFP, respectively, and transformed into the ΔmeaB mutant. For construction of the nmr uidA reporter gene construct pnmr_prom::uidA, the nmr promoter was amplified using primers nmr-Sph-F and nmr-NcoI-R with integrated restriction sites. The promoter fragment was cloned into the SphI/NcoI-restricted vector pAS (kindly provided by Amir Sharon, Tel Aviv University) carrying the β-glucuronidase (uidA) gene from E. coli and a hygromycin resistance cassette. The GFP fusion vector pccg1_prom-sGFP::meaB was constructed by amplifying the full-length meaB cDNA using the primers MeaB-GFP-f and MeaB-GFP-r. The resulting meaB fragment was cloned into the EcoRI/SacI-restricted GFP vector pWMS carrying the synthetic GFP (sGFP) gene under the control of the N. crassa ccg1 promoter. For visualization of MeaB^S, the vector pmeaB^S-GFP was used. PCR amplification of meaB by using a reverse primer, meaB-Hind-HA, and a forward primer, meaB-Sac-1kb-for, revealed a meaB gene copy with the 3′-hemagglutinin (HA) epitope extension. The HA-tagged fragment was cloned into the SacI/HindIII-restricted vector pUCH2-8, which was named pmeaB-HA and transformed into the wild type (WT).

Screening of genomic library.

About 40,000 recombinant phages of the F. fujikuroi IMI58289 genomic library (19) were screened by plaque hybridization (31). Plaques in E. coli strain LE 392 were blotted onto Gene Screen nylon membranes (DuPont) according to the manufacturer's instructions. A [³²P]dCTP-labeled F. fujikuroi PCR fragment containing about 1.0 kb of the meaB coding region was used as a homologous probe. Hybridizations and washing were performed at 65°C. Putative positive phages were selected, plated, and purified by a second round of hybridization. Phage DNA was isolated as described by Sambrook et al. (31) and used for restriction analysis.

Nucleic acid isolation and analysis.

Lyophilized mycelium was ground into a fine powder and dispersed (in the case of DNA for use in PCR) in extraction buffer as described by Cenis (5). DNA for Southern hybridization experiments was prepared following the protocol of Doyle and Doyle (10). Plasmid DNA was extracted using the Genomed (Germany) plasmid extraction kit. Total F. fujikuroi RNA was isolated using the RNAgents total RNA isolation kit (Promega, Mannheim, Germany). For quantification of transcript levels, X-ray films were scanned and band intensities were analyzed by using ImageMaster Totallab version 2.0 (Amersham Biosciences/GE Healthcare). For Southern analysis, genomic, plasmid, or phage DNA was digested to completion with appropriate restriction enzymes, fractionated in 1.0% (wt/vol) agarose gels, and transferred to nylon membranes (N⁺; Amersham). DNA probes were randomly labeled, and hybridizations were carried out overnight at 65°C. For A. nidulans transcript analysis, growth of mycelia, RNA preparation, and quantitative Northern analysis were performed as described previously (23). Gels were analyzed on a Storm PhosphorImager, and data were analyzed using ImageQuant software.

PCR and RT-PCR.

For cDNA synthesis, 1 μg total RNA was reverse transcribed using the oligo(dT)_12-18 primer and SuperScript II (Invitrogen) according to the manufacturer's instructions. PCR mixtures contained 25 ng DNA, 5 pmol of each primer, 200 nM deoxynucleotide triphosphates, and 1 unit BioTherm DNA polymerase (GeneCraft GmbH, Lüdinghausen, Germany). The reactions were started with 4 min at 94°C, followed by 35 cycles of 1 min at 94°C, 1 min at 56° to 65°C, and 1 min at 70°C, and a final 10 min at 70°C. PCR products were cloned into pCR2.1-TOPO (Invitrogen). For RNA ligase-mediated rapid amplification of cDNA ends (RLM-RACE-PCR), the FirstChoice RLM-RACE kit (Ambion) was used according to the manufacturer's instructions. After the ligation of an adapter fragment, reverse transcription-PCRs (RT-PCRs) were performed using the primers RACE-intern and RACE-RT_rev.

To amplify the approximate full-length meaB cDNA fragment, the RT primers meaB-prom1 to -4 were used. Diagnostic PCR for correct integration of the GUS reporter and GFP reporter fragments utilized primer pairs nmr-Seq-F/GUS-rev and ccg1-prom-F/sGFP-R1. Homologous integration of the meaB replacement cassette was proven with primers meaB-LF-D/meaB-RF-D. All primer sequences are listed in Table S1 of the supplemental material.

F. fujikuroi transformations.

Preparation of protoplasts of F. fujikuroi was carried out as described previously (38). A total of 10⁷ protoplasts of strains IMI58289 (wild type), Δnmr-T20 (22), or ΔareA-T19 (39) were transformed with 10 μg of the SacI/SalI fragment of the replacement vector pΔmeaB to generate single meaB or double meaB areA and meaB nmr knockout mutants. Transformed protoplasts were regenerated at 28°C in a complete regeneration agar [0.7 M sucrose, 0.05% yeast extract, 0.1% (NH₄)₂SO₄] containing 100 μg ml⁻¹ nourseothricin (Werner Agents, Jena, Germany) for 6 to 7 days. Single conidial cultures were established from nourseothricin-resistant transformants and used for DNA isolation and Southern blot analysis. For overexpression of meaB, 10 μg of vector pglnA::meaB was cotransformed with 10 μg of vector pGPC1 (18) carrying the hygromycin resistance cassette into wild-type protoplasts. For complementation and visualization, the ΔmeaB mutant was transformed with 10 μg vector pccg1_prom::sGFP::meaB, pmeaB^S-GFP, or pmeaB^L-cDNA, all carrying hygromycin resistance. Ten micrograms of the vector pmeaB-HA was transformed into the wild-type IMI58289. Hygromycin B-resistant transformants were isolated from complete regeneration agar with 120 μg ml⁻¹ hygromycin B (Cayla, France). For GUS reporter gene assays, the wild type and the ΔmeaB mutant were transformed with the circular vector pnmr_prom::uidA.

