Suppressor Mutations Bypass the Requirement of fluG for Asexual Sporulation and Sterigmatocystin Production in Aspergillus nidulans

Corresponding author: Jae-Hyuk Yu, University of Wisconsin, 1925 Willow Dr., Madison, WI 53706-1187., jyu1{at}wisc.edu (E-mail)

Communicating editor: J. LOROS

	ABSTRACT

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Asexual sporulation (conidiation) in the filamentous fungus Aspergillus nidulans requires the early developmental activator fluG. Loss of fluG results in the blockage of both conidiation and production of the mycotoxin sterigmatocystin (ST). To investigate molecular mechanisms of fluG-dependent developmental activation, 40 suppressors of fluG (SFGs) that conidiate without fluG have been isolated and characterized. Genetic analyses showed that an individual suppression is caused by a single second-site mutation, and that all sfg mutations but one are recessive. Pairwise meiotic crosses grouped mutations to four loci, 31 of them to sfgA, 6 of them to sfgB, and 1 each to sfgC and sfgD, respectively. The only dominant mutation, sfgA38, also mapped to the sfgA locus, suggesting a dominant negative mutation. Thirteen sfgA and 1 sfgC mutants elaborated conidiophores in liquid submerged culture, indicating that loss of either of these gene functions not only bypasses fluG function but also results in hyperactive conidiation. While sfg mutants show varying levels of restored conidiation, all recovered the ability to produce ST at near wild-type levels. The fact that at least four loci are defined by recessive sfg mutations indicates that multiple genes negatively regulate conidiation downstream of fluG and that the activity of fluG is required to remove such repressive effects.

ASEXUAL sporulation in Aspergillus nidulans is a continual progression from growth to development and is a precisely timed and genetically programmed event in the life cycle in response to internal and external cues. It involves formation of multicellular reproductive organs termed conidiophores, each of which produces thousands of mitotically derived spores (for review see ADAMS 1994 ; ADAMS et al. 1998 ). In previous studies, we showed that two antagonistic signaling pathways control growth and asexual development in A. nidulans. Growth signaling is mediated by FadA and SfaD, the - and ß-subunits for a heterotrimeric G protein, respectively. When FadA (G) is active, GTP-bound FadA and the heterodimer SfaD(Gß):G are dissociated and both signal to enhance proliferative growth, which in turn represses both asexual sporulation and sterigmatocystin (ST) production (Fig 1; YU et al. 1996 ; HICKS et al. 1997 ; ROSEN et al. 1999 ). Constitutive activation of FadA growth signaling results in the fluffy autolytic phenotype. Initiation of asexual development requires the activity of two major genes, flbA and fluG. FluG is responsible for the production of an extracellular factor and it stimulates both development-specific events and activation of FlbA, which in turn inactivates FadA signaling (LEE and ADAMS 1994B ; YU et al. 1996 ). FlbA is a regulator of G protein signaling (RGS) domain protein and, like other RGS proteins, is predicted to negatively regulate G protein signaling by facilitating the intrinsic GTPase activity of the G (FadA)-subunit (LEE and ADAMS 1994A ; BERMAN et al. 1996 ; KOELLE and HORVITZ 1996 ; YU et al. 1996 ; HEPLER et al. 1997 ). Asexual development requires both inhibition of growth and activation of sporulation.

View larger version (22K): In this window In a new window Download PPT slide   Figure 1. Genetic model for growth and development in A. nidulans. In this model, the activities of two antagonistic signaling pathways control growth and asexual sporulation (and ST production). Growth signaling is mediated by a heterotrimeric G protein composed of FadA (G) and SfaD (Gß):G. Activated FadA-GTP and SfaD:G transduce signals to downstream effectors that include PkaA (a catalytic subunit of protein kinase A; SHIMIZU and KELLER 2001 ). Activation of proliferative growth represses both asexual sporulation and ST biosynthesis. FlbA is an RGS protein that rapidly turns off FadA-mediated growth signaling likely by acting as a GTPase-activating protein (GAP) for FadA. At least partial inhibition of FadA-mediated growth signaling is required for conidiation and ST production. Activation of asexual sporulation requires the activities of FluG and other downstream developmental genes, flbB, flbC, flbD, flbE, and brlA (for review see ADAMS et al. 1998 ).

A key step in the formation of conidiophores is activation of the brlA gene, which encodes a C₂H₂ zinc finger transcriptional activator required for expression of sporulation-specific genes (Fig 1; ADAMS et al. 1988 ; CHANG and TIMBERLAKE 1992 ). Genes affecting brlA expression have been identified and characterized and these include fluG, flbA, flbB, flbC, flbD, and flbE. Mutations in all of these genes result in undifferentiated fluffy colonies (WIESER et al. 1994 ). Two of the delayed conidiation loci, flbC and flbD, are predicted to encode DNA-binding proteins and represent potential direct activators of brlA expression (WIESER and ADAMS 1995 ). Because mutations in flbC and flbD have additive effects on development, it has been proposed that these genes control independent steps in a nonlinear pathway (Fig 1). The flbB gene is predicted to encode a bZip-like transcription factor and flbE does not have clear homologs in the available databases (J. WIESER and T. H. ADAMS, personal communication). By testing the genetic requirements for the inappropriate conidiation observed following overexpression of fluG, flbA, or flbD, the gene order, fluG flbE flbD flbB, has been proposed (WIESER and ADAMS 1995 ). As fluG functions first in this regulatory network, fluG overexpression requires the activities of flbC, flbA, and flbD for activating brlA and causing conidiation in submerged culture (LEE and ADAMS 1996 ).

