ABSTRACT
In the fungus Aspergillus nidulans, inactivation of the flbA to -E, fluG, fluF, and tmpA genes results in similar phenotypes, characterized by a delay in conidiophore and asexual spore production. flbB to -D encode transcription factors needed for proper expression of the brlA gene, which is essential for asexual development. However, recent evidence indicates that FlbB and FlbE also have nontranscriptional functions. Here we show that fluF1 is an allele of flbD which results in an R47P substitution. Amino acids C46 and R47 are highly conserved in FlbD and many other Myb proteins, and C46 has been proposed to mediate redox regulation. Comparison of ΔflbD and flbDR47P mutants uncovered a new and specific role for flbD during sexual development. While flbDR47P mutants retain partial function during conidiation, both ΔflbD and flbDR47P mutants are unable to develop the peridium, a specialized external tissue that differentiates during fruiting body formation and ends up surrounding the sexual spores. This function, unique among other fluffy genes, does not affect the viability of the naked ascospores produced by mutant strains. Notably, ascospore development in these mutants is still dependent on the NADPH oxidase NoxA. We generated R47K, C46D, C46S, and C46A mutant alleles and evaluated their effects on asexual and sexual development. Conidiation defects were most severe in ΔflbD mutants and stronger in R47P, C46D, and C46S strains than in R47K strains. In contrast, mutants carrying the flbDC46A allele exhibited conidiation defects in liquid culture only under nitrogen starvation conditions. The R47K, R47P, C46D, and C46S mutants failed to develop any peridial tissue, while the flbDC46A strain showed normal peridium development and increased cleistothecium formation. Our results show that FlbD regulates both asexual and sexual differentiation, suggesting that both processes require FlbD DNA binding activity and that FlbD is involved in the response to nitrogen starvation.
INTRODUCTION
All living organisms require sensing and integration of environmental signals in appropriate ways to generate responses that determine transitions between cell growth, programmed cell death, and differentiation. The filamentous fungus Aspergillus nidulans is an excellent model for studying these processes, as it presents complex but defined cell cycle transitions between growth (mycelia) and asexual (conidiation) and sexual (cleistothecium and ascospore development) differentiation and is amenable to genetic analysis.
A. nidulans asexual reproduction is induced by environmental signals such as exposure to air (2, 70) or nutrient starvation (63). It involves the production of chemical signals (9, 25, 38, 40, 44, 64, 72) and depends on activation of the brlA gene (10), which encodes a transcription factor (TF) of the Zn finger family (1, 29, 48, 54). Normally, conidiation involves the production of a mycelial cell compartment, from which a conidiophore stalk develops. After the stalk reaches a determined length, polar growth stops for development of a multinucleated vesicle, from which two successive uninucleate cell types are produced by budding (metulae and phialides). The phialides are conidiogenic cells that produce uninucleate spores through several rounds of consecutive mitosis (reviewed in references 2, 5, 16, and 81).
fluffy mutants, which show a cotton-like morphology and a notable delay in asexual development, were proposed to define genes upstream of brlA that are required for conidiation (13, 64, 68, 77). Indeed, extensive genetic analysis led to the identification of developmental regulators acting upstream of brlA (fluG and flbA to -E [flbA-E] genes) and required for its proper expression (77). FlbA is a regulator of a heterotrimeric G protein signaling pathway that stimulates vegetative growth and inhibits conidiation (83). In contrast, FluG is responsible for the production of an extracellular signaling factor required for activation of the flbD-E genes in vegetative cells and to induce conidiation, along with other unidentified compounds (38, 39, 44, 56, 64). FlbB is a TF of the bZIP type that unexpectedly also localizes at the hyphal tips of vegetative hyphae, forming a complex with FlbE (17, 18, 21). flbE encodes a small protein with two conserved but uncharacterized domains, and FlbE and FlbB are functionally interdependent and proposed to regulate the transition from vegetative growth to conidiation (20, 34). FlbB is found in nuclei, but notably, it is detected only at the most apical nucleus in vegetative hyphae (18). flbD encodes a TF of the Myb family (76) which, together with FlbB, is required to jointly activate brlA expression and conidiation (21). FlbC is a zinc finger TF that binds the brlA promoter in vitro, and the highest expression level of flbC correlates with cessation of apical extension and swelling of the conidiophore vesicle, which are phenotypes also observed for flbC overexpression (33). Two other mutants with delayed asexual development defined the tmpA and fluF genes. tmpA encodes a flavoenzyme required for the production of a sporulation chemical signal independent of the FluG factor (64). A fluF1 mutant was isolated after treatment with 5-azacytidine (68), but the identity of the gene remained unknown.
