The nsdD gene encodes a putative GATA-type transcription factor necessary for sexual development of Aspergillus nidulans

Authors


Abstract

The ability to reproduce both sexually and asexually is one of the characteristics of the homothalic ascomycete Aspergillus nidulans. Unlike the other Aspergillus species, A. nidulans undergoes sexual development that seems to be regulated by internal and external stimuli. To begin to understand the sexual reproduction of A. nidulans we previously isolated and characterized several NSD (never in sexual development) mutants that failed to produce any sexual reproductive organs, and identified four complementation groups, nsdA, nsdB, nsdC, and nsdD. The nsdD gene has been isolated, and it is predicted to encode a GATA-type transcription factor with the type IVb zinc finger DNA-binding domain. The mRNA of the nsdD gene started to accumulate in the early phase of vegetative growth, and the level increased as sexual development proceeded. However, it decreased during asexual sporulation and no nsdD mRNA was detected in conidia. Deletion of nsdD resulted in no cleistothecia (fruiting bodies) formation, even under the conditions that preferentially promoted sexual development, indicating that nsdD is necessary for sexual development. In contrast, when the nsdD gene was over-expressed, sexual-specific organ (Hülle cell) was formed even in submerged culture, which normally completely blocked sexual development, and the number of cleistothecia was also dramatically increased on solid medium. These results lead us to propose that the nsdD gene functions in activating sexual development of A. nidulans. Multiple copies of the nsdD gene could suppress nsdB5 and veA1, indicating that either nsdD acts downstream of these genes or possibly functions in overlapping pathway(s).

Introduction

Aspergillus nidulans is a homothallic ascomycetes that has two major reproductive cycles – sexual and asexual. During reproductive differentiation several morphological changes occur. For asexual development, specialized reproductive structures called conidiophores are elaborated, and millions of asexual spores called conidia are produced. For sexual development, mycelia are aggregated to yield Hülle cells, and ascospore-bearing organs called cleistothecia are generated (for review, see Zonneveld, 1977). Most studies concerning the development of A. nidulans have focused on asexual sporulation. A number of genes have been identified that are associated with asexual development and the regulatory pathway or network of these genes is now well established (for review, see Adams et al., 1998).

However, little is known about the process of sexual sporulation, and only a few genes that seem to be involved in this process have been identified. One example is the veA gene, which was first identified as a gene responsible for a velvet-like phenotype (Käfer, 1965). The veA1 mutation delays and reduces the development of sexual organs, resulting in the preferential development of asexual sporulation (Champe et al., 1981). Compared with the wild type (e.g. FGSC4), asexual development of the veA1 mutant is much less affected by various environmental factors such as nutrients (Han et al., 1994a), light (Mooney and Yager, 1990) and temperature (Champe et al., 1981). Therefore, it was proposed that veA acted as a negative regulator of asexual development and/or an activator of sexual development in response to the various environmental factors (Timberlake, 1990; Han et al., 1994a). Besides the veA gene, several other genes including phoA, stuA, and medA are known to affect sexual development (Clutterbuck, 1969; Miller et al., 1992; Busby et al., 1996;Bussink and Osmani, 1998).

Protection of the culture plate from aeration and light inhibits asexual sporulation of the veA+ strains almost completely (Han et al., 1990; Mooney and Yager, 1990). Under these conditions, sexual differentiation could be easily observed in all colonies. To initiate studies on sexual development in A. nidulans, we developed screening conditions partly using this property (Han et al., 1990), and we were able to isolate various mutants that were defective in sexual development (Han et al., 1990). These were classified into three groups: (1) mutants that were unable to form any sexual structures (NSD: never in sexual development); (2) those that produced immature sexual organs (BSD: block in sexual development); and (3) those that produced fully matured sexual organs but showed differences in amount or timing compared with the wild type (ASD: abnormal in sexual development). Several NSD mutants were analysed for their genetic and morphological characteristics (Han et al., 1994b; Han et al., 1998). At least four complementation groups were identified (nsdA, nsdB, nsdC and nsdD). In addition to the lack of initiation of sexual development, the NSD mutants shared several common phenotypes such as apical growth retardation, early development of asexual spores and production of dark pigments on the underside of colonies.

In this work, we present isolation and characterization of the nsdD gene whose loss-of-function mutations (e.g. nsdD19) result in a complete block of sexual development in A. nidulans. The nsdD gene is predicted to encode a putative GATA-type transcription factor that contains a type IVb zinc finger domain at its C terminus. Deletion of nsdD resulted in absence of sexual development. Forced expression of nsdD induced sexual sporulation, even under conditions that do not allow sexual development. Based on this evidence, we propose that nsdD positively controls the early step(s) of sexual reproduction, and is necessary and sufficient for sexual development of A. nidulans. This is the first study to present evidence that a specific transcription factor is involved in sexual development in A. nidulans.