DNA sequencing and sequence homology searches.

DNA sequencing of recombinant plasmid clones was accomplished with an automatic LI-COR 4000 sequencer (MWG, München, Germany). Sequence homology searches were performed using the NCBI database server. Protein homology was based on BlastX searches (2). For further investigations the programs of DNAStar Inc. (Madison, WI) were used.

Plate tests.

Growth of the wild-type and meaB, nmr, and areA deletion and overexpression mutants was compared on ICI medium with 100 mM glutamine as nitrogen source, with or without 20 mM potassium chlorate. Direct comparison of A. nidulans and F. fujikuroi wild-type and mutant strains was performed on minimal medium, except for tests with p-fluoro-dl-phenylalanine, where complete medium was used. The agar media contained 10 mM ammonium or glutamine, 10 mM ammonium and glutamine (NH₄ and Gln), or 10 mM sodium nitrate. NH₄ and Gln were supplemented with 100 mM KClO₃, 0.2 mg ml⁻¹ of p-fluoro-dl-phenylalanine, or 30 mM NO₂⁻. In the case of methyl ammonium, plates contained 10 mM NaNO₃ as a nitrogen source and were supplemented with 50 mM methyl ammonium. Plates were incubated at 28°C for 3 days.

Western blot analysis.

Total protein extraction was performed as described in reference 37. Fifty micrograms of the protein extract was used per lane and separated by discontinuous SDS-polyacrylamide gel electrophoresis. The 5% loading gel was used at pH 6.8, while the 10% separation gel had a pH of 8.8. The resulting gel was electroblotted (semidry) to a nitrocellulose transfer membrane (Whatman). For detection of the MeaB-HA fusion protein, a horseradish peroxidase-conjugated HA antibody (Miltenyi) was used.

Fluorescence microscopy.

For studying subcellular localization of MeaB, transformants were inoculated on minimal Aspergillus medium in glass-bottom culture dishes and incubated at 25°C for 16 h. Confocal fluorescence imaging of hyphae was undertaken using Zeiss microscopes: an LSM710 with a Zeiss Plan-Apochromat Fluar 40×, 1.3-numerical aperture (NA) objective and an LSM510 with a Zeiss Plan-Apochromat Fluar 63× 1.4-NA objective with 3× confocal scan zoom, utilizing Zeiss Zen and ZeissLSM software, respectively. MeaB-GFP was excited with a 488-nm laser and imaged using a fluorescence emission bandwidth of 492 to 530 nm, and H1-red fluorescent protein (RFP) was imaged with a 561-nm digital signal processing excitatory laser and using a fluorescence emission bandwidth of 571 to 630 nm. Visualization of the nuclei in F. fujikuroi was conducted at 810 nm using bis-benzimide (Hoechst 33258; Sigma-Aldrich, Germany) at a 1:1,000 dilution and McIlvaine buffer, pH 7.3 (Kangatharalingam and Ferguson [15a]). Prior to microscopy, cultures were incubated at room temperature in the presence of the dye for 40 min followed by a minimal medium wash. All image fields were selected randomly, and a z-series compilation of four to six images through the entire depth of the cells was taken in each x-y time series.

A. nidulans strains and genetic techniques.

A. nidulans strains carried markers in standard use (6, 7). Standard genetic techniques were used (6). Growth media were as described by Cove (8). The ΔmeaB strain was constructed by direct gene replacement with a recombinant PCR construct utilizing A. fumigatus pyrG as a selectable marker and a strain with the genotype pyrG89 pabaB22 ΔnkuA::argB⁺ (argB2) riboB2, as described by Szewczyk et al. (33). This ΔmeaB strain was then used as the recipient for F. fujikuroi meaB, which was integrated at the meaB locus. This utilized a PCR construct which fused the coding region of the F. fujikuroi meaB gene flanked by the A. nidulans meaB promoter and downstream sequences. Transformants were selected for loss of A. fumigatus pyrG and consequent 5-fluoroorotic acid (1.5 mg ml⁻¹) resistance. The sGFP fusion to A. nidulans meaB was constructed by introducing the 5′ 1,120 bp of the meaB coding region into a derivative of pMCB17apx in which the alcA promoter had been replaced with the gpdA promoter (R. Fischer, personal communication). This vector contains N. crassa pyr-4 as a selectable marker. The vector was transformed into a strain with the genotype pyrG89 ΔnkuA::argB⁺ (argB2) pyroA4 veA1. A. nidulans strains used for Northern analysis and growth tests were as follows: biA1 (wild type), pabaA1 areA49, an areA allele with an 8-bp deletion beginning in codon 75 (4), ΔnmrA, and ΔnmrA ΔmeaB.

RESULTS

Cloning of the meaB gene.

To identify the F. fujikuroi meaB gene, we performed a comparison of the A. nidulans meaB sequence against the Fusarium verticillioides genome sequence (http://www.broad.mit.edu) using the tBlastN algorithm (2). The best match was identified on chromosome 3 (locus FVEG_05452.3). We amplified a 1.5-kb genomic fragment of the putative F. fujikuroi meaB homologue and used the PCR product to screen a genomic library of F. fujikuroi wild-type strain IMI58289. A 4.0-kb genomic SacI fragment was identified, subcloned, and sequenced (GenBank accession number FM994929). This revealed an open reading frame of 1,513 bp, interrupted by two putative introns. Splicing of the putative 243-nucleotide (nt) and 52-nt introns was confirmed by RT-PCR (data not shown). Assuming translation initiates at the first AUG, meaB is predicted to encode a 406-amino-acid protein showing high levels of sequence identity to other MeaB-like proteins from Fusarium graminearum (82%; accession number EAA77600.1); N. crassa (75%; accession number EAA28458.1), Magnaporthe grisea (75%; accession number EAA60978.1), and to the already characterized A. nidulans MeaB (57% identity; accession number CAA66668.1).