Frequently, sexual development (formation of fruiting bodies containing ascospores) occurs after conidiation, usually under low-nutrient and oxygen limitation conditions and in the absence of light (9). The first sign of sexual development is the formation of Hülle cells, which later surround the mature fruiting body, or cleistothecium. Cleistothecium formation involves the differentiation of a central ascogenous tissue that gives rise to asci and ascospores and the development of a network of sterile hyphae surrounding the ascogenous tissue, which finally develops into a melanized cleistothecial wall, or peridium (9, 62). Signaling involved in initiation of fruiting body development is mediated by membrane-bound G-protein-coupled receptors (GprA, GprB, and GprD) (23, 61). Besides signaling through the G protein (FadA, SfaD, and GpgA) (57, 61, 83), the mitogen-activated protein kinase (MAPK) module SakA-AtfA (31, 36), the COP9 signalosome (7), VeA (5, 32, 65, 81), and several transcription factors, including SteA (73), NsdD (23), StuA (47), CpcB (27), and NosA (75), are required for or regulate fruiting body formation. However, the target genes of these TFs are mostly unknown. The noxA gene, encoding an NADPH oxidase, is also essential for sexual differentiation. Null noxA mutants are blocked at a very early stage (primordia) of cleistothecium formation and fail to produce mature cleistothecia and ascospores (35), whereas in Neurospora crassa (8) and Podospora anserina (42), the elimination of the NADPH oxidase NOX-1 results in complete female sterility, indicating that NOX-generated reactive oxygen species (ROS) play important roles in cell differentiation.
How ROS signaling triggers differentiation remains a key question. However, several proteins, including protein tyrosine phosphatases and transcription factors, have been shown to be targets for oxidation by ROS (11, 74). TFs of the Myb family are found in animals, plants, fungi, protozoa, and algae (14, 45, 51, 58, 66, 76). These proteins play important roles in various cellular processes, such as cell proliferation, apoptosis, differentiation, metabolism, and environmental sensing. Myb TFs are regulated by various mechanisms, one of which involves redox regulation of an invariant cysteine residue whose reduction is essential for DNA binding (49).
Here we show that fluF and flbD are the same gene and that the fluF1 allele corresponds to a point mutation that results in an R47P replacement within the FlbD Myb domain. Furthermore, we report a new and specific role for flbD in sexual development by showing that flbD mutants fail to produce the cleistothecial peridium but nevertheless develop viable “naked” ascospores. Analyses of different flbD mutant alleles indicate that FlbD plays different roles in asexual and sexual development, depending on DNA binding activity, perhaps involving redox regulation of conserved Myb domain C46.