Results

nsdD encodes a putative GATA-type transcription factor with a type IVb zinc finger domain

We have screened a number of mutants blocked in cleistothecial development, even under the conditions that forced colonies to undergo sexual development (Han et al., 1990). Among these mutants, those that failed to develop cleistothecia (never in sexual development: NSD) were further analysed in detail. The morphological and genetic properties of several mutants in four complementation groups are summarized in Table 1. As shown in Table 1, most NSD mutants showed retarded growth rate and developed conidiophores earlier than the wild type. NSD204, NSD206 and NSD219 strains, carrying nsdA4, nsdC6 and nsdD19 mutation, respectively, almost completely failed to develop sexual sporulation. These mutants were backcrossed with a developmentally wild-type strain to look for segregation of the NSD phenotype. Because these mutations were all recessive (possibly loss-of-function mutations), we were able to examine the meiotic segregation of each mutation, and approximately 50% of the progeny of each cross showed the NSD phenotype. This indicates that the NSD phenotype was caused by a single mutation. Genetic analyses revealed that nsdA and nsdC resided on linkage group II, nsdB on linkage group V and nsdD on linkage group VI (Table 1). Four different alleles were identified in nsdD, nsdD13, nsdD16, nsdD19 and nsdD20, and these allelic mutants showed typical NSD phenotype except the mutant carrying nsdD13, which developed a few cleistothecia in point-inoculated cultures.

Table 1.   Characteristics of NSD mutants.
StrainLocus/alleleLinkage groupGrowth(%)aConidiation(h)bCleistothecia formationc
  • a.

    Relative radial growth rate compared with wild type.

  • b.

    Conidia development time was determined by observing the time at which 50% colonies on a plate had formed conidiophores (mean 

  •  ±  standard error).

  • c. Cleistothecia formation in the culture after whole plate/point inoculation. The number of mature cleistothecia per cm 2 were averages for 10 different areas of a plate: –, < 1; +, 1–10; ++, 10–50; +++, 50–100.

FGSC4  10027.5 ± 0.5+++/++
NSD204 nsdA4 II65 ± 323.5 ± 1–/–
NSD205 nsdB5 V77 ± 224.0 ± 1–/+
NSD206 nsdC6 II65 ± 228.0 ± 1–/–
NSD113 nsdD13 VI68 ± 424.8 ± 1–/+
NSD216 nsdD16 VI67 ± 325.5 ± 2–/–
NSD219 nsdD19 VI64 ± 525.2 ± 1–/–
NSD220 nsdD20 VI68 ± 526.0 ± 3–/–
KHH52 ΔnsdD VI82 ± 225.5 ± 1–/+

The nsdD19-complementing DNA fragment has been isolated and localized within a SmaI–BamHI 4-kb fragment (pNSD19–SB4) that retained full complementing activity (Chae et al., 1995). The restriction map of a 4-kb DNA fragment schematic structure of nsdD gene is shown in Fig. 1A. The nucleotide sequencing revealed that the nsdD gene had an open reading frame (ORF) of 1565 bp, initiated by AUG and interrupted by three introns of 74, 49, and 56 bp in size respectively.

Figure 1.

 Structure of the nsdD gene.

A. The restriction enzyme map of a SmaI–BamHI 4.0-kb fragment containing the nsdD gene is shown. The direction and an approximate start site of nsdD transcription was deduced by analysis of two cDNA clones and is represented by the arrow located above the map. Three introns are indicated by discontinuity of the arrow. The location of the nsdD coding region is represented by an open rectangle, and the closed box in the coding region represents approximate position of a conserved GATA-type zinc finger domain. Restriction enzyme sites are abbreviated: B, BamHI; H, HindIII; S, SacI; E, EcoRI; X, XhoI; Sl, SalI; Sm, SmaI.

B. Mutations in different alleles of nsdD. Numbers at the left and right of each line indicate nucleotide positions. The changed amino acid caused by the mutations is shown under the nucleotide sequence. See text for the mutations in detail. The underlined nucleotides of wild-type lines represent the deletion and the bold nucleotides in mutant lines insertion. The nucleotides in box represent the lariat-forming sequence.

C. A schematic representation of the different mutant alleles of the NsdD protein compared with wild-type NsdD. Blank boxes represent the NsdD part of polypeptides and the boxes with slashed lines the extra polypeptide of non-frame. Corresponding amino acids changes are shown where ATG is indicated as 1. Asterisk indicates an introduced stop codon; fs, frame shift. The complete genomic and cDNA sequences were submitted to the GenBank database under accession number U70043 and U70044 respectively.

The predicted NsdD protein consists of 461 amino acid residues (Mr = 49 250) and is rich in proline (11.3%) and serine (13.4%). Peptide sequence comparison shows that NsdD has the conserved amino acid sequence of type IVb C-X2-C-X18-C-X2-C zinc finger domain typically found in the GATA-type transcription factors (Teakle and Gilmartin, 1998).