The fluG gene encodes a cytoplasmically localized 96-kD (864 amino acid) protein that is present at relatively constant levels throughout the life cycle (LEE and ADAMS 1994B ). The C-terminal half of FluG shares 30% identity with prokaryotic glutamine synthetase I whereas the N-terminal half of the protein shares no significant similarity with any functionally characterized proteins in the databases. Moreover, the entire N-terminal region could be deleted without affecting sporulation (D'SOUZA et al. 2001 ), indicating that FluG activity resides in the C-terminal half of the protein, which could be involved in constitutive synthesis of a small diffusible molecule related to glutamine or glutamate (LEE and ADAMS 1994B ). Although FluG activity is known to be required for activation of conidiation and for post-transcriptional activation of FlbA (ADAMS et al. 1988 ; WIESER and ADAMS 1995 ; LEE and ADAMS 1996 ), specific molecular events responding to fluG activity and leading to the developmental switch from vegetative growth remain to be uncovered.

In this article, we describe the isolation, characterization, and genetic analyses of 40 suppressors of fluG (SFGs) that bypass the need of fluG in conidiation and production of ST in A. nidulans. Such second site mutations will be extremely useful to further dissect early regulatory mechanisms of asexual sporulation. As the fluG-dependent asexual sporulation is independent of and parallel to FadA- and SfaD:G-mediated growth signaling, no mutations in FadA or SfaD were able to suppress fluG (YU et al. 1996 ; ROSEN et al. 1999 ). Thus, these SFGs will likely identify genes that specifically function in the conidiation regulatory pathway downstream of FluG. The fact that at least four loci are defined by 39 sfg mutations and all but 1 of the sfg mutations are found to be recessive strongly indicates that genes defined by sfg mutations likely function as negative regulators of conidiation. Moreover, all SFGs are found to produce ST in the absence of fluG activity, suggesting that these sfg genes may also negatively affect flbA activity and mutations in sfg genes might cause at least partial inhibition of FadA growth signaling to allow ST production to occur. A new model for upstream regulation of asexual development and growth is also presented.

	MATERIALS AND METHODS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Aspergillus strains, media, growth conditions, and genetic analysis:
The A. nidulans strains used in this study are listed in Table 1. Genotypes of SFGs are essentially the same (pabaA1, yA2; fluG::trpC⁺; trpC801, veA1; sfg^S) except for the sfg locus and mutant alleles. Standard culture and genetic techniques were employed (PONTECORVO et al. 1953 ; KAFER 1977 ). The growth rates of SFGs were checked on minimal medium (MM) from 2 to 5 days. All liquid cultures were inoculated with 5 x 10⁷ spores in 100 ml of liquid MM or MM with 0.1% yeast extract (YE) with 250 rpm shaking at 37°. Submerged development in each SFG was observed under a microscope at 1-hr intervals after an initial 18-hr growth period in liquid shake culture. Asexual developmental induction was performed as previously described (ADAMS et al. 1988 ) and samples for RNA isolation were collected at designated time points.

All SFGs were crossed with a developmentally wild-type strain, FGSC773. From these crosses, fluG;sfg^S strains carrying the pyrG89 allele were also isolated to carry out pairwise crosses to determine the number of loci and for future gene cloning (Table 1). The dominance or recessiveness of each sfg mutation was tested by generating diploid strains between each SFG strain (fluG;sfg^S) and RJH128 (fluG;sfg^WT) or RJA4.4 (fluG;sfg^WT).

Mutagenesis and isolation of SFGs:
A fluG strain (TTA127.4) was point inoculated on supplemented solid MM and incubated at room temperature for 7–10 days and conidia were collected for mutagenesis. Approximately 10⁸ conidiospores of TTA127.4 were treated with 1 mg/ml or 10 mg/ml of 4-nitroquinoline-1-oxide (4-NQO; BAL et al. 1977 ) for 0, 30, and 60 min, respectively, as previously described (WIESER et al. 1994 ). The mean survival rate of the 1-mg 4-NQO with a 30-min treatment was 67% and survivors of this condition were further screened.

ST extraction and TLC analysis:
Spores (10⁶) of each SFG were inoculated into 2 ml liquid complete medium (CM) in 8-ml tubes and the stationary cultures were incubated at 37° for 7 days as previously described (YU and LEONARD 1995 ). ST was extracted from 7-day-old cultures by adding 1 ml of CHCl₃ to 8-ml tubes and vortexing for 1–2 min. The organic phase was transferred to 1.5-ml tubes and centrifuged at 500 x g for 5 min. The CHCl₃ layer was collected, dried, and resuspended in 50 µl of CHCl₃ and 4–5 µl of each sample was applied onto a thin-layer chromatography (TLC) silica plate containing fluorescence indicator (Kiesel gel 60, 20 cm x 20 cm, 0.25 mm thick; Merck). ST standard was purchased from Sigma and 5 µg was applied onto the TLC plate with other samples. The plate was then developed with toluene:ethyl acetate:acetic acid (80:10:10, v/v/v), where the R_f value of ST is 0.65. At this step ST exhibits dark red color under the long-wave UV (320 nm). To enhance visibility and detection limit of ST, aluminum chloride (20% AlCl₃ · 6H₂O in 95% ethanol) is sprayed on to the TLC plate and the plate is baked at 80° for 5 min. The color of ST changes from red to exhibit bright light green by this process (STACK and RODRICK 1971). To compare ST production with stcU mRNA levels (presented in Fig 5), duplicate samples were prepared and collected from days 1–4, one for total RNA isolation and the other for ST extraction as described above.

View larger version (144K): In this window In a new window Download PPT slide   Figure 2. Phenotypes of SFGs. Each top row shows the colonies of wild type, fluG, and designated SFG strains. The bottom shows the close-up views of the colonies (40x magnification). Note that a fluG strain (TTA127.4) shows undifferentiated hyphal mass (fluffy) whereas SFG strains show varying levels of restored conidiation. Most of the delayed sporulators restore conidiation to near wild-type levels within 5 days after inoculation on solid medium. Photographs show colonies that were point-inoculated onto MM with 0.1% YE and incubated for 3 days at 37°. Arrows indicate conidiophores in delayed sporulators.