F. fujikuroi meaB complements A. nidulans ΔmeaB.

To determine if F. fujikuroi meaB is functional in A. nidulans, we replaced the coding region of Aspergillus meaB with that of F. fujikuroi meaB⁺ but retained the A. nidulans promoter. In A. nidulans, disruption of meaB results in resistance to the toxic ammonium analogue methylammonium and increased sensitivity to both chlorate and nitrite in the presence of ammonium (26). The transformants showed almost-wild-type sensitivity to methylammonium and resistance to chlorate and nitrite under nitrogen-repressing conditions (Fig. 1). This demonstrates that F. fujikuroi meaB is a functional orthologue of A. nidulans meaB.

Fig. 1.

Complementation of the A. nidulans ΔmeaB mutant with the F. fujikuroi meaB gene. The A. nidulans WT, ΔmeaB mutant (ΔmeaB) and two strains in which the F. fujikuroi meaB gene (F.f. meaB) replaced the native gene were grown on synthetic complete medium supplemented with NH₄⁺ (10 mM) (a), methylammonium (5 mM) plus NO₃⁻ (b), nitrite (30 mM) plus NH₄⁺ (10 mM) (c), or 300 mM potassium chlorate plus NH₄⁺ (10 mM) (d). Plates were incubated for 2 days at 37°C.

Regulation of meaB expression.

It has been reported that in A. nidulans, meaB is constitutively transcribed (43). In order to determine if this is the case in F. fujikuroi, we utilized Northern blot analysis to monitor meaB transcript levels in both wild-type and ΔareA strains (39). After initial cultivation the strains were transferred to either nitrogen-free or glutamine-supplemented medium. In the wild type, two distinct transcripts were observed, and their expression was dependent on nitrogen availability, the smaller transcript (meaB^S) being prevalent under nitrogen starvation and a larger transcript (meaB^L) predominant in the presence of glutamine (Fig. 2A). In the ΔareA strain, only meaB^L is expressed under the conditions tested, consistent with both the expression of meaB^S and repression of meaB^L being AreA dependent.

Fig. 2.

Regulation of meaB gene expression in F. fujikuroi and physical map of the F. fujikuroi meaB gene. (A) The wild-type, ΔareA, and ΔmeaB strains were grown for 3 days on ICI medium, and the washed mycelia were then transferred for 2 h into either nitrogen-free medium or medium with 100 mM glutamine. Total RNA (15 μg) was subjected to Northern analysis using the genomic meaB fragment as probe. 18S rRNA was imaged as a loading control. The two different transcripts of meaB were named meaB^S (small transcript) and meaB^L (large transcript). (B) The meaB gene structure, with the coding region (open box), introns (gray box), and the two transcription start sites, identified by 5′-RACE PCR, indicated. (C) Sequence of the 5′-untranslated region of meaB^S, with the two GATA sites (yellow) and the putative translation start (red) indicated.

The switch between two transcript sizes could result from differential splicing and/or variation in promoter or terminator utilization. To identify the precise location of transcription start sites, we performed RLM-RACE. In the presence of glutamine, the meaB 5′-untranslated region (UTR) was located 473 nt (meaB^L) upstream from the putative translation start site. The meaB^S transcript, which predominates under nitrogen-starved conditions in the wild type and is not observed in the ΔareA mutant, was found to initiate 36 nt downstream from the 5′ consensus sequence of meaB^L intron 1 (Fig. 2B). The first in-frame AUG within the shorter transcript is located 376 nt downstream from the transcription start site. In addition, we identified two potential AreA binding sites 20 to 40 bp downstream from this transcription start site (Fig. 2C). This observation is consistent with AreA being directly involved in activation of meaB^S.

Based on transcript mapping, it is predicted that meaB^L and meaB^S encode two different protein products (Fig. 2A and B), and their relative protein levels will vary with nitrogen availability. In order to verify this we introduced the coding region for the HA epitope tag at the 3′ end of the meaB conding region. MeaB-HA was monitored by Western analysis in a time course experiment. The fungus was grown for 3 days in the synthetic ICI medium with 20 mM glutamine, and samples were taken after 24, 48, and 72 h. At day 3, the mycelium was shifted for two more hours into ICI medium either without any nitrogen source or with glutamine (nitrogen sufficiency). From this analysis (Fig. 3) a strong signal for full-length MeaB-HA (MeaB^L-HA) was observed at 24 h, when nitrogen would still be present in the medium. At 48 and 72 h, when the fungus would be starving for nitrogen, only a very faint MeaB^L-HA band was detected. Conversely, after shifting into synthetic ICI medium with glutamine, a strong MeaB^L-HA band appeared. These data correlate directly with the meaB^L transcript abundance (Fig. 2A). Under no condition was a smaller band, with the expected size of MeaB^S-HA, visible, indicating that meaB^S is either not translated or its product is only present in very small amounts.

Fig. 3.

Western blot analysis of MeaB-HA expression. MeaB-HA expression was monitored in a MeaB-HA transformant grown for 3 days in synthetic ICI medium with 20 mM glutamine, and samples were taken at 24 h and 48 h. After 72 h the washed mycelia were shifted for 2 h into ICI medium without any nitrogen or with 100 mM glutamine. A wild-type control was included as a negative control. The anti-HA antibody was directly fused to horseradish peroxidase (HRP), which was used for chemiluminescent detection of the tagged proteins.

Deletion of meaB and functional analysis of meaB^S and meaB^L.

For the generation of meaB deletion mutants, the wild-type IMI 58289 was transformed with the SacI/SalI fragment from the replacement vector pΔmeaB (see Fig. S1 in the supplemental material). Four transformants were shown to have undergone homologous integration of the replacement cassette by diagnostic PCR and Southern blot analysis (data not shown). Comparison of the ΔmeaB mutant with the wild type, for transcription of nitrogen-regulated GA and bikaverin biosynthesis genes, revealed a slight upregulation under nitrogen-starved conditions (see Fig. S2 in the supplemental material), suggesting that MeaB contributes to nitrogen regulation, probably acting as a repressor. Consistent with this, the ΔmeaB strain showed reduced growth on chlorate (Fig. 4A). However, deletion of meaB did not result in significant deregulation under nitrogen-sufficient conditions, suggesting that there are additional regulators (see Fig. S2).