To confirm that the isolated gene is actually the nsdD gene rather than a suppressor, we determined the nucleotide sequence of the mutant alleles. The nsdD region from NSD219 was amplified using polymerase chain reaction (PCR) and the nucleotide sequence was determined by directly sequencing the PCR product. We found that the wild-type CG nucleotides (position 2060–2061) were replaced by single T in the nsdD19 allele, and the mutation was predicted to cause a reading frame shift. This reading frame shift results in early termination of translation in nsdD19 and the NsdD19 truncated polypeptide is predicted to have only 81 amino acids (deletion of aa 82–461) (Fig. 1B and C).

With this confirmation that the cloned gene was indeed nsdD, we examined the nucleotide changes in other mutant alleles, i.e. nsdD13, nsdD16 and nsdD20, by amplifying each mutant allele followed by sequencing each PCR product. In the nsdD16 allele, cytosine (2252) of 130th codon was deleted, resulting in a reading frame change. As in the nsdD19 allele, an early termination codon (TAG) is generated after 30 out-of-frame codons, resulting in truncation of the NsdD polypeptide (Δ 130–461aa). The nsdD13 and nsdD20 alleles both have mutations in the consensus lariat-forming sequence of the 2nd intron, and such mutations are predicted to interfere with the splicing of introns (Fig. 1C). With reverse transcription (RT)-PCR, we confirmed that the 2nd intron of these mutant alleles were not spliced (data not shown). These mutants were predicted to produce truncated polypeptides composed of 283 amino acids of NsdD and an extra 52 and 14 amino acids for nsdD13 and nsdD20 respectively. All of these predicted NsdD mutant polypeptides lacked the distal portion of the polypeptide including zinc finger, indicating that the zinc finger domain is necessary for the normal function(s) of NsdD in the induction of sexual development.

Expression of the nsdD gene increases as sexual development proceeds

Because the nsdD gene is expected to control sexual development, it is possible that nsdD expression is coupled to sexual development. Northern blot analysis was carried out for total RNA extracted from conidia, 4-h-germinated conidia, mycelia, an asexually induced culture and a sexually induced culture (Fig. 2). As shown in Fig. 2, the nsdD mRNA was not found in dormant or germinating conidia. The nsdD transcript was detected in vegetative hyphae, and the level was unchanged until early stages of development. As conidiation (asexual sporulation) proceeded, the level of the nsdD transcript decreased. However, when mycelia were forced to develop only sexually, the nsdD mRNA accumulation increased up to the stage of Hülle cell aggregation. These results indicate that although the nsdD gene expression was not specific to sexual developmental stages, its transcript level has a close relationship with the process of sexual development.

Figure 2.

 Expression of nsdD in conidia, vegetative hyphae, asexually and sexually differentiated cells.

A. Conidia were harvested with 0.08% Tween 80 from 3-day-cultured plates. They were crushed with glass beads immediately after harvest or after 4-h germination, and then total RNA was extracted. No nsdD transcript was detected in the dormant conidia (Con; 0 h) nor germinating conidia (4 h). Total RNA from 18-h-cultured mycelia was used for comparison (M) and the nsdD transcript was detected. The actin coding region was used as probe for hybridization controls.

B. The nsdD gene is expressed constantly in vegetative growth culture condition. The total RNA was prepared from mycelia with submerged culture for 16 h or more.

C. Expression of nsdD in conditions that induce asexual sporulation decreased as conidiation proceeded. Mycelia grown for 18 h in liquid MM were transferred onto solid medium and cultured further under normal aeration conditions, which induce asexual sporulation. At appropriate time intervals, samples were collected and RNA was extracted from each differentiated sample.

D. Expression of nsdD in sexually induced conditions was greatly increased at 12 h, a time that coincides with development of abundant Hülle cells and primordia. After that, the expression gradually decreased as development proceeded but recovered at 30 h, when the first mature ascospores were found. Mycelia grown for 18 h in liquid MM were transferred onto solid media plates. The plates were sealed with parafilm and placed in a dark box completely free from light to induce sexual reproduction. These were incubated for 20 h and the seals were removed. Expression of nsdD increased greatly at 12 h after induction of sexual development. Equal loading of total RNA was evaluated by ethidium bromide staining (rRNA).

nsdD is required for sexual development in A. nidulans

In order to characterize the function of nsdD, a disruption construct (pKH41) was generated to replace the nsdD coding region with the intact argB gene as a selective marker. Using site-directed mutagenesis, the nsdD coding region (1383 bp: from 1665 bp to 3288 bp) was deleted and an EcoRV site was introduced, yielding pDHV2. The SmaI–SmaI-digested argB gene fragment was obtained from pJYargB and cloned into pDHV2, resulting in pKH41, and the insert of pKH41 was amplified using PCR (see Experimental procedures). This PCR product was then used to transform VER7 (ΔargB, veA+) strain. Several transformants showing nsdD phenotype were isolated and deletion of the nsdD gene was verified by genomic DNA Southern blot analysis using the 5′-flanking region of the deleted part as a probe (data not shown).