View larger version (114K): In this window In a new window Download PPT slide   Figure 3. Mutations in sfgA or sfgC result in hyperactive conidiation. SFG44 (fluG; sfgA44) and SFG5 (fluG; sfgC5) started to form vesicles (V) within 18–20 hr and produced complete conidiophores after 20–22 hr of incubation. Like most SFGs, SFG38 (fluG; sfgA38) and SFG51 (fluG; sfgA51) do not produce conidiophores in liquid submerged culture. Approximately 5 x 107 conidia were inoculated into 100 ml of liquid MM with supplements and 0.1% YE in 250-ml flasks and incubated at 37° with 250 rpm shaking. Photographs were taken at 22 hr of submerged culture. C, conidia; S, sterigmata; V, vesicle.

View larger version (110K): In this window In a new window Download PPT slide   Figure 4. SFGs restore accumulation of brlA and stcU transcripts. While brlA and stcU transcripts were readily detected in wild-type strains upon asexual developmental induction (Asexual), no brlA or stcU transcripts are detected at any time points of a fluG strain. As SFG5 (fluG; sfgC5) produces conidiophores within 22 hr postinoculation in submerged culture, it shows brlA mRNA accumulation at 24 hr of vegetative growth (Veg) as well as 12 and 24 hr post-asexual induction. SFG38 (fluG; sfgA38) shows brlA mRNA accumulation at 12 and 24 hr postinduction but not in vegetative cultures. SFG44 and SFG51 show similar Northern blot results to SFG5 and SFG38, respectively (data not shown). While wild type shows the highest level of stcU transcript accumulation at 8 hr postinduction, SFG5 and SFG38 accumulate stcU in a delayed manner.

View larger version (84K): In this window In a new window Download PPT slide   Figure 5. Sterigmatocystin (ST) production and stcU mRNA accumulation in SFGs. Duplicates of wild type (FGSC26), TTA127.4 (fluG), SFG5 (fluG; sfgC5), SFG38 (fluG; sfgA38), SFG44 (fluG; sfgA44), and SFG51 (fluG; sfgA51) were grown in stationary liquid CM for 1, 2, 3, and 4 days at 37°. One set was then used for ST analysis and the other was used for total RNA isolation. An isogenic wild-type strain (RKH51.117) was also examined and showed the same ST accumulation patterns as FGSC26 (not shown). The top shows ST analysis on TLC and arrows indicate the position of ST determined by the ST standard. Accumulation of ST was confirmed by (1) the migration distance (Rf value 0.65), (2) dark red color of the spot before AlCl3 spray, and (3) bright light green fluorescence after spraying AlCl3 (see MATERIALS AND METHODS). SFGs with sfgC5, sfgA44, and sfgA51 mutations accumulated relatively low amounts of ST until 4 days. Bright spots below and above ST shown in these SFGs, which have different Rf values and colors (light blue), are not ST but unknown compounds. Although low amounts of ST and increased accumulation of unknown compounds in these SFGs make it difficult to clearly view ST accumulation, they all produced ST at day 4 and even accumulated ST at near wild-type levels at 7 days of culture. Northern blot analyses of stcU mRNA are shown at the bottom. Equal loading of total RNA was evaluated by ethidium bromide staining of rRNAs.

Extracellular rescue of sporulation defect of fluG by SFGs:
The green conidial fluG strains (RJA23.1 and RJA23.2) were point inoculated at the center of MM with 0.5% YE and three SFGs were inoculated both sides of each fluG strain in duplicate. The strains were incubated at 37° for 2–3 days and examined under a stereomicroscope for possible extracellular rescue as previously described (LEE and ADAMS 1994B ).

Nucleic acid isolation and manipulation:
Total RNA was isolated by adding 0.6 ml of silica/zirconium beads (Biospec, Bartlesville, OK) and 1 ml of Trizol (Invitrogen, San Diego) and homogenizing in a Mini Bead Beater (Biospec) for 2 min and then subsequently following manufacturer's instructions (Invitrogen). Total RNA (15 µg/lane) was separated by electrophoresis using a 1.1% agarose gel containing 6% formaldehyde and ethidium bromide and the nucleic acids were transferred to a MagnaProbe Nylon membrane (0.45 µm; Osmonics, Minnetonka, MN). Probes for brlA and stcU mRNA were prepared by amplifying coding regions of brlA and stcU from wild-type (FGSC4) genomic DNA. A 1.48-kb brlA and a 1.12-kb stcU (BROWN et al. 1996 ) amplicon were labeled with ³²P-dCTP using a random-priming kit (Promega, Madison, WI) and used as probes for Northern blot analyses. Hybridization was carried out using modified Church buffer (1 mM EDTA, 0.25 M Na₂HPO₄7H₂O, 1% hydrolysated casein, 7% SDS, adjusted to pH 7.4 with 85% H₃PO₄; YU and LEONARD 1995 ).

Genomic DNA of wild-type and SFG strains was isolated by adding 0.3–0.5 ml of silica/zirconium beads, 0.5 ml of breaking buffer [2% Triton X-100, 1% SDS, 100 mM NaCl, 10 mM Tris-HCl (pH 8.0), 1 mM EDTA], and 0.5 ml of phenol:chloroform:isoamylalcohol (25:24:1) to mycelial samples followed by homogenizing in a Mini Bead Beater for 2 min. DNA in the aqueous phase was collected and ethanol precipitated. The purified genomic DNA in 50 µl of Tris-EDTA buffer was diluted 10 times for PCR reactions.

Microscopy:
Photomicrographs were taken using an Olympus BH2 compound microscope with the Kodak MDS290 digital imaging system. All other photographs were taken using a SONY digital camera (DSC-F707).