Fig. 4.

Complementation of the ΔmeaB mutant with the full-length meaB^L cDNA clone. (A) Growth of the wild type, the ΔmeaB mutant, and three randomly selected meaB^L cDNA transformants (T-23, T-4, and T-9) on agar with 20 mM potassium chlorate. (B) Northern blot analysis of the wild type and meaB^L cDNA transformants, incubated for 2 h in ICI medium without nitrogen or with glutamine. The meaB^L transcript was observed and found to be differentially regulated in the meaB^L cDNA transformants, but the meaB^S transcript was not detected,.

To investigate the function of meaB^S and meaB^L, we transformed the ΔmeaB mutant with full-length cDNA clones derived from both transcripts. Only meaB^L cDNA, including the approximately 500-bp nontranslated 5′-UTR, fully restored the wild-type chlorate resistance (Fig. 4A), while the mutants transformed with meaB^S were indistinguishable from ΔmeaB, being sensitive to chlorate (data not shown). Northern analysis of the meaB^L cDNA transformants, in which meaB^L was transcribed from its native promoter, revealed only the large transcript (Fig. 4B). These data are consistent with the meaB^S transcript initiating within intron 1 of meaB. Importantly, the meaB^L transcript retained appropriate regulation, being highly expressed under conditions of nitrogen sufficiency. Based on this analysis we were unable to identify any function for meaB^S.

Characterization of ΔmeaB Δnmr and ΔmeaB ΔareA double knockouts.

In A. nidulans, AreA activity is partially repressed by binding the negative regulator NmrA under nitrogen-sufficient conditions (3), and it was reported that MeaB regulates nmrA transcription (43). To investigate whether this is the case in F. fujikuroi, we created ΔmeaB Δnmr and ΔmeaB ΔareA double deletion mutants to compare their phenotypes with those of the wild type and the respective single mutants. To determine the impact of AreA, Nmr, and MeaB on the transcription of nitrogen-regulated genes, we performed Northern blot analysis with the wild type and the single ΔareA, Δnmr, and ΔmeaB as well as the double ΔmeaB Δnmr and ΔmeaB ΔareA knockout mutants. We confirmed that transcription of meaB^S is dependent on AreA, since deletion of areA abolished the meaB^S transcript (Fig. 5A; for quantification see Fig. S3A in the supplemental material). Consistent with this we observed that the Δnmr strain has increased levels of meaB^S transcription, possibly as a consequence of enhanced AreA activity. The transcription of genes involved in GA (cps/ks and des) and bikaverin (bik1 and bik2) biosynthesis as well as the AreA target genes MTD1 and aap1, encoding a peptide and an amino acid permease (32), are partially upregulated under starvation conditions in both the ΔmeaB and Δnmr mutants. However, as already shown for Δnmr mutants (32), deletion of meaB did not result in derepression of nitrogen-repressed genes under glutamine-sufficient conditions. An interesting expression pattern was observed for glnA, which encodes glutamine synthetase. Consistent with our previous data, expression of glnA is partially dependent on AreA and is significantly downregulated in the ΔareA mutant (reference 34 and this paper). However, the expression of glnA is altered in opposite ways by the disruption of meaB and nmr, being downregulated in the ΔmeaB strain but upregulated in the Δnmr mutant in the presence of glutamine (Fig. 5A; see also Fig. S3A in the supplemental material). This demonstrates that Nmr and MeaB have distinct roles with respect to glnA expression.

Fig. 5.

Transcription of several nitrogen-regulated genes in the single and double knockout mutants of meaB, areA, and nmr. (A) All strains were grown for 3 days in synthetic ICI medium with 10 mM glutamine. After 3 days, the washed mycelia were shifted for 2 h into ICI medium without any nitrogen or with 100 mM glutamine. Abbreviations: MTD1, peptide transporter gene; glnA, glutamine synthetase gene; AAP1, amino acid permease. (MTD1, glnA, and AAP1 are AreA target genes identified by a microarray approach [30].) (B) All strains were grown in synthetic ICI medium with an initial glutamine concentration of 50 mM. Samples were taken after 24, 48, and 72 h without any additional supplementation of glutamine.

The most striking result was the significant upregulation of the bikaverin biosynthesis genes in the ΔmeaB ΔareA double mutant under both starvation and nitrogen-sufficient conditions, in contrast to both the ΔmeaB and ΔareA single mutants (Fig. 5A; for quantification, see Fig. S3A in the supplemental material). Previously, we showed that AreA is not essential for expression of bikaverin genes (41). In contrast, the expression of the GA genes is strictly AreA dependent, being fully repressed in both the ΔmeaB ΔareA double and ΔareA single mutants. These data are consistent with both AreA and MeaB independently repressing the bikaverin biosynthesis genes and demonstrate that, for a specific subset of genes, MeaB mediates nitrogen metabolite repression independently of AreA. To confirm this result we did a time course experiment, starting in liquid medium with repressing amounts of glutamine (50 mM) that do not allow any expression of GA or bikaverin biosynthetic genes in the wild type. The cultures were harvested after 24, 48, and 72 h. The subsequent Northern experiments revealed deregulation of the bikaverin cluster genes (shown for bik1 and bik2) in the ΔmeaB ΔareA double knockout mutant after just 1 day of growth, despite the high nitrogen concentrations, while the expression of the GA biosynthesis gene cps/ks was completely abolished, consistent with the strong dependency on AreA. After 3 days the expression of bik1 declined, while at this time point the wild type began to show weak expression of the bikaverin genes, consistent with nitrogen depletion being the key signal for expression of these genes. Interestingly, the ΔmeaB and Δnmr single mutants displayed weak upregulation of bik2. This effect was not seen for bik1 (Fig. 5B).