MATERIALS AND METHODS
Strains, media, and growth conditions.Strains used in this work are listed in Table 1. For flbD analysis, we used a strain carrying a truncated allele that would result in a FlbD protein lacking its C terminus (76), referred to here as flbD1. The flbC8 allele contains a point mutation that produces an H237Q change in the second putative zinc finger, consistent with a loss of function (46). The flbB allele, originally isolated as fluH1 (69), corresponds to a point mutation changing codon 57 into a stop codon (AAG to TAG) that results in a truncated peptide (this work). flbE58 has not been sequenced yet but very likely corresponds to a loss-of-function allele (77). ΔflbD strains in a veA+ background were obtained from a cross between strains TJAQ15 and FGSCA4. The presence of the veA+ allele was confirmed by PCR using genomic DNAs from selected progeny and the primers veA+ forward and veA+ reverse, with digestion of the PCR product with the BstXI enzyme (NEB, Ipswich, MA), as reported previously (24). All strains were grown at 37°C in glucose-supplemented minimal nitrate medium (26). Genetic manipulations, including diploid formation, were performed according to the methods of Pontecorvo et al. (53). Asexual developmental cultures and nutrient starvation were carried out as reported previously (3, 63). To induce sexual development in confluent plates, conidia from 5 days of growth were plated on top agar at 1 × 105 conidia per plate, incubated at 37°C for 24 h (0-h time point), and then sealed with masking tape. To induce sexual development in point-inoculated plates, spores were inoculated on water-agar plus supplements (no carbon or nitrogen source) and grown at 37°C for 5 days, 4 small agar cylinders of synthetic crossing medium for Neurospora crassa (6) were then placed symmetrically on the border of the colony, and the plate was sealed with masking tape and incubated at 37°C for 7 to 10 days.
Aspergillus nidulans strains used in this work
Disruption of flbD, flbD point mutations, and GFP tagging.Genomic DNA was used as the template to produce an flbD gene replacement construct by double-joint PCR (82). The 5′ flbD fragment was obtained with primers 5FORflbD and 5REV+TAILflbD (see Table 2 for primer sequences). The 3′ flbD fragment was amplified with primers 3FOR+TAILflbD and 3REVflbD. An Aspergillus fumigatus pyrG marker was amplified with primers pyrGforward and pyrGreverse, using plasmid pFNO3 (50) as the template. The three fragments were purified, mixed, and used in a fusion PCR with primers FORnestedflbD and RevnestedflbD. The final, 4,000-bp flbD-AfpyrG-flbD cassette was purified and used to transform A. nidulans strain CFL3 by conidium electroporation (59). Point mutations in C46 and R47 in the FlbD Myb domain were introduced by overlap extension PCR. A 5′ flbD fragment including the mutation was amplified with primers MutForflbD and RevflbD(C-D), RevflbD(C-S), RevflbD(C-A), or RevflbD(R-K). A 3′ flbD fragment including the mutation was amplified with primers MutRevflbD and ForflbD(C-D), ForflbD(C-S), ForflbD(C-A), or ForflbD(R-K). These fragments were combined in overlap extension reaction mixtures with primers ForflbDpENTER and RevflbDpENTERs/stop to clone these PCR fragments as reported previously (71), placing the flbD open reading frame (ORF) (wild-type or mutant alleles) under the control of the alcA promoter and tagging the FlbD C terminus with monomeric red fluorescent protein (mRFP). To obtain the flbDR47P allele, the flbD ORF was amplified with the primers ForflbDpENTER and RevflbDpENTERs/stop, using genomic DNA from the fluF1 mutant as the template. Plasmids obtained in this way were designated pFlbDwtENTR, pFlbDC46DENTR, pFlbDC46SENTR, pFlbDC46AENTR, pFlbDR47KENTR, and pFlbDR47PENTR. After confirming the flbD ORF sequence in each plasmid, the pFlbDENTR plasmids were used for recombination with the destination vector pMT-mRFP (71), using Gateway LR Clonase enzyme mix (Invitrogen, Carlsbad, CA) as indicated by the manufacturer. The resulting plasmids, palcA-flbD-mRFP plasmids (pJAQ9 to -14), were used to transform strain CJAQ25 by protoplast fusion. Three PCR products were used to generate an flbD C-terminal green fluorescent protein (GFP) wild-type or mutant construct according to the method of Yang et al. (79). First, a 5′ fragment upstream of the stop codon, including the entire flbD ORF, was amplified with primers flbDGSP1 and flbDGSP2 (for the wild-type construct) or with primers flbDGSP1b and RevflbDpENTERs/stop, using palcA-flbD-mRFP plasmids as the templates. Second, a 3′ flbD fragment was amplified with primers flbDGSP3 and flbDGSP4. Third, the GFP gene and the A. fumigatus pyrG marker were amplified with primers flbDGFP1 and flbDGFP2, using plasmid pFNO3 as the template (50). Purified fragments were mixed and used in a fusion PCR with primers flbDGSP4 and flbDGSP1 or flbDGSP1b. The 4,630-bp flbD-gfp-AfpyrG cassette was used to transform A. nidulans strain A1155 by conidium electroporation or protoplast fusion (59, 60, 80). Transformants with point mutations were confirmed by Southern blot analysis and DNA sequencing.