A typical nsdD null mutant, KHH52 (ΔnsdD, veA+), showed somewhat retarded hyphal growth, about 25% slower than a developmentally wild-type control strain KHH60 (nsdD+, veA+), but faster than NSD219 (nsdD19, veA+) (Table 1). This growth pattern was almost the same whether the strain was grown on minimal medium (MM) or complete medium (CM), indicating that growth rate is independent of nutritional status. Conidiophores of all nsdD mutants were formed earlier than the wild type by 2–3 h on plate cultures. No conidiophores were formed in submerged cultures, as was found in wild type.

The nsdD deletion strains were not able to produce cleistothecia or Hülle cells on the standard culture condition, even after a prolonged incubation (more than 10 days (Fig. 3B, Table 2), confirming that nsdD is necessary for sexual development. Furthermore, no sexual development occurred even under conditions such as aeration reduction by plate-sealing, which forces wild-type mycelia to differentiate sexually (data not shown). However, cultures of KHH52 grown from point inoculation on complete medium formed a few cleistothecia bearing mature ascospores at a later time point than wild type (Fig. 3C). However, the nsdD19 mutant did not form any kind of sexual cells under any conditions (Han et al., 1994b; Table 1, Fig. 3D). The fact that the nsdD19 mutation, which contained a frame-shift mutation within the coding region of nsdD, showed more extreme phenotypes in growth and sexual development than the deletion mutation suggested additional negative effects of the truncated NsdD polypeptide formed by the nsdD19 mutation (see Discussion).

Figure 3.

 Phenotypes of the nsdD19 and nsdD deletion mutant strains. Colonies of (A) a developmentally wild-type strain (KHH60) that is isogenic to ΔnsdD (KHH52) (B) and (C) ΔnsdD (KHH52) and (D) nsdD19 (NSD219) are shown. The mutant phenotypes were observed in two different culture conditions, one cultured after plateful inoculation (106 spores/plate; A and B) and the other cultured after point inoculation in the centre of the plate (C and D). When the mutants were point inoculated the ΔnsdD strain formed one or two cleistothecia per plate (the arrow in C), but the nsdD19 strain never produced any cleistothecia (D; see text). The photographs were taken from 5-day cultures (maginification ×30).

Table 2.   Sexual development in various conditions.
StrainRelevant genotypeGlucoseaC SourcebN SourcecOsmolaritydLighte
1%3%GluLacAcetNH3NO3KClSorbitol
  • a.

    Concentration of glucose as a sole carbon source. N source was 0.1% sodium nitrate. The nitrogen source was limited to examine the effect of glucose concentration on development more effectively.

  • b.

    Types of carbon source. 1% glucose (Glu), 1% lactose (Lac) and 2% acetic acid (Acet) were used with 0.6% sdium nitrate as a nitrogen source.

  • c. Types of nitrogen source. 0.2% ammonium tartrate (NH 3) and 0.6% sodium nitrate (NO3) were used. Carbon source was 1% glucose.

  • d.

    Types of osmostabilizer. Potassium chloride(KCl) (0.6 M) and 1.2 M sorbitol (Sorbitol) were added to minimal medium with 1% glucose and 0.6% sodium nitrate as carbon and nitrogen source respectively.

  • e. Illumination with 10–15 W m 2 white light.

  • f. Average number of mature cleistothecia per cm 2 of 10 different areas of a plate: –, < 1; +, 1–10; ++, 10–50; +++, 50–100; ++++, > 100.

  • g.

    An isogenic wild-type strain of KHH62.

  • h.

    Not determined. For strains KHH71, 72, and 73, simply the ability of nsdD overexpression on suppression of nsdA4, nsdB5, and nsdC6 mutations was examined.

FGSC4WT, veA+++f++++++++++++–(+)++
FGSC26WT, biA1, veA1+++++++++++++
NSD219 nsdD19, veA+ NDh
KHH60gWT, veA++++++++++++++++++
KHH52ΔnsdD, veA+
KHH62 niiA(p)::nsdD, veA++++++++++++++++++++++++++++ND+++
KHH41 niiA(p)::nsdD, veA1++++++++++++++++++NDND
KHH71 niiA(p)::nsdD, nsdA4, veA+ NDNDNDNDNDND
KHH72 niiA(p)::nsdD, nsdB5, veA+ +NDNDND+++NDNDND
KHH73 niiA(p)::nsdD, nsdC6, veA+ NDNDNDNDNDND