	RESULTS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Isolation of suppressors of fluG:
To gain further insights into the molecular events arising from fluG activity and leading to activation of developmental switch, we took an unbiased approach involving the isolation of second-site mutations that overcome the sporulation defects of a loss of fluG function. Unlike the previous study where a homozygous fluG diploid strain was employed to specifically look for dominant suppressor mutations (D'SOUZA et al. 2001 ), we used a fluG haploid strain (TTA127.4), expecting that various recessive and/or dominant suppressor mutants would be isolated. We visually screened 125,000 survivors from 4-NQO mutagenesis and isolated 40 putative suppressors of fluG (SFGs) that restored conidiation to clearly distinguishable levels. These 40 SFG mutants show varying levels of conidiation recovery and on the basis of the phenotypes on solid MM, the mutants were tentatively grouped into three classes: 9 high sporulators (HS), 18 wild-type level sporulators (W), and 13 delayed sporulators (DS). Delayed sporulators initially resemble the fluG mutant for 2–3 days, but most of them achieve wild-type sporulation levels within 5 days. Representative SFG strains are presented in Fig 2. To be consistent, we designate a mutant strain with SFG#, a mutant allele with sfg#, suppressor mutations collectively with sfg^S, and a wild-type allele with sfg^WT.

Individual suppression is caused by a single second-site mutation unlinked to fluG:
As a first step in genetic analyses, each SFG strain was meiotically crossed with a developmentally wild-type strain (FGSC773) and all of the crosses generated matured cleistothecia. If suppression was caused by a single second-site mutation, segregation of the relevant genotypes (with corresponding phenotypes) of progeny would be fluG;sfg^S (conidiating), fluG⁺;sfg^S (probably conidiating like wild type), fluG;sfg^WT (fluffy due to fluG), and fluG⁺;sfg^WT (wild type), thus generating 25% fluffy progeny. All of our successful crosses showed 25% recovery of fluG (fluffy phenotype) progeny, indicating that a single gene mutation, not linked to fluG, caused individual suppression. In addition, no new phenotypes among the conidiating progeny were evident, indicating that individual sfg^S mutation does not cause readily detectable morphological changes with respect to wild-type FluG function. In these crosses, we have also isolated multiple recombinant SFG strains carrying the auxotrophic marker pyrG89 and these were used to determine the number of sfg loci (see below). Isolation of fluG; sfg^S; pyrG89 strains was accomplished by examining 10 independent uracil-requiring conidiating progeny from each cross for the fluG pattern by genomic DNA PCR. Conidial strains with the fluG PCR pattern are expected to be fluG with sfg^S (Table 1).

Dominance and recessiveness of sfg^S:
As a prerequisite of the SFG gene identification we tested whether each sfg^S is dominant or recessive to its wild-type allele by generating diploids that are homozygous for fluG and heterozygous for sfg (sfg^S; sfg^WT) by fusing an individual SFG strain with another fluG strain, RJH128 or RJA4.4. Diploid strains of 39 SFGs exhibited a fluffy phenotype like fluG, indicating that these sfg^S mutations are recessive to their wild-type alleles (Table 2). One diploid strain (dSFG38) sporulated to wild-type level, suggesting that sfg38 defines a dominant (interfering) mutation, which was later found to be an allele of recessive mutants (see below).

At least four loci are defined by sfg mutations:
In an attempt to determine the number of genes defined by sfg mutations, each primary SFG strain was meiotically crossed with recombinant SFG strains carrying different auxotrophic markers (fluG; sfg^S; pyrG89; and/or pyroA4). While pairwise crosses of two allelic (or tightly linked) sfg^S mutations would produce only conidial progeny, crosses of nonallelic sfg^S mutations are expected to generate 25% fluffy progeny due to fluG (sfg^wt1;sfg^wt2;fluG). The results of pairwise crosses showed that 39 sfg^S mutations represented at least four linkage groups (sfgA D), where 31 mapped to sfgA, 6 mapped to sfgB, and 1 each mapped to sfgC and sfgD, respectively (Table 1 and Table 2). Although SFG8 could not be assigned to a specific linkage group due to extreme difficulty in sexual crosses, it was found not to be an allele of sfgA or sfgC. Isolation of 31 suppressors mapped to sfgA (or a tightly linked locus) presents a strongly biased distribution of suppressor mutations in sfgA. Moreover, mutations in the sfgA locus seem to result in varying ranges of recovered conidiation (Fig 2; Table 2; HS, W, and DS) indicating that partial loss of sfgA function might be sufficient to cause suppression of fluG.

Two previously reported mutations, sfdA15 and sfdB38, which were originally isolated as suppressors of flbD, were found to suppress fluG (KELLNER and ADAMS 2002 ). Thus, it was of interest to see whether these sfd mutations could be mapped to any of four sfg loci. REK65.13 and REK88.23 were meiotically crossed with SFG strains from each linkage group and sfdA15 was found not to be linked to sfgA, sfgB, or sfgC. Meiotic crosses between sfdB38 mutant and selected SFG strains or sfdA15 with sfgD mutants failed to generate matured cleistothecia.

Mutations in sfgA and sfgC cause submerged conidiation:
One of the phenotypic characteristics of hyperactive conidiation is the formation of conidiophores in liquid submerged culture, conditions under which wild-type strains do not sporulate. To test whether some SFGs show hyperactive conidiation even in the absence of FluG activity, we examined an individual SFG strain's ability to form conidiophores in liquid shake culture and found that 14 SFGs elaborated conidiophores in submerged culture within 25 hr, where 13 and 1 belong to the linkage groups A and C, respectively. Particularly, SFG43 (fluG; sfgA43) and SFG44 (fluG; sfgA44) began to form vesicles at 18 hr and produced complete conidiophores within 20 hr (Table 2; Fig 3). SFG5 (fluG; sfgC5) produced complete conidiophores within 22 hr in liquid MM with 0.1% YE, but not in liquid MM alone (Fig 3). SFG5 exhibited delayed conidiation phenotype with enhanced growth (25% more than wild type) on solid MM (Fig 2 and Table 2). Addition of YE to SFG5 cultures, however, caused reduced growth (Table 2), slightly increased conidiation on solid medium (not shown), and inappropriate conidiation in liquid culture (Fig 3). None of the mutants belonging to linkage group B or D produced conidiophores in submerged culture.