DNA primers used in this work
RNA extraction and Northern blot analysis.Samples were frozen in liquid nitrogen and stored at −70°C until use, at which time they were ground with a mortar and pestle under liquid nitrogen. Total RNA was isolated with TRIzol (Invitrogen, Carlsbad, CA) according to the manufacturer's instructions. Ten to 15 μg of RNA was separated in a 1% agarose gel containing formaldehyde, transferred to a Hybond N+ membrane, and hybridized using specific probes.
RESULTS
fluF1 is an flbD allele.A fluffy mutant isolated after treatment with 5-azacytidine defined the fluF1 gene, which was assigned to chromosome VIII (68). fluF is located in the standard genetic map of A. nidulans as the most centromere-proximal identified locus in the long arm of chromosome VIII (http://www.gla.ac.uk/ibls/molgen/aspergillus/viiimap.html). This location is based on its reported close linkage (4 centimorgans [cM]) with choC. However, in the same article, a linkage of 6.5 cM is reported for another marker, fwA1, which is not linked to choC (67) in the standard map. We confirmed by mitotic haploidization the location of fluF1 in chromosome VIII; however, we failed to detect any linkage with choC (identifiable as ANID_1376) (http://www.gla.ac.uk/ibls/molgen/aspergillus/viiicontigs.html; our unpublished observations) or with sonA (ANID_1379) (78), which is centromere proximal to choC. We thus investigated whether fluF1 is allelic with other mutations in chromosome VIII that result in a fluffy phenotype. To identify the affected gene, we carried out complementation tests on diploids with mutations in other genes located in chromosome VIII whose mutation results in a fluffy phenotype map (flbD [76] and tmpA [64]). Diploid strains obtained by crosses between fluF1 mutants and wild-type (Fig. 1A) or ΔtmpA (not shown) strains showed wild-type conidiation. In contrast, diploids obtained by crosses between fluF1 and flbD1 mutants maintained a fluffy phenotype, indicating that fluF1 and flbD1 are alleles (Fig. 1A). Standard genetic crosses between fluF1 and flbD1 strains failed to yield wild-type recombinants among 329 progeny analyzed (Fig. 1B), confirming the allelism of fluF1 and flbD1. We used fluF1 strain genomic DNA as the template to amplify the cognate flbD gene. Direct sequencing of two independent PCR products showed a single G-to-C substitution in codon 47 (CGG → CCG), which results in replacement of the highly conserved arginine 47 by proline (R47P) in the FlbD protein (Fig. 1D; see Fig. S1 in the supplemental material).
fluF1 is an flbD allele in which arginine 46 codon CGG is replaced by proline codon CCG. (A) A diploid between fluF1-1 and flbD1 fluffy gene mutants shows a fluffy mutant phenotype. Strains FGSC26 (WT; top panel), PW-1 (WT; middle panel), CAF2 (fluF1-1), G1059A3 (fluF1-2), TJW30.1 (flbD1), CJAQ10 (fluF1-1/flbD1), CJAQ18 (WT/flbD1), and CJAQ21 (WT/fluF1-2) were point inoculated onto supplemented minimal medium (MM) plates and incubated at 37°C for 3 days. (B) fluF1 (CJAQ8) and flbD1 (TJW30.1) strains show very little recombination in sexual crosses. Ascospores from a hybrid cleistothecium were inoculated onto supplemented MM plates containing yeast extract and incubated at 37°C for 3 days, and then the fluffy phenotype was scored. (C) FlbD, a transcription factor of the Myb domain family, contains two repeats with three alpha helices, indicated as H1 to H3. The second repeat contains the highly conserved amino acids cysteine (C46) and arginine (R47). (D) fluF1 mutation results in replacement of arginine 47 by proline (R47P). Genomic DNAs from wild-type (CLK43) and fluF1 mutant (G1059A3) strains were amplified by PCR and sequenced as shown in the pherogram. See Table 1 for full strain genotypes.