nsdD is sufficient for activation of sexual development

In order to determine whether nsdD could activate sexual development, we examined the effect of forced expression of the nsdD gene on developmental pattern. This was accomplished by fusing the nsdD gene to the niiA promoter [niiA(p)] followed by tranformation into a wild-type strain. As shown in Fig. 4A, the nsdD transcript was greatly increased in a transformant KHH62, but not in a control strain KHH60 when cultured on sodium nitrate as a sole nitrogen source. However, in the presence of ammonium tartrate, expression was reduced to the control level indicating that the controllable expression was achieved. Over-expression of the nsdD gene resulted in preferential development of sexual organs (Fig. 4B-4). Conidiation was almost completely repressed, whereas profuse elongated aerial hyphae were produced, which later developed into Hülle cells and cleistothecia. We examined the conidia yield of KHH62 by removing an agar block with cork borer (1-cm diameter) to Tween80-containing vials, and counting them with a haemacytometer after vigorous vortexing. The average number of conidia from the culture of KHH62 under induced condition did not exceed 5 × 104, whereas that from under-repressed conditions or that from wild-type control exceeded 1 × 107. Also the KHH62 strain developed cleistothecia at almost the normal level in the presence of 0.6 M KCl or 10 W m−2 white light, which completely inhibited cleistothecia formation in a wild-type strain (Table 2). In addition, over-expression of nsdD resulted in formation of sexual structures even in submerged culture which does not allow any sexual development. As shown in Fig. 4C-4, when KHH62 was cultured in a submerged state in the presence of 0.6% sodium nitrate, whole mycelial pellets developed into Hülle cell aggregates within 3 days. All of these phenotypic changes induced by over-expression of the nsdD gene clearly suggest that nsdD functions in positive regulation of sexual development rather than in negative regulation of asexual sporulation.

Figure 4.

 Forced expression of nsdD induced inappropriate sexual development. Controlled expression of nsdD was accomplished by fusing the nsdD coding region to the niiA promoter [niiA(p):: nsdD] with inducing (nitrate; 2 and 4 for all panels) or repressing (ammonia; 1 and 3 for all panels) nitrogen sources.

A. Northern blot analysis shows that the controlled expression of the nsdD gene was established.

B. Strains with nsdD+ (KHH60: 1, 2) and niiA(p):: nsdD (KHH62: 3, 4) were cultured on MM plates containing 0.2% ammonium tartrate (1, 3) and 0.6% sodium nitrate (2, 4) as the sole nitrogen source (Magnification X30).

C. KHH60 (1, 2) and KHH62 (3, 4) were submerged cultures in MM broth containing 0.2% ammonium tartrate (1, 3) and 0.6% sodium nitrate (2, 4) as the sole nitrogen source. Lots of Hülle cells were observed when nsdD was over-expressed (4) (magnification ×200).

There are several conditions that promote or inhibit cleistothecia formation. For instance, when lactose was supplied as a sole carbon source cleistothecia formation was greatly increased (Table 2). Thus, lactose was considered as one of the sexual inducers. However, on medium with acetate as a sole carbon source, sexual development was completely repressed and growth was retarded (Table 2). Besides acetate, 0.6 M KCl, which was used as an osmotic stabilizer, repressed sexual development completely. However, 1.2 M sorbitol, which was also used as osmotic stabilizer, did not. Interestingly, the nsdD deletion mutants carrying either veA+ or veA1 alleles showed the NSD phenotype, even when the cells were grown on lactose-containing medium (Table 2). This result suggests that sexual development cannot take place normally without the function of the nsdD gene even under the conditions in which sexual development is preferentially induced. On the other hand, when the over-expression strains KHH62 and KHH42 were grown in the presence of acetate as a sole carbon source or additional 0.6 M KCl, many cleistothecia were produced. This result suggests that forced expression of the nsdD gene could overcome the conditions that are inhibitory to sexual development (Table 2).

Suppression of veA1 and nsdB5 mutation by multicopy or over-expression of nsdD

In previous work, we described the relationship between the copy number of the nsdD gene and the rate of cleistothecial development (Chae et al., 1995). Increased copy number of the nsdD gene resulted in a proportional increment in the rate of development of the sexual organs. When the nsdD gene was transformed into several other sexually defective mutant strains, it resulted in near wild-type sexual developmental phenotype in some mutants such as the veA1 and nsdB5. This result indicates that these mutations were suppressed by multiple copies of the nsdD gene (data not shown). These two mutations (KHH41 and KHH72) were also suppressed by over-expression of the nsdD gene (Table 2). These results imply that the nsdD gene may function downstream of these genes or function in possible overlapping pathways.

FlbA is required for sexual development and the nsdD expression

In previous studies, a group of genes that function as major determinants of fungal growth, sporulation and biosynthesis of the mycotoxin (sterigmatocystin: ST) have been identified and characterized. The fadA and sfaD genes encode the α and β subunits of a heterotrimeric G protein, respectively, and both positively mediate growth signalling (Yu et al., 1996b; Rosen et al., 1999). Continuous activation of FadA (and thus SfaD:Gγ) blocks sporulation and ST production while stimulating growth, which results in fluffy-autolytic phenotype. FlbA is a member of the Regulator of G protein Signalling (RGS) proteins that functions as GTPase activating protein. It negatively controls FadA–mediated growth signalling, to allow development and ST production to occur. fluG is responsible for the production of an unidentified extracellular factor and fluG function is apparently necessary for activation of development-specific genes as well as FlbA. Based on the molecular genetic analyses of these and other developmental genes, it was proposed that there are two antagonistic signalling pathways that control fungal growth and asexual sporulation. Because some of these asexual developmental mutants also showed alterations in completing sexual development, it was of interest to examine the expression of nsdD in these mutants.