We selected four SFG mutants, SFG5, SFG38, SFG44, and SFG51, and examined accumulation of brlA transcript at various stages of the life cycle. As presented in Fig 4, a wild-type strain (FGSC26) does not show brlA transcript accumulation during vegetative growth phase or before 8 hr post-asexual induction. However, SFG5, which elaborated conidiophores within 22 hr of vegetative growth, accumulated brlA transcript at 24 hr in liquid culture and at 12 and 24 hr post-asexual developmental induction (Fig 4). Almost identical brlA accumulation patterns were observed for SFG44 (not shown). Both SFG5 and SFG38 strains exhibit an 4-hr delay of brlA (and stcU, see below) transcript accumulation. SFG51 (a delayed sporulator) accumulated detectable levels of brlA transcript at 24 hr postinduction (not shown). These data show that the timing and levels of brlA transcript accumulation upon asexual developmental induction are closely related to phenotypes of SFGs on solid medium. Unlike SFG5 and SFG44, SFG38 and SFG51 do not form conidiophores in liquid submerged culture.

Most sfg mutations affect hyphal growth:
Asexual development and hyphal growth are antagonistic in that elevation of one process causes downregulation of the other process. The fact that SFGs bypass the requirement for fluG in asexual development and that some SFGs exhibit a hyperactive conidiation phenotype leads us to think that some sfg mutations might also affect hyphal growth, probably due to elevated asexual development. To test this, growth rates of each SFG strain on solid MM and MM with 0.1% YE (not shown) were measured in triplicate and compared with those of a wild-type strain (FGSC26). As shown in Table 2, SFG mutants show varying levels of growth, 49–125% of that of wild type, and all SFGs but five (SFG2, SFG5, SFG36, SFG46, and SFG50) show reduced hyphal growth compared to wild type. Particularly, two sfgB mutant alleles, sfgB27 and sfgB52, caused reduction of growth rates to 50% of wild type on MM. The fact that mutations in sfgB cause reduced growth yet have low levels of conidiation (delayed conidiation) suggests that sfgB might elucidate a new cross-talking network between growth and asexual development.

No sfg mutations extracellularly rescue the conidiation defect of fluG:
One of the phenotypic characteristics of fluG strains is that the conidiation defect can be rescued by growing fluG strains next to either wild-type or other developmental mutants (LEE and ADAMS 1994B ). We tested whether any of 40 SFGs could extracellularly rescue the sporulation defect of fluG strains. All primary SFGs originated from TTA127.4 (pabaA1, yA2; fluG::trpC⁺; trpC801, veA1) and produce yellow conidia due to the yA2 mutation. Thus, for efficient rescue experiments, we generated two fluG strains that produce green conidia (RJA23.1 and RJA23.2; see Table 1) and tested all SFGs in duplicate and found that no SFGs were able to rescue the conidiation defect of fluG strains.

All SFG mutants regain the ability to produce ST:
Previously, it was shown that FluG is required for the production of the mutagenic and carcinogenic mycotoxin ST in A. nidulans. It has been proposed that this requirement for FluG is via activating FlbA, which in turn inactivates FadA growth signaling (HICKS et al. 1997 ). To test whether SFGs regain the ability to produce ST, we initially examined 40 SFGs for ST production at 7 days of culture and found that all were able to produce ST at near wild-type levels without fluG activity (data not shown). Two developmentally wild-type strains, FGSC26 and RKH51.117 (an isogenic wild-type strain, Table 1), have been compared with SFG strains and both wild-type strains showed the same ST accumulation patterns. To confirm that ST recovery is through restored stc gene expression we examined the accumulation of ST and stcU mRNA in selected SFGs days 1–4 as previously described (YU and LEONARD 1995 ). As presented in Fig 5, while the fluG; sfg^WT strain (TTA127.4) does not accumulate stcU transcript or ST, four selected SFGs, including dominant SFG38, show certain levels of restored stcU mRNA and ST accumulation. Although until day 4 some SFGs produced less ST than did wild type, they all produced ST at near wild-type levels at 7 days of culture. We also examined stcU transcript accumulation in SFG5, SFG38, wild-type, and fluG strains at various stages of the life cycle. SFG5 does not show stcU transcript accumulation in liquid culture, even though it conidiates and accumulates brlA transcript (Fig 4). Upon induction of asexual development, however, both SFG5 and SFG38 show similar patterns of brlA and stcU transcript accumulation (Fig 4). Recovery of ST production is an important difference between sfg^S mutations and the previously reported dominant suppressor of fluG, dsgA1, because dsgA1 cannot bypass the need for fluG in ST production (D'SOUZA et al. 2001 ). These results indicate that sfg^S mutations, even without FluG activity, can (at least partially) inhibit FadA growth signaling and allow ST biosynthesis to occur.

Dominant negative sfgA38 is different from dsgA1:
At present, only two (sfgA38 and dsgA1) dominant suppressors of fluG have been isolated. As described, sfg38 and 30 other recessive sfg mutations mapped to linkage group A, suggesting that sfgA38 is likely a dominant interfering (negative) mutant allele. We attempted to test whether dsgA1 can be mapped to any of the four sfg linkage groups by meiotic crosses between SFGs and a fluG; dsgA1 strain (HDCD15.1). While HDCD15.1 readily formed heterokaryons with most fluG; sfg^S mutant strains, no matured cleistothecia were formed even under conditions that enhance sexual development (D'SOUZA et al. 2001 ).

Despite unsuccessful crosses between SFGs and HDCD15.1, on the basis of their clear phenotypic differences, one can speculate that sfgA38 and dsgA1 might define different genes. As shown in Fig 6, HDCD15.1 readily generates fluffy sectors consistently whereas SFG38 never does that. The fluffy sector stays fluffy in subsequent generations, indicating that the dominant nature of the dsgA1 mutation is somehow lost permanently in this sector. While SFG38 does not form fluffy sectors, it produces circles of enriched cleistothecia (Fig 6C). The interval and width of cleistothecia-enriched bands indicate that SFG38 likely undergoes elevated sexual development at 24-hr intervals, each interval lasting 12 hr. Additional critical differences between sfgA38 and dsgA1 are: (1) while sfgA38 restores ST production to wild-type level without fluG, dsgA1 requires fluG for ST production; and (2) sfgA38 does not cause conidiophore formation in liquid submerged culture whereas dsgA1 causes formation of conidiophores in the absence of fluG (D'SOUZA et al. 2001 ).