Plate assays with different mutants.

To better understand the role of MeaB in the nitrogen regulation network, we performed plate assays comparing the growth of the wild-type strain with that of the following deletion and overexpression (OE) strains: ΔmeaB, Δnmr, ΔareA, ΔmeaB Δnmr, ΔmeaB ΔareA, OE::meaB, OE::nmr, and OE::areA. On both solidified synthetic ICI medium and CM, all except the ΔmeaB ΔareA double mutant on ICI medium revealed a near-wild-type phenotype (Fig. 6A). However, on medium with the toxic nitrate analogue chlorate, the Δnmr and ΔmeaB strains and the ΔmeaB Δnmr double mutant showed significantly reduced growth, indicative of partial derepression of nitrate metabolism (8). Chlorate sensitivity was more marked in the Δnmr strain and the ΔmeaB Δnmr double mutant than in the ΔmeaB mutant. In contrast, deletion of areA resulted in increased chlorate resistance, consistent with AreA being required for the expression of genes involved in nitrate utilization, such as nitrate reductase (niaD). Overexpression of meaB did not result in significantly better growth on chlorate, while the nmr-overexpressing strain (OE::nmr) had a phenotype comparable to that of the ΔareA mutant (Fig. 6A). These data are consistent with Nmr and MeaB having independent roles, especially with respect to the regulation of genes involved in nitrate metabolism.

Fig. 6.

Plate assays comparing the growth of the wild-type and mutant strains. (A) F. fujikuroi wild-type and mutant strains were grown for 3 days on CM with glutamine (a) and for 6 days on the same medium supplemented with 20 mM potassium chlorate (b). The strains were also grown for 4 days on synthetic ICI medium with glutamine (c) and for 7 days on the same medium supplemented with 20 mM potassium chlorate (d). (B) Direct comparison of A. nidulans and F. fujikuroi wild-type and mutant strains on minimal medium or, in the case of p-fluoro-dl-phenylalanine, complete medium, supplemented with 10 mM NH₄⁺ and Gln (NH₄⁺ and Gln) (e), NH₄⁺ and Gln with 100 mM KClO₃ (f), NH₄⁺ and Gln with 0.2 mg ml⁻¹ of p-fluoro-dl-phenylalanine (g), NH₄⁺ and Gln supplemented with 30 mM NO₂⁻ (h), or with NO₃⁻ as a nitrogen source supplemented with 50 mM methyl ammonium (i). Plates were incubated at 28°C for 3 days.

To compare the phenotypes of F. fujikuroi wild-type and mutant strains with those of the corresponding strains in the model fungus A. nidulans, we performed growth assays with both species together on media with different toxic analogues of nitrogen sources (Fig. 6B). In F. fujikuroi the ΔmeaB and Δnmr strains showed similar sensitivity to chlorate and the ΔmeaB Δnmr double mutant was slightly more sensitive. In A. nidulans, the ΔmeaB mutant showed greater sensitivity than the Δnmr mutant, and more clearly than with F. fujikuroi, the double mutant revealed additivity, with greater sensitivity than either single mutant. For p-fluoro-dl-phenylalanine, both A. nidulans and F. fujikuroi ΔmeaB and ΔmeaB Δnmr strains revealed resistance, while the Δnmr mutants show wild-type sensitivity. In both organisms, in the presence of repressing nitrogen sources (ammonium and glutamine), nitrite is toxic to ΔmeaB but not ΔnmrA strains, indicating that Nmr and MeaB play different roles in this case. As described previously (36), methylammonium is not toxic to F. fujikuroi (32), in contrast to A. nidulans, where the ΔmeaB but not ΔnmrA mutation leads to resistance.

In summary, similar growth responses were shown in corresponding mutants of both fungi for effects of toxic nitrogen analogues, with the exception of methylammonium, which had no inhibitory effect on F. fujikuroi. Importantly, in both species the pleiotropic responses differed between nmr and meaB mutants, which is not consistent with MeaB and NmrA being involved in a single regulatory cascade (Fig. 6B).

Is MeaB the regulator of nmr transcription in F. fujikuroi?

To determine if the transcription of nmr is regulated by MeaB, as reported for A. nidulans (43), we performed Northern blot analysis with the wild type and ΔmeaB mutant. Surprisingly, nmr has a similar expression profile in both the wild type and ΔmeaB mutant (Fig. 7A), being optimally expressed in response to nitrogen starvation and repressed by glutamine. To determine if the expression of nmr is regulated in an AreA-dependent manner, we monitored transcript levels in wild-type, ΔareA, and ΔmeaB ΔareA strains over a 72-h time course, after inoculation into medium supplemented with 100 mM glutamine (Fig. 7B; for quantification see Fig. S3B in the supplemental material). In the wild type, optimal expression was seen at 48 h. Deletion of areA resulted in significant downregulation of nmr, particularly at the first time point, 24 h. In the ΔmeaB strain, nmr was expressed throughout the time course, being highest at 72 h. In the ΔmeaB ΔareA double mutant, the nmr transcript levels diminished, with no expression being observed by 72 h. These data are not consistent with nmr transcription being regulated exclusively by MeaB but do suggest that it contributes to regulation along with AreA and other, as-yet-uncharacterized transcription factors.

Fig. 7.

nmr transcription in wild-type and ΔmeaB strains. (A) The wild type and ΔmeaB mutant were grown for 3 days in synthetic ICI medium with 10 mM glutamine. After 3 days, the washed mycelia were shifted for 2 h into ICI medium without any nitrogen or with 100 mM glutamine. The filter was probed with the genomic fragment of nmr. (B) The wild type as well as ΔmeaB, Δnmr, ΔareA, and ΔareA ΔmeaB mutant strains were grown for 3 days in synthetic ICI medium with 10 mM glutamine, and samples were taken at 24 h, 48 h, and 72 h after inoculation. The filter was probed with the genomic fragment of nmr.