Some of asexual developmental mutants were grown in the sexually induced culture conditions for 45 h, and total RNA was isolated and the expression of nsdD was examined by Northern blot analysis using the nsdD coding region as a probe. As it was shown in Fig. 5, all mutants tested except a ΔflbA mutant showed accumulation of the nsdD mRNA indicating that FlbA function is required for the nsdD gene expression.

Figure 5.

 Loss of flbA function results in lack of nsdD transcription. Total RNA was isolated from various mutants grown under the sexually induced conditions (Han et al., 1990) for 45 h and the resulting blot was hybridized with the nsdD-coding region. nsdD transcript was accumulated in all tested mutants except ΔflbA. Lane 1, FGSC26 (Wild type); lane 2, AMBP2 (ΔbrlA); lane 3, RJY918 (ΔfadA); lane 4, TBN39.5 (ΔflbA); lane 5, FLUGVE + (ΔfluG, veA+); lane 6, TTA127.4 (ΔfluG, veA1); lane 7, RSRB1.19 (ΔsfaD, veA+); lane 8, TJYGV2 (ΔsfaD, veA1). Equal loading of total RNA was evaluated by ethidium bromide staining.

Finally, we examined whether ΔflbA mutants were able to form any cleistothecia under the sexually induced conditions. ΔflbA mutants are distinguished from the other developmental mutants tested in this study by the fact that they begin as a fluffy aconidial colony but 3 days after inoculation, the centre of the colony begins to disintegrate and by 5 days after inoculation the entire colony has autolysed so that only a few hyphal strands remain (Lee and Adams, 1994a; Wieser et al., 1994). We found that even under the sexually induced conditions ΔflbA mutants showed fluffy-autolytic phenotype and did not form any cleistothecia or Hülle cell aggregate (data not shown). Along with our Northern analysis, this result suggests that FlbA functions in controlling FadA-mediated growth and signalling is necessary for sexual as well as asexual development (see Discussion).

Discussion

The molecular genetic mechanisms controlling initiation of sexual reproduction of A. nidulans are almost unknown. A few genes such as veA (Champe et al., 1981), tubB (Kirk and Morris, 1991), medA (Busby et al., 1996) and stuA (Wu and Miller 1997), are known to affect the cleistothecial development. Previously, we have isolated a number of mutants that were unable to complete normal sexual development and classified them into three phenotypic groups (see Introduction ; Han et al., 1990). Genetic analyses of these mutants resulted in the identification of four complementation groups, nsdA, nsdB, nsdC and nsdD, which are expected to function in determination of sexual development (Han et al., 1994b). We have cloned a DNA region that complemented the nsdD19 mutation completely (Chae et al., 1995), and confirmed it to be the true nsdD gene (this study). Structural analysis of the nsdD gene suggests that it encodes a putative GATA-type transcription factor, which carries a type IVb zinc finger domain (Harrison, 1991; Teakle and Gilmartin, 1998).

As shown in Fig. 3 andTable 2, nsdD deletion mutants failed to initiate sexual reproduction under the standard culture conditions described in Han et al. (1990) and even under the conditions that force wild-type mycelia to differentiate sexually, e.g. restricted aeration (Han et al., 1990). Moreover, nsdD deletion mutants produced conidia somewhat earlier than the wild type (Table 1). When nsdD was over-expressed, cleistothecia were preferentially developed even under unfavourable conditions for sexual development, such as the presence of a high concentration of KCl (Table 2), whereas conidiation was very limited. These observations indicate that the nsdD gene may function as a major determinant of asexual or sexual reproduction. It may be argued that the NSD phenotypes appear as a result of the premature determination of asexual reproduction i.e., the wild-type NsdD protein may act as a negative regulator of asexual development. However, for several reasons it seems more likely that NsdD acts positively on the sexual determination. First, the increased level of nsdD expression was found at a stage which sexual development has proceeded to a considerable extent. It is improbable that a gene necessary for negative regulation of asexual sporulation would be strongly expressed after sexual sporulation has reached an irreversible stage. Second, when nsdD is artificially over-expressed, submerged cultures produce Hülle cells, the sexual-specific organ (Fig. 4C-4), which are not observed in a wild-type strain. It has been shown that over-expression of the asexual development genes including brlA (Adams et al., 1988), abaA (Mirabito et al., 1989) and flbA (Lee and Adams, 1994a) resulted in inappropriate asexual sporulation in submerged culture. Similarly, we observed that over-expression of nsdD was sufficient to cause the appearance of sexually differentiated cells in submerged culture. These results lead us to propose that the nsdD gene positively controls sexual development.