View larger version (112K): In this window In a new window Download PPT slide   Figure 6. Phenotypes of dominant fluG suppressor mutants. Colonies of two dominant fluG suppressors, HDCD15.1 (fluG; dsgA1) and SFG38 (fluG; sfgA38), grown on MM for 7 days at 37° are shown. Magnified images (40x) of conidial (a) and fluffy (b) regions of HDCD15.1 and a region of enriched cleistothecia (c) in SFG38 are also presented. CP, conidiophores; H, hyphae; CLS, cleistothecia.

	DISCUSSION

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

The genus Aspergillus encompasses the most common fungi in our environment. Members of this genus reproduce asexually by forming long chains of conidiospores radiating from a central structure known as a conidiophore. One of the primary questions has been how such a complex structure differentiates from vegetatively growing hyphae. Previous studies in A. nidulans showed that FluG is required for this developmental switch by activating the conidiation process and indirectly regulating the G protein-mediated growth-signaling cascade via activating FlbA (LEE and ADAMS 1994A , LEE and ADAMS 1996 ; YU et al. 1996 ). Balanced control of these two parallel signaling pathways is essential for maintaining the genetically programmed life cycle. For the last decade, heterotrimeric G proteins and their signaling/regulatory mechanisms have been intensively studied and are found to be conserved from yeasts to humans. Such structural and functional conservancy of the G protein signaling components makes reverse genetics and functional genomics favorable approaches. However, genes functioning in the conidiation process and associated regulatory mechanisms are largely unknown primarily due to uniqueness of the conidiation processes in filamentous fungi. Moreover, like FluG, many genes are expected to be novel in their structures and functions. Therefore, we believe that unbiased genetic studies represent the most suitable way to further dissect upstream regulatory mechanisms of conidiation in A. nidulans.

FluG is required for activation of conidiation and it functions upstream of other developmentally specific genes including flbE, flbD, flbC, flbB, and brlA (see Fig 1). Previous genetic studies looking for developmentally defective mutants from a wild-type strain were specifically aimed at the identification of positive regulators of conidiation. A recent study of fluG describing the isolation of dominant suppressors of fluG was also biased to the identification of activating components of conidiation (D'SOUZA et al. 2001 ). If negative elements were involved in the regulatory cascade of conidiation, previous studies would have missed these critical components. Thus, to investigate molecular mechanisms associated with the FluG-mediated activation of conidiation in an unbiased way, we have isolated and characterized a large number of SFGs, employing a haploid fluG strain. Because the fluG-dependent initiation of asexual sporulation is independent of and parallel to FadA- and SfaD:G-mediated growth signaling and no mutations in FadA or SfaD are able to suppress fluG (YU et al. 1996 , YU et al. 1999 ; ROSEN et al. 1999 ), genes presented by suppressors of fluG are expected to specifically function in the conidiation pathway downstream of FluG.

Characterization and genetic analyses of 40 SFG mutants have been carried out. Because the fluG deletion mutant was used for the isolation of suppressor, all SFGs are expected to be extragenic and bypass suppressors of fluG. Genetic analyses of SFGs can be summarized as follows: (1) each SFG is derived from a single second-site mutation; (2) 39 sfg mutations are recessive to their wild-type alleles and only 1 (sfgA38) is dominant; (3) at least four loci are defined by sfg^S mutations; (4) 31, 6, 1, and 1 mutations are mapped to linkage groups sfgA, sfgB, sfgC, and sfgD, respectively; and (5) the dominant mutation sfgA38 is an allele of sfgA and is different from dsgA1. The fact that at least four loci are defined by recessive sfg mutations indicates that multiple genes are involved in negative control of conidiation downstream of fluG, which supports our reason for using a haploid fluG strain. In this study, the most interesting findings are that 31 SFGs including the dominant mutant SFG38 are mapped to the sfgA linkage group and that they show varying levels of recovered conidiation (Table 2), where 28 are wild-type level or hyperactive sporulators and 4 are delayed sporulators. These results strongly indicate that sfgA functions as a key negative regulator of conidiation and it might have multiple functional domains. Any mutations causing (at least partial) loss of sfgA function(s) may be sufficient to restore conidiation to certain levels. Depending on the levels of remaining functionality of the SfgA mutant products, varying ranges of suppression, i.e., delayed sporulation to hyperactive sporulation, might result. On the other hand, incremental loss of sfgA function would result in elevated levels of restored conidiation and even hyperactive conidiation. Supporting this idea, a relatively large number (13 of 31) of sfgA mutations are found to cause submerged conidiation in the absence of fluG activity. Unlike sfgA mutations, however, most mutations in sfgB seem to result in delayed conidiation, suggesting that partial (or even complete) loss of sfgB function might not be sufficient to cause full recovery of conidiation. Furthermore, the fact that only one suppressor mutation each has been mapped to sfgC or sfgD after screening 125,000 survivors suggests that only specific mutation(s) in sfgC or sfgD, e.g., a complete loss of function, might bypass fluG function. Similar to sfgA, a complete loss of sfgC function might be sufficient to cause hyperactive conidiation, but only with yeast extract (see Fig 3). However, regardless of levels of recovered conidiation, all SFG mutants regained the ability to produce ST to near wild-type levels, suggesting that all sfg mutations could cause (at least partial) activation of FlbA.

On the basis of our findings, a new genetic model for upstream regulation of asexual development in A. nidulans is proposed (Fig 7). In this model, dsgA is positioned in the box of conidiation-specific functions because dsgA1 suppresses conidiation but not the ST production defects of fluG (D'SOUZA et al. 2001 ). While genetic interactions among four sfg genes need to be studied, the products of the sfgA, sfgB, sfgC, and sfgD genes, represented as [Sfg], negatively regulate conidiation-specific functions as well as certain FlbA functions, which are necessary for ST production through inhibition of FadA growth signaling. We propose that the primary role of FluG is to remove these repressive effects imposed by multiple sfg genes. In the proposed model, a complete loss of negative regulation by the sfg genes would cause a full elimination of repressive effects, resulting in hyperactive sporulation as seen in 13 sfgA and 1 sfgC5 mutants. Conversely, partial functions of the sfg genes would maintain certain levels of negative regulation, which cause low levels of remediation of conidiation, i.e., delayed conidiation.