Comparison of different alleles identifies a new role for flbD during sexual development.To determine the phenotype of a null flbD mutant, we deleted the entire flbD coding region by using the deletion construct 5′flbD-AfpyrG-3′flbD (see Fig. S2A in the supplemental material), generated by double-joint PCR, and used it to transform electrocompetent conidia from an A. nidulans CFL3 strain. Eight fluffy PyrG+ transformants analyzed by Southern blotting had the correct gene deletion event (see Fig. S2A and B). When conidiation levels of wild-type, flbDR47P (fluF1), flbD1, and ΔflbD strains were compared, we found that although all strains carrying flbD mutant alleles were strikingly impaired in conidiation, this phenotype was more severe in ΔflbD strains (Fig. 2A). As flbD mRNA levels were not affected in the flbDR47P mutant (see Fig. S3), our results suggest that flbDR47P and flbD1 alleles produce proteins with partial function in the regulation of asexual sporulation.
flbD is required for peridium but not noxA-dependent ascospore development during sexual differentiation. (A) Strains CLK43 (WT), CJAQ24 (fluF1; referred to as flbDR47P), TJW30.1 (flbD1), and TJAQ15 (ΔflbD) were point inoculated onto supplemented MM plates and incubated at 37°C for 5 days. Total conidiospores were harvested and counted for each colony. Error bars indicate standard deviations for three independent experiments. (B) Schematic representation of early stages of cleistothecium development (upper left). Strains TJAQ15 (ΔflbD) and CJAQ24 (fluF1; referred as flbDR47P) were induced to undergo sexual development by use of a method adapted from an N. crassa method (see Materials and Methods), and samples were observed using a stereoscopic microscope (magnification, ×8) after 7 days of development (lower left) or processed for scanning electron microscopy after 8 days of development (right panels). Red aggregates of ascospores lacking the surrounding peridial tissue are indicated with black arrows. Scanning electron micrographs show a wild-type cleistothecium and ΔflbD ascospore aggregates next to some filamentous cells (hyphae). (C) Strains CAH4 (ΔnoxA) and CJAQ40 (ΔflbD ΔnoxA) were grown as described above. Images were taken after 8 days of sexual development by use of scanning electron microscopy. For scanning electron microscopy images, white bars indicate the magnification. PT, peridial tissue; AT, ascogenic tissue; As, ascospores; Cl, cleistothecia; H, Hülle cells; Pr, primordium. See Table 1 for full strain genotypes.
In performing sexual crosses between flbD1 and flbDR47P (fluF1) strains, we noticed that it was very difficult to detect and isolate cleistothecia, as putative fruiting bodies appeared very soft and wet. To follow sexual development more closely in isolated fruiting bodies, we successfully adapted a method developed for Neurospora crassa (6, 12). Under these conditions, we observed that all flbD mutants produced atypical fruiting structures (Fig. 2B), characterized by the presence of many Hülle cells and naked ascospores, which lacked the external tissue (peridium) that normally surrounds and protects the ascospores (9, 62). Ascospores produced in this way presented normal viability (not shown), and their production was denoted by the deposition of a reddish purple pigment on the plate surface (Fig. 2B). Having shown previously that the NADPH oxidase NoxA is required for the production of ROS at the peridial tissue and for cleistothecium development (35), we asked if the NoxA requirement could be bypassed in flbD mutants, which can develop ascospores in the absence of peridial tissue. We generated ΔflbD ΔnoxA double mutants and followed sexual development as described before. As shown in Fig. 2C, a ΔflbD ΔnoxA double mutant was arrested at a very early stage of development (initial cleistothecia) and, similarly to a ΔnoxA single mutant, was unable to produce any ascospores. These results demonstrate that peridium and ascospore development represents two separable developmental processes and that in addition to its role in asexual development, flbD is required for peridium formation during sexual development but dispensable for ascospore production, while the absence of NoxA results in a blockage of sexual development at a very early stage and prevents both developmental outcomes.