The nsdD deletion mutant never initiated sexual development under the standard culture condition in which heavy inoculum was spread over the whole plate. However it did produce mature cleistothecia, although delayed and in small number, when cultured after point inoculation in the centre of the plates. It is not easy to explain what the differences between two cultures are and how development is affected differently. One possible explanation may be the difference in time of nutrient depletion. On the other hand, the NSD219 (nsdD19) mutant did not develop any kind of sexual cells under any conditions tested (Han et al., 1994b). Furthermore, its growth was more retarded than that of deletion mutants (Table 1). The fact that the nsdD19 mutation, which is predicted to yield the truncated NsdD polypeptide (Fig. 1C), exhibits more extreme phenotypes both in growth (Table 1) and sexual development (Fig. 3D) than a complete loss-of-function (deletion) mutation suggests that the NsdD polypeptide might react with some other protein(s) in the nucleus. A similar result has been observed in mutations of the creA gene that regulates carbon utilization negatively (Shroff et al., 1997). A point mutation in the DNA-binding domain of the creA gene caused a more extreme phenotype than a deletion (Shroff et al., 1997). The mutant CreA protein cannot bind to the promoter region but can bind to other proteins, which results in titration of other proteins, and eventually leads to dominant negative effects. Similarly, the truncated NsdD protein may carry a domain that can interact with other regulatory proteins, but it cannot bind to the promoter and thus dilute out the proteins which, otherwise, would function in induction of sexual development at a later stage.

The regulatory pathway or network of the genes that are involved in the asexual reproduction process, from signalling of differentiation to completion of conidiation, is now well established (Adams et al., 1998). Developmental activation of asexual reproduction requires activities of FlbA, an RGS domain protein, which functions mainly by inactivating FadA signalling of proliferation (Yu et al., 1996b). Activation and completion of an asexual development-specific pathway requires the products of other genes including fluG, flbB, flbC, flbD, flbE and brlA. Our studies suggest that initiation and completion of sexual development also require the antagonistic activity of FlbA on the FadA-proliferation signalling system. We found that the flbA deletion mutant did not form any cleistothecia or Hülle cell aggregate. Furthermore, the expression no nsdD transcript was accumulated in a ΔflbA mutant (Fig. 5), indicating that the expression of nsdD is dependent on FlbA activities. It was shown that the expression of brlA, a key transcription factor in asexual sporulation, was also dependent on FlbA functions (Lee and Adams, 1994a). These results lead us to propose that FadA-mediated growth signalling has to be modulated by FlbA activities in order for both sexual and asexual developments to proceed.

Because sexual reproduction occurs in competition with asexual reproduction in response to various environmental changes, there should be some other signalling and regulatory pathway specific to sexual development, which is comparable to that established for asexual development. In addition to NsdD, the products of nsdA, nsdB and veA seem to be involved in this pathway. Most NSD mutants initiate asexual sporulation earlier than the wild type by several hours, regardless of environmental conditions (Han et al., 1998). When a wild-type strain was cultured on plates sealed with parafilm for more than 30 h, the mycelia did not produce any differentiated cells, but were irreversibly determined to develop sexually and produced few asexual spores after the seals were removed (Han et al., 1990). The NSD mutants, however, began to conidiate immediately after removal of seals, indicating that they never were determined to develop sexually. A fluG deletion mutant that carries the wild-type veA allele no longer shows fluffy phenotype. Instead, it develops plenty of cleistothecia (J-H. Yu, unpublished data). All of these results suggest that the two reproductive cycles are determined in a mutually exclusive way according to environmental conditions, and the process is genetically programmed. Further study of interactions among genes controlling sexual and asexual development including fluG, flbA, nsdD, veA and other flbs and nsds is required to clarify this process.

Experimental procedures

Fungal strains, growth conditions and genetic manipulations

All A. nidulans strains used in this study are listed in Table 3. RMS011 (Stringer et al., 1991), G34 (Gems et al., 1991) and VER7 were used as recipients in transformation experiments following standard A. nidulans techniques (Yelton et al., 1984). Standard A. nidulans genetic procedures were followed (Pontecorvo et al., 1953). Minimal medium (MM) with appropriate supplements was prepared as described (Pontecorvo et al., 1953; Käfer, 1977; Harsanyi et al., 1977). Complete medium (CM) is standard minimal medium with 2.5 g of yeast extract, 2.5 g of casein hydrolysate and 1 ml of vitamin solution per 1 litre. Mutants were isolated on complete media with 0.01% sodium deoxycholate added to reduce colony size. All strains were incubated at 37°C. Developmental cultures were grown on appropriately supplemented MM and sexual induction was performed as previously described (Han et al., 1994b). Light illumination was carried out in a growth chamber equipped with white fluorescent and metal SP lamps (max. 20,000 Lux) and temperature control system.

Table 3.  Strains used in this study.
Strain.GenotypeSource
  1. FGSC; Fungal Genetics Stock Center.