View larger version (15K): In this window In a new window Download PPT slide   Figure 7. A new model for upstream regulation of asexual sporulation. DsgA is positioned within the developmental-specific functions that do not affect FlbA-mediated ST remediation. FluG functions the most upstream and its primary role is to remove repressive effects imposed by downstream sfg genes. Elimination of Sfg-mediated negative regulation is necessary for activation of conidiation-specific functions and FlbA, which in turn confers ST production.

However, removal of negative regulation is not sufficient to alter developmental competence. Previously, it has been shown that conidiation does not occur until cells have gone through a defined period of vegetative growth (18 hr), during which cells acquire competence to respond to developmental signaling or induction (for review see ADAMS et al. 1998 ). Although many SFGs produced conidiophores in liquid culture in a relatively short time, no SFGs were able to shorten the proposed time (18 hr) for acquisition of developmental competence. Furthermore, no mutations in FadA or SfaD result in earlier than 20-hr conidiophore development in liquid culture (YU et al. 1996 , YU et al. 1999 ; ROSEN et al. 1999 ). These results are consistent with the observation that overexpression of fluG does not alter the time required for conidiophore development to begin in an air-exposed colony (LEE and ADAMS 1996 ). In contrast, dsgA1 causes submerged conidiation between 9 and 11 hr of incubation even without FluG activity, indicating that time required for competence in a dsgA1 mutant is greatly shortened (D'SOUZA et al. 2001 ). Collectively, acquisition of developmental competence may be defined by activation of downstream developmentally specific functions, which involve coordination of upstream positive and negative regulatory functions.

It has been proposed that FluG is required for the production of the extracellular sporulation factor. The fact that both FluG mRNA and protein are present at relatively constant levels throughout the life cycle implies that accumulation of the factor above a certain level (threshold) might be necessary to trigger the switch from vegetative growth to conidiation (LEE and ADAMS 1996 ; D'SOUZA et al. 2001 ). No SFG mutants were able to rescue the conidiation defect of fluG by proximal growth, suggesting that suppression of fluG is due to alterations in intracellular regulation, not through the recovery of the production of the extracellular sporulation factor(s). Overexpression of fluG results in inappropriate production of conidiophores that are remarkably similar to wild-type conidiophores and have all the cell types including stalks, vesicles, metulae, phialides, and conidia, indicating that activation of fluG results in activation of all the genes necessary to form a complete conidiophore (LEE and ADAMS 1996 ). Similarly, 14 SFGs produce complete conidiophores within 25 hr of liquid submerged culture, implying that sfg genes likely function immediately downstream of fluG (see Fig 3). Taken together, we propose that sfg genes act to repress conidiation during the early vegetative growth phase and that accumulation of the sporulation factor above certain levels serves as a signal to remove this negative regulation.

	FOOTNOTES

¹ Present address: Department of Environmental Health Sciences, School of Public Health, University of Michigan, 109 Observatory St., 1506 SPH I, Ann Arbor, MI 48109.

	ACKNOWLEDGMENTS

We thank Guiping Yang for her assistance on characterization of SFGs and colleagues in the laboratory for helpful discussions and suggestions. Special thanks go to Byron Brehm-Stecher and Ellin Doyle in our institute for critically reviewing the manuscript. This work was supported in part by sponsors of the Food Research Institute and by a Hatch Grant to J.H.Y. from the College of Agriculture & Life Sciences at University of Wisconsin, Madison.

Manuscript received November 19, 2002; Accepted for publication July 17, 2003.

	LITERATURE CITED

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

ADAMS, T. H., 1994 Asexual sporulation in higher fungi, pp. 367–382 in The Growing Fungus, edited by N. A. R. GOW and G. M. GADD. Chapman & Hall, London.

ADAMS, T. H., M. T. BOYLAN, and W. E. TIMBERLAKE, 1988  brlA is necessary and sufficient to direct conidiophore development in Aspergillus nidulans.. Cell 54:353-362.[Medline]

ADAMS, T. H., J. K. WIESER, and J.-H. YU, 1998  Asexual sporulation in Aspergillus nidulans.. Microbiol. Mol. Biol. 12:3827-3833.

BAL, J., E. M. KAJTANIAK, and N. J. PIENIAZEK, 1977  4-nitroquinoline-1-oxide: a good mutagen for Aspergillus nidulans.. Mutat. Res. 56:153-156.

BERMAN, D. M., T. M. WILKIE, and A. G. GILMAN, 1996  GAIP and RGS4 are GTPase activating proteins (GAPs) for the Gi subfamily of G protein subunits. Cell 86:445-452.[Medline]

BROWN, D. W., J.-H. YU, H. S. KELKAR, M. FERNANDES, and T. C. NESBITT et al., 1996  Twenty-five co-regulated transcripts define a sterigmatocystin gene cluster in Aspergillus nidulans.. Proc. Natl. Acad. Sci. USA 93:1418-1422.[Abstract/Free Full Text]

CHANG, Y. G. and W. E. TIMBERLAKE, 1992  Identification of Aspergillus brlA response elements (BREs) by genetic selection in yeast. Genetics 133:29-38.