flbD, but not other fluffy genes, has a veA-independent role in sexual development.As indicated before, inactivation of the flbB, flbC, flbD, flbE, and tmpA genes results in fluffy phenotypes related to a delay in asexual development and less expression of the brlA gene (fluffy low brlA) (64, 77). Furthermore, FlbB and FlbD act together to regulate asexual sporulation as well as polarized growth (21). We therefore asked if these or other types of fluffy gene interactions were also required for peridium development or general sexual differentiation. In contrast to the ΔflbD mutant, flbB, flbC, flbE, and tmpA mutants were all able to differentiate normal fruiting bodies (Fig. 3), which produced normal and viable ascospores like those of a wild-type strain. These results indicate that FlbD's role in peridium development is specific and independent of the FlbD partner FlbB and the flbC, flbE, and tmpA genes.
flbD is unique among other fluffy genes with regard to its specific role in sexual development. Strains CLK43 (WT), TJAQ15 (ΔflbD), CJA16 (flbB), RJF009 (flbC), RJW120 (flbE), and TGS6 (ΔtmpA) were point inoculated onto supplemented agar plates and incubated at 37°C for 5 days. Sexual induction was carried out as indicated. Images were taken with a stereoscopic microscope. Magnification, ×8. Some heterogeneity in fruiting body size was observed in all strains. See Table 1 for full genotypes.
The A. nidulans veA gene encodes a 573-amino-acid protein required for cleistothecium formation (32). Because our laboratory strains contain a defective veA1 allele that results in more production of conidia and smaller numbers of cleistothecia (30), we also evaluated the flbD mutant's sexual phenotype in the presence of a wild-type veA allele. As shown in Fig. S4 in the supplemental material, a ΔflbD veA+ strain developed fruiting structures which also lacked the peridium and produced naked ascospores, indicating that FlbD's role in peridium development is independent of the veA gene.
FlbD Myb domain residue C46 plays critical roles in both asexual and sexual development.Many reports indicate that Myb transcription factors can be regulated by redox modification. This mechanism involves a cysteine conserved in all R1R2R3 and R2R3 Myb family members from plants, fungi, and animals (see Fig. S1 in the supplemental material). This cysteine can act as a redox sensor, and in fact, its oxidation impairs DNA binding in vitro (49). Adjacent to this cysteine in these Myb proteins is a highly conserved arginine residue whose positive charge affects cysteine's pKa and therefore its redox properties (11). In FlbD, these residues correspond to C46 and R47, and the fact that the R47P substitution results in a virtually complete lack of function (Fig. 4 and 5) suggests that this change might affect C46's redox properties. However, the role of C46 in FlbD's function has not been evaluated. To test this, we generated mutant alleles with mutations in C46 and R47. C46D and C46S changes were expected to impair DNA binding (22, 49), while a C46A change was expected to improve or not affect DNA binding (49). R47 was replaced by K, also a positively charged residue, with the expectation that this should not affect FlbD function, as in some Myb proteins from Entamoeba histolytica the residue next to C corresponds to K (45). We fused wild-type and mutant alleles to the reporter gene mRFP and placed the construct under the control of the inducible alcohol dehydrogenase gene (alcA) promoter (71). Plasmids were used to transform the ΔflbD strain CJAQ25, directing the integration to the argB locus (for details, see Materials and Methods). Integration into argB was confirmed via Southern blotting (see Fig. S5 in the supplemental material), and each flbD point mutation was confirmed by DNA sequencing using genomic DNA as the template (not shown).