FGSC4Glasgow wild typeFGSCa
FGCS26 biA1; veA1 FGSC
FGSC851 pabaA1 yA2;ΔargB::trpC;trpC801, veA1FGSC
G34 yA2; argB2 methH2; veA1 Gems et al. (1991)
VER7 pabaA1 yA2;ΔargB::trpC;trpC801This study
VE4 biA1; sB3; chaA1; veA1 Han et al. (1994b)
Wx17 biA1, npgA1;sB3;chaA1;veA1 Han and Han (1993)
Wx24 biA1, npgA1;sB3;chaA1;trpC801, veA1 Han and Han (1993)
NSD113 nsdD13 Han et al. (1994b)
NSD204 biA1; nsdA4; sB3; chaA1 Han et al. (1994b)
NSD205 biA1; nsdB5; sB3; chaA1 Han et al. (1994b)
NSD206 biA1; nsdC6, sB3; chaA1 Han et al. (1994b)
NSD216 biA1; nsdD16 sB3; chaA1 Han et al. (1994b)
NSD219 biA1; nsdD19 sB3; chaA1 Han et al. (1994b)
NSD210 biA1; nsdD10 sB3; chaA1 Han et al. (1994b)
KHH22 yA2; argB2 methH2; nsdD19 sB3; chaA1; veA1 This study
KHH60 pabaA1 yA2;ΔargB::trpC;trpC801This study
KHH52 pabaA1 yA2;ΔargB::trpC;ΔnsdD::argB;trpC801This study
KHH62pabaA1 yA2; ΔargB::trpC; niiA(p)::nsdD; trpC801This study
KHH32 pabaA1 yA2;ΔargB::trpC;niiA(p)::nsdD;trpC801 veA1This study
KHH71 pabaA1 yA2; nsdA4;ΔargB::trpC;niiA(p)::nsdD;trpC801This study
KHH72 pabaA1 yA2;ΔargB::trpC;nsdA4;niiA(p)::nsdD;trpC801This study
KHH73 pabaA1 yA2; nsdC6;ΔargB::trpC;niiA(p)::nsdD;trpC801This study
RM31.8 biA1;ΔbrlA::argB;veA1 Han et al. (1993)
RJY918 pabaA1 yA2;ΔargB::trpC;ΔfadA::argB;trpC801 veA1 Yu et al. (1996b)
TBN39.5 biA1 yA2;ΔargB::trpC;ΔflbA::argB;methG1;veA1 Lee and Adams (1994a)
TTA127.4 pabaA1 yA2;ΔargB::trpC;ΔfluG::argB;trpC801 veA1 Lee and Adams (1994b)
FLUGVE+ pabaA1 yA2;ΔargB::trpC;ΔfluG::argB;trpC801This study
RSRB1.19 pabaA1 yA2;ΔargB::trpC;trpC801ΔsfaD::argB veA1 Rosen et al. (1999)
TJYGV2 pabaA1 yA2;ΔargB::trpC;trpC801ΔsfaD::argB Rosen et al. (1999)

NSD219 is the original nsdD mutant strain and was isolated from VE4 after UV mutagenesis (Han et al., 1994b). NSD219 was crossed to G34 (Gems et al., 1991) to yield KHH22 (argB2, nsdD19). RMS011 was crossed to Wx24 to yield VER7 (ΔargB, veA+). KHH52 was constructed by transformation of VER7 with the nsdD deletion plasmid pKH41. KHH62 was constructed by transformation of VER7 with the niiA(p)::nsdD fusion plasmid pKH42 (see below). The argB+ transformants were analysed using genomic Southern blots to confirm that the plasmid had integrated and deleted the nsdD gene. The induction of the niiA promoter is performed as described previously (Yu et al., 1996a).

Plasmid construction

Plasmids were generated using standard techniques (Sambrook et al., 1987). pKH41 was used to replace the A. nidulans nsdD gene and was generated in two steps. First, pDH2 was constructed by cutting pNSD19-SB4 (Chae et al., 1995) with SmaI and EcoRV, and re-ligating to remove the EcoRV site. pDHV2 was then constructed by introducing a EcoRV site 41 bases 5′ of the ATG codon and 3′ of the termination codon (TAA) using in vitro mutagenesis with the synthetic oligonucleotide 5′-TAA TGC TAG CGT GTT AAG ATA T CA CAA AAA CCA GGT GTT G-3′ as described (Kunkel, 1985). The entire nsdD coding region was deleted, and a 1.8-kb XhoI fragment from pJYargB containing the argB gene was inserted into the EcoRV site, yielding pKH41.

Nucleic acid manipulation

Small-scale genomic DNA isolations from fungal mycelia for Southern hybridization and PCR were carried out as described previously (Lee and Taylor, 1990). Southern blot analysis and other DNA manipulations were carried out using standard techniques (Sambrook et al., 1987). For total RNA isolations 1–2 g of lyophilized ground mycelium was homogenized in TrizolTM (GibcoBRL, Grand Island, NY) and isolated following the manufacture's recommendation. Northern blot analyses were performed as described previously with 32P-labelled probes (Yu et al., 1996a).

Microscopy

Photomicrographs presented in this study were taken using an Olympus BX50 or CH30 microscope.

Acknowledgements

We thank Suhn-Kee Chae and our colleagues in the laboratory for many helpful discussions. This work was supported by the Korea Science and Engineering Foundation (KOSEF) grant KOSEF 95-0401-07-01-3 and Wonkwang University grant 2000.

Ancillary