D'SOUZA, C. A., B. N. LEE, and T. H. ADAMS, 2001  Characterization of the role of the FluG protein in asexual development of Aspergillus nidulans.. Genetics 158:1027-1036.[Abstract/Free Full Text]

HEPLER, J. R., D. M. BERMAN, A. G. GILMAN, and T. KOZASA, 1997  RGS4 and GAIP are GTPase-activating proteins for Gq and block activation of phospholipase Cß by -thio-GTP-Gq. Proc. Natl. Acad. Sci. USA 94:428-432.[Abstract/Free Full Text]

HICKS, J. K., J.-H. YU, N. P. KELLER, and T. H. ADAMS, 1997  Aspergillus sporulation and mycotoxin production both require inactivation of the FadA G protein-dependent signaling pathway. EMBO J. 16:4916-4923.[Medline]

KÄFER, E., 1977  Meiotic and mitotic recombination in Aspergillus and its chromosomal aberrations. Adv. Genet. 19:33-131.[Medline]

KELLNER, E. M. and T. H. ADAMS, 2002  Mutations in sfdA and sfdB suppress multiple developmental mutations in Aspergillus nidulans.. Genetics 160:159-168.[Abstract/Free Full Text]

KOELLE, M. R. and H. R. HORVITZ, 1996  EGL-10 regulates G protein signaling in the C. elegans nervous system and shares a conserved domain with many mammalian proteins. Cell 84:115-125.[Medline]

LEE, B. N. and T. H. ADAMS, 1994a  Overexpression of flbA, an early regulator of Aspergillus asexual sporulation leads to activation of brlA and premature initiation of development. Mol. Microbiol. 14:323-334.[Medline]

LEE, B. N. and T. H. ADAMS, 1994b  The Aspergillus nidulans fluG gene is required for production of an extracellular developmental signal. Genes Dev. 8:641-651.[Abstract/Free Full Text]

LEE, B. N. and T. H. ADAMS, 1996  fluG and flbA function interdependently to initiate conidiophore development in Aspergillus nidulans through brlAß activation. EMBO J. 15:299-309.[Medline]

PONTECORVO, G., J. A. ROPER, L. M. HEMMONS, K. D. MACDONALD, and A. W. J. BUFTON, 1953  The genetics of Aspergillus nidulans.. Adv. Genet. 5:141-238.[Medline]

ROSEN, S., J.-H. YU, and T. H. ADAMS, 1999  The Aspergillus nidulans sfaD gene encodes a G protein ß subunit that is required for normal growth and repression of sporulation. EMBO J. 18:5592-5600.[Medline]

SHIMIZU, K. and N. P. KELLER, 2001  Genetic involvement of a cAMP-dependent protein kinase in a G protein signaling pathway regulating morphological and chemical transitions in Aspergillus nidulans.. Genetics 157:591-600.[Abstract/Free Full Text]

STACK, M. and J. V. RODRICKS, 1971  Method for analysis and chemical confirmation of sterigmatocystin. J. Assoc. Off. Anal. Chem. 54:86-90.[Medline]

WIESER, J. and T. H. ADAMS, 1995  flbD encodes a Myb-like DNA binding protein that regulates initiation of Aspergillus nidulans conidiophore development. Genes Dev. 9:491-502.[Abstract/Free Full Text]

WIESER, J., B. N. LEE, J. W. FONDON, and T. H. ADAMS, 1994  Genetic requirements for initiating asexual development in Aspergillus nidulans.. Curr. Genet. 32:218-224.

YU, J.-H. and T. J. LEONARD, 1995  Sterigmatocystin biosynthesis in Aspergillus nidulans requires a novel type I polyketide synthase. J. Bacteriol. 177:4792-4800.[Abstract/Free Full Text]

YU, J.-H., J. WIESER, and T. H. ADAMS, 1996  The Aspergillus FlbA RGS domain protein antagonizes G-protein signaling to block proliferation and allow development. EMBO J. 15:5184-5190.[Medline]

YU, J.-H., S. ROSEN, and T. H. ADAMS, 1999  Extragenic suppressors of loss-of-function mutations in the Aspergillus FlbA regulator of G-protein signaling domain protein. Genetics 151:97-105.[Abstract/Free Full Text]

This article has been cited by other articles:

				I. Vargas-Perez, O. Sanchez, L. Kawasaki, D. Georgellis, and J. Aguirre Response Regulators SrrA and SskA Are Central Components of a Phosphorelay System Involved in Stress Signal Transduction and Asexual Sporulation in Aspergillus nidulans Eukaryot. Cell, September 1, 2007; 6(9): 1570 - 1583. [Abstract] [Full Text] [PDF]

				J.-H. Yu, J.-H. Mah, and J.-A. Seo Growth and Developmental Control in the Model and Pathogenic Aspergilli Eukaryot. Cell, October 1, 2006; 5(10): 1577 - 1584. [Full Text] [PDF]

				J.-H. Mah and J.-H. Yu Upstream and Downstream Regulation of Asexual Development in Aspergillus fumigatus Eukaryot. Cell, October 1, 2006; 5(10): 1585 - 1595. [Abstract] [Full Text] [PDF]

				J.-A. Seo, Y. Guan, and J.-H. Yu FluG-Dependent Asexual Development in Aspergillus nidulans Occurs via Derepression Genetics, March 1, 2006; 172(3): 1535 - 1544. [Abstract] [Full Text] [PDF]

				J.-A. Seo and J.-H. Yu The Phosducin-Like Protein PhnA Is Required for G{beta}{gamma}-Mediated Signaling for Vegetative Growth, Developmental Control, and Toxin Biosynthesis in Aspergillus nidulans Eukaryot. Cell, February 1, 2006; 5(2): 400 - 410. [Abstract] [Full Text] [PDF]

				J.-A. Seo, K.-H. Han, and J.-H. Yu Multiple Roles of a Heterotrimeric G-Protein {gamma}-Subunit in Governing Growth and Development of Aspergillus nidulans Genetics, September 1, 2005; 171(1): 81 - 89. [Abstract] [Full Text] [PDF]

				M. Ni, S. Rierson, J.-A. Seo, and J.-H. Yu The pkaB Gene Encoding the Secondary Protein Kinase A Catalytic Subunit Has a Synthetic Lethal Interaction with pkaA and Plays Overlapping and Opposite Roles in Aspergillus nidulans Eukaryot. Cell, August 1, 2005; 4(8): 1465 - 1476. [Abstract] [Full Text] [PDF]