Asexual development of ΔflbD mutants carrying different flbD alleles, encoding proteins with amino acid substitutions at C46 and R47, in trans. Plasmids pJAQ9 to -14 (directed to the argB locus), containing wild-type or mutated flbD alleles (R47P, R47K, C46D, C46S, and C46A) expressed from the inducible alcA promoter and tagged with mRFP at the C terminus, were used to transform strain CJAQ25 (ΔflbD). Strains TJAQ15 (ΔflbD), TJAQ20 (flbDwt), TJAQ24 (flbDR47P), TJAQ30 (flbDR47K), TJAQ35 (flbDC46D), TJAQ41 (flbDC46S), and TJAQ46 (flbDC46A) were point inoculated onto glucose-supplemented MM plates and incubated at 37°C for 5 days. The total number of conidiospores per colony area was calculated. Error bars indicate standard deviations for three independent experiments.
FlbD residues C46 and R47 in the Myb domain play critical roles in both asexual and sexual development. (A) Constructs containing wild-type or mutated (R47P, C46D, and C46A) flbD alleles, with the A. fumigatus pyrG gene as a selective marker and with C-terminal GFP tagging, were used to transform strain A1155 (nkuA). Strains CJAQ51 (WT), CJAQ52 (ΔflbD), TJAQ48 (flbD::gfp), TJAQ49 (flbDR47P), TJAQ50 (flbDC46D), and TJAQ51 (flbDC46A) were point inoculated onto glucose-supplemented MM plates and incubated at 37°C for 5 days. (B) The strains from panel A were point inoculated onto supplemented agar plates and incubated at 37°C for 5 days for sexual development. Images were taken by stereoscopic microscopy. Magnification, ×8. (C) Conidia from strains CJAQ51 (WT), TJAQ48 (flbD::gfp), and TJAQ51 (flbDC46A::gfp) were plated on top agar at 1 × 105 conidia/plate and incubated at 37°C. After 24 h, plates were sealed with Parafilm and samples were taken at the indicated times. The number of cleistothecia per square centimeter was calculated as reported previously (31). See Table 1 for full genotypes.
The results in Fig. 4A and B show that on solid medium, wild-type flbD fused to the mRFP allele complemented the conidiation defects of the ΔflbD strain, even in the presence of glucose on the plate, a condition in which alcA is known to be expressed weakly (43). However, the mRFP tag was not useful to localize FlbD::mRFP, as its signal was undetectable by fluorescence microscopy under growth, asexual or sexual differentiation, or even inducing (ethanol or threonine) conditions. Under the same conditions, the expression of the flbDR47P and flbDC46D alleles resulted in only a modest increase in conidiation. The flbDR47K allele showed improved conidiation in comparison to the flbDR47P allele, but without reaching wild-type levels. This indicates that a positive charge at position 47 is important for FlbD function in conidiation but that K is not functionally equivalent to R in FlbD. Strains carrying an flbDC46S allele showed about half of wild-type conidiation levels, similar to the level exhibited by the flbDR47K strain. Notably, the flbDC46A allele resulted in conidiation levels that were indistinguishable from those found in the wild type, consistent with the expectation that a C46A change should improve or not affect FlbD DNA binding (49). These results highlight the regulatory role of C46 in FlbD during conidiation.
To test the functionality of these flbD alleles during sexual differentiation, we evaluated the strain phenotypes by using a standard assay for inducing sexual development in which conidiation occurs first and is then followed by fruiting body development. The results in Fig. S6 in the supplemental material show that under these conditions, alcA-driven expression of wild-type and flbDC46A alleles was also enough for conidiation (see conidiophores and yellow conidia) but not enough to rescue normal sexual development (note that water droplets were formed around clumps of naked ascospores).
As an alternative approach to evaluate the effects of point mutations within the FlbD Myb domain on sexual development, we generated a new set of constructs to express wild-type and mutant alleles fused to GFP from the endogenous flbD promoter. These constructs were used to transform A. nidulans strain A1155 (nkuA), and integration was directed to the flbD locus. Each point mutation was confirmed by DNA sequence analysis of the entire flbD ORF (data not shown). The results in Fig. 5A show that strains containing wild-type and flbDC46A alleles showed wild-type conidiation, indicating that the GFP tag did not affect FlbD's function under these conditions. However, the GFP fluorescence signal was not detectable during growth or asexual or sexual differentiation