To confirm that MeaB is not essential for the expression of nmr, we transformed both the wild type and the ΔmeaB mutant with a reporter gene construct carrying the F. fujikuroi nmr promoter fused to the E. coli glucuronidase (uidA) gene. Randomly selected transformants of both the wild-type and ΔmeaB strains were indistinguishable with respect to GUS expression, with both classes showing deep blue staining while the control strains had no glucuronidase activity (see Fig. S4 in the supplemental material). These data confirm that the nmr promoter is active in both the wild-type and ΔmeaB background, confirming that MeaB is not required for nmr transcription in F. fujikuroi.

meaB and nmrA expression in A. nidulans.

Considering the apparent disparity between the expression profiles and regulation of meaB and nmrA in F. fujikuroi and that reported for A. nidulans (43), we undertook quantitative Northern analysis of A. nidulans by utilizing wild-type, ΔmeaB, ΔnmrA, and areA49 strains (Fig. 8; for quantification see Fig. S5 in the supplemental material). In addition to meaB and nmrA, we monitored meaA, which encodes an ammonium transporter whose expression is exquisitely sensitive to nitrogen metabolite repression (22), and areA transcript levels. Strains were grown overnight on ammonium as the sole nitrogen source and then transferred to fresh medium containing either no nitrogen, ammonium, glutamine, or alanine and incubated for 4 h (Fig. 8). As can be seen in Fig. 8, both meaB and nmrA transcript levels reflect the nitrogen regimen, with the highest levels being observed in the presence of ammonium and glutamine. The downregulation of meaB transcript levels observed in response to nitrogen starvation is dependent on the presence of a functional areA allele, with repression being lost in the areA49 strain. With respect to nmrA, repression by nitrogen starvation was only partially reversed in the areA49 strain, the expression profile being similar to that observed for the wild type but at approximately three times the level. Therefore, in both F. fujikuroi and A. nidulans, transcription levels of meaB and nmrA are differentially regulated by nitrogen, being directly or indirectly affected by AreA. Surprisingly, we found that nmrA transcription was not dependent on meaB, as had previously been reported (43). However, quantification of the Northern assay results did reveal that nmrA transcript levels varied less dramatically in response to nitrogen regime in the ΔmeaB strain (see Fig. S5 in the supplemental material), consistent with MeaB contributing to this regulation.

Fig. 8.

Northern analysis of meaB, nmrA, areA, and meaA transcript levels in A. nidulans. Northern analysis was conducted on four A. nidulans strains: WT, ΔmeaB, ΔnmrA, and areA49, as indicated. After initial growth for 15 h, mycelia were transferred to fresh medium containing either no nitrogen (-N), NH₄⁺, glutamine (Gln), or alanine (Ala) and incubated for a further 4 h. Blots were sequentially probed for meaB, nmrA, areA, and meaA and, as a control for quantification, 18S rRNA. A parallel experiment in which the final incubation was conducted for 2 h gave a similar profile (see Fig. S5 in the supplemental material). Replicate experiments were conducted and quantified using a phosphorimager.

Intracellular localization and posttranslational control of MeaB.

Previously, the intracellular localization of MeaB had not been characterized in any organism. Of particular interest was if, and under which conditions, MeaB translocates to the nucleus. In order to investigate this in F. fujikuroi, we fused the meaB coding region to the 3′ end of sGFP, which was located downstream from the constitutive N. crassa ccg1 promoter. The resulting construct, pccg1_prom::sGFP::meaB was transformed into the ΔmeaB mutant. Two transformants were selected for MeaB localization studies using confocal microscopy. After overnight growth in the presence of ammonium or glutamine, the fluorescent signal was localized to distinct regions of the cell (Fig. 9A and B) which, based on Hoechst staining, are the nuclei (Fig. 9E). On transfer to nitrogen-free medium the signal became evenly distributed throughout the cells within 60 min. Upon addition of Gln or NH₄⁺ to the medium, the fluorescence signal again localized with the nuclei within 15 to 30 min (Fig. 9A and C; see also Movies S1 to S3 in the supplemental material). After extended starvation (15 h), the nuclear localization triggered by the addition of ammonium (Fig. 9B and D) or glutamine (data not shown) was significantly delayed, taking 90 min. This lag suggests that MeaB localization does not directly respond to extracellular nitrogen levels but requires the metabolic state of the cell to be restored. In the presence of nitrate as sole nitrogen source, nuclear localization is apparent, but a significant proportion of the GFP signal remains outside the nucleus (Fig. 9F). Variation in carbon source, carbon starvation, or oxidative stress did not alter MeaB localization (data not shown). These data are consistent with MeaB functioning as a transcriptional regulator primarily under nitrogen-sufficient conditions.

Fig. 9.

Subcellular localization of MeaB. (A to F) In F. fujikuroi, MeaB-GFP accumulates in the nuclei in the presence of 10 mM glutamine (A) or 10 mM ammonium tartrate (B). MeaB-GFP disperses throughout the cell upon nitrogen starvation. The mean fluorescence intensity was monitored over the full time course for randomly selected nuclei after both short (C) and long (D) periods of nitrogen starvation. The speed with which nuclear localization occurs depends on the duration of starvation: within 15 min after 1 h of starvation (A and C) and about 1.5 h after 16 h of starvation (B and D). Nuclear localization was confirmed by Hoechst staining (E [ammonium tartrate] and F [nitrate]). (G and H) In A. nidulans, MeaB-GFP accumulates in the nuclei (G) under nitrogen sufficiency conditions (10 mM ammonium tartrate for 16 h; the green signal corresponds to MeaB-GFP, and the red signal corresponds to H1-RFP). In the presence of 10 mM ammonium tartrate (H), MeaB-GFP accumulates in the nuclei but during nitrogen starvation disperses within 30 min.

To investigate the smaller product, MeaB^S, a second construct was made in which the GFP coding region was introduced immediately upstream from the first AUG of meaB^S. Very low levels of fluorescence were observed (data not shown), consistent with the absence of a clear signal for MeaB^S detectable in Western blot analysis (Fig. 3). Nuclear localization did not occur under any of the nitrogen regimens tested. This is not surprising, as MeaB^S does not include the putative nuclear localization signal or the bZIP DNA binding domain.

Equivalent experiments were conducted in A. nidulans expressing MeaB tagged with GFP at the N terminus and expressed from the constitutive A. nidulans gpd promoter (28). As in F. fujikuroi, in the presence of ammonium, nuclear localization was observed, based on the MeaB-GFP signal colocalizing with RFP-tagged histone H1 (Fig. 9G and H). Growth on nitrogen sources other than ammonium or glutamine did not result in nuclear localization. After nitrogen starvation the GFP signal became evenly distributed throughout the cell.

In both F. fujikuroi and A. nidulans, quantification of the total fluorescent signal revealed no variation (data not shown). As constitutive promoters were used in both cases, this is consistent with MeaB not being regulated at the level of protein degradation.

DISCUSSION

Most of our understanding of nitrogen regulation in filamentous fungi is based on A. nidulans and N. crassa. However, recent data on nitrogen regulation in F. fujikuroi, especially in relation to the biosynthesis of two nitrogen-regulated secondary metabolites, GAs and bikaverin, have revealed some striking differences. First, in contrast to A. nidulans (20), glutamine synthetase plays a distinct role in regulating expression of nitrogen-regulated genes in F. fujikuroi (34). Second, while the deletion of nmrA and nmr1 in A. nidulans and N. crassa, respectively, resulted in significant deregulation of nitrogen-repressed genes (3, 15), the functional homologue in F. fujikuroi does not appear to play a major role in nitrogen regulation (21, 32). However, as can be seen from the data presented here, in A. nidulans the ΔnmrA phenotype is also quite subtle, leading to no observed derepression of areA transcription and only partial derepression of the ammonium permease meaA in Northern blot assays (Fig. 8) and a marginal effect on chlorate sensitivity in plate tests (Fig. 6). Therefore, additional factors must operate in both F. fujikuroi and A. nidulans to repress AreA activity or independently mediate nitrogen metabolite repression, as in the case of bikaverin genes.

In A. nidulans, besides NmrA, the bZIP transcription factor MeaB was shown to affect the expression of some AreA target genes (26). Recently, MeaB was reported to regulate the expression of nmrA by directly binding to a consensus sequence in the nmrA promoter (43). As such, MeaB is implicated as a key factor in the regulatory cascade that modulates AreA activity through an interaction with NmrA. However, the complex phenotype displayed by nmrA mutants suggests that its role extends beyond the modulation of AreA activity (26, 43).

To determine the role MeaB plays in F. fujikuroi, we cloned the meaB orthologue and confirmed conservation of function by complementation of ΔmeaB in A. nidulans. However, in F. fujikuroi transcriptional regulation of meaB is very different from that reported for A. nidulans (43). In contrast to the almost constitutive transcription of A. nidulans meaB, in F. fujikuroi we observed differential expression of two distinct transcripts whose relative abundance is regulated by nitrogen availability in an AreA-dependent manner (Fig. 2A). In the wild type, the large transcript (meaB^L) is optimally expressed when the fungus is grown under nitrogen-sufficient conditions, while an approximately 500-bp-shorter transcript, meaB^S, which initiates within the first intronic region of meaB^L, accumulates under nitrogen starvation conditions. In the ΔareA mutant, only meaB^L is observed, suggesting AreA plays a negative role in its regulation. Three GATA sequence elements, at positions bp −270, −310, and −370 upstream of the translation start of MeaB^L, are putative binding sites for AreA. Furthermore, three adjacent GATT sequence elements are also present in front of the MeaB^L start codon. These sequence elements were shown to be AreA binding sites in the promoter of the ornithine transaminase gene in A. nidulans, which is strictly repressed by AreA (11). Consistent with AreA acting as a repressor of meaB^L, this transcript was not detected in the Δnmr mutant (Fig. 5A), probably due to elevated AreA activity.

Interestingly, our analysis of meaB transcription in A. nidulans also revealed differential expression in response to nitrogen availability and apparent repression by AreA. Superficially, at least, the mechanisms appear different in the two fungi, as there is no apparent variation in transcript size associated with the regulation. However, in A. nidulans the first intron within meaB does contain four GATA sequences which could potentially bind AreA.

Studying the functionality of the small transcript, we showed that meaB^S did not restore wild-type growth on chlorate when transformed into the ΔmeaB mutant (data not shown). This is consistent with no detectable protein in Western blot analysis and the very low levels of MeaB^S detected by fluorescence microscopy. Furthermore, MeaB^S did not demonstrate nuclear localization under either starvation or nitrogen-sufficient conditions. Searching for nuclear localization signal sequences within the putative MeaB^L protein, we identified three interlaced nuclear localization signal sequences (PSORT II): two pat-7 motifs, initiating at amino acids 55 and 58, respectively, and one pat-4 motif extending from amino acid positions 58 to 61. All three occurred upstream from the putative MeaB^S translational start, consistent with the putative small form, MeaB^S, not entering the nucleus. Additionally, the bZIP domain present in MeaB^L is not encoded by meaB^S. Therefore, MeaB^S is unlikely to function as an active transcription factor.

To better understand the role of MeaB in regulating transcription in F. fujikuroi, we performed Northern blot analyses with the wild type and areA and nmr single deletion mutants as well as ΔmeaB ΔareA and ΔmeaB Δnmr double mutants. We showed that a set of MeaB target genes, such as the secondary metabolism genes for GA and bikaverin biosynthesis, the ammonium transporter-encoding genes, and the gene encoding the amino acid permease AAP1, are all partially upregulated in the ΔmeaB strain under nitrogen-limiting conditions, consistent with MeaB acting as a repressor at these loci. As seen in the time course experiment (Fig. 3), detectable amounts of MeaB^L are still present under these nitrogen-limiting conditions. Recently, we found a similar effect of derepression of bikaverin genes by deletion of PacC not only under repressing alkaline conditions, where PacC is most active, but also under inducing acidic conditions (41). Interestingly, MeaB promotes the expression of glnA under high glutamine concentrations, while Nmr1 represses its expression, indicating complementary roles for MeaB and Nmr in the regulation of glutamine synthetase.

We have shown that MeaB contributes to the regulation of nitrogen-controlled genes, but generally it only moderates their expression and does not overcome the requirement for AreA. This is exemplified by the inability of ΔmeaB strains to deregulate nitrogen-repressed genes under nitrogen-sufficient conditions. The one exception to this is very striking: expression of the bikaverin genes in the ΔmeaB ΔareA double mutant in the presence of glutamine. This is consistent with our recent finding that the mechanism of nitrogen metabolite repression for this secondary metabolite gene cluster differs dramatically from the fungal paradigm, with AreA not being required for expression (41). Deletion of either areA or meaB alone does not lead to significant derepression of the bikaverin genes, indicating that both transcription factors can act independently to regulate these genes. Recently, we showed that the expression of bik genes is regulated in a very complex manner, being negatively regulated by PacC-dependent pH regulation under alkaline conditions and a noncanonical (AreA-independent) nitrogen regulation mechanism in addition to induction mediated by the positively acting pathway-specific transcription factor Bik5 (41). Furthermore, we have also shown that the global regulator Velvet negatively affects bik gene expression, while it is essential for the expression of GA genes (42). Therefore, all these components must be included in a general model for the regulatory network for GA and bikaverin gene clusters (Fig. 10).

Fig. 10.

Model for regulation of GA and bikaverin biosynthesis genes in F. fujikuroi. The GA biosynthetic genes are subject to nitrogen metabolite repression mediated by AreA, while the bikaverin (bik) biosynthetic genes are nitrogen repressed in a noncanonical manner independently by AreA and MeaB. Consequently, the double deletion of meaB and areA results in strong derepression of the bik genes under nitrogen-sufficient conditions. AreA has a dual role as an activator of meaB^S and a repressor of meaB^L. The expression of nmr1 is repressed by large amounts of nitrogen and activated upon nitrogen starvation, but MeaB is not essential for expression of nmr1. While no pathway-specific transcription factor is present in the GA gene cluster, the bik genes are positively regulated by the pathway-specific transcription factor Bik5 and repressed by PacC under alkaline pH conditions (41). Both the GA and bik genes are subject to epigenetic regulation, mediated by the Velvet complex: the global regulator Velvet activates the expression of the GA and inhibits the expression of bik genes, while the histone methyltransferase LaeA activates the GA genes and the bik genes.

The independent repression of bik gene expression by MeaB and AreA is novel for two reasons. First, it defines MeaB as a transcription factor which can mediate nitrogen metabolite repression independently of AreA. Second, it reveals a previously undescribed function for AreA. Generally, AreA acts as an activator and more rarely as a repressor, but such activities are predominantly observed under conditions of nitrogen limitation. Recently, we identified a set of genes in F. fujikuroi by microarray studies which are repressed by AreA under starvation conditions, e.g., cpc1 and idi4, encoding the cross-pathway control regulator and a transcription factor involved in autophagy, respectively (32). However, in this instance AreA is either directly or indirectly repressing the bikaverin gene cluster under conditions of glutamine sufficiency.

Another interesting result is that, in contrast to published data for A. nidulans (43), MeaB is not essential for the transcription of nmr in F. fujikuroi, with near-wild-type levels of nmrA mRNA being expressed in the ΔmeaB strain. Furthermore, in A. nidulans nmrA transcription is only slightly reduced in the ΔmeaB strain. Surprisingly, in F. fujikuroi nmr is upregulated under nitrogen starvation conditions and repressed under nitrogen sufficiency, with meaB^L having the opposite profile. Furthermore, the consensus MeaB binding element (TTGCACCAT), found in the promoter sequence of nmrA genes in a number of Aspergillus species (43), is not present in the 5′-UTR of the F. fujikuroi, F. verticillioides, F. oxysporum, or F. graminearum nmr genes (Broad Institute [http://meme.sdsc.edu/]). Therefore, from our data, there is no evidence that in F. fujikuroi MeaB and Nmr act in the same regulatory cascade. This is supported by the distinct pleiotropic nature of ΔmeaB and ΔnmrA mutants. The most striking example of this is the opposite effect of ΔmeaB and Δnmr mutations on the expression of the glutamine synthetase structural gene, glnA (Fig. 5A; see also Fig. S3A in the supplemental material).

Based on the mutant phenotypes, MeaB appears in most cases to be a negatively acting transcription factor involved in coordinating the cellular response to nitrogen availability. In this respect it appears to act in opposition to AreA, although various features suggest that only a subset of genes are coregulated by both proteins. As with AreA, nitrogen quality and availability appear to be the key signals for MeaB. Not only is the functional meaB^L optimally expressed in response to nitrogen sufficiency, but also the MeaB^L protein is found to localize to the nuclei under these conditions, consistent with it acting as a transcriptional regulator primarily under these conditions. The dynamic intracellular response of the MeaB^L protein to changes in the nitrogen regimen, with the rapid localization to the nucleus in the presence of either ammonium or glutamine, which is reversed on transfer to nitrogen-limiting conditions (Fig. 9; see also the movies in the supplemental material), reveals posttranslational modulation that is fully coordinated with its transcriptional regulation. Previously in A. nidulans, AreA was shown to localize to the nucleus primarily under conditions of nitrogen starvation, where transcript levels are highest, and this is rapidly reversed upon addition of ammonium (37). This again underlines the apparently reciprocal roles of these two transcription factors.

Based on this work, it appears that for many genes regulated by MeaB, their expression is dependent on AreA, with the striking exception of the bikaverin biosynthesis pathway, in which both transcription factors contribute to regulation independently, repressing transcription under conditions of nitrogen sufficiency. This reveals that although their functions are often linked this is not necessary and that MeaB can independently mediate nitrogen metabolite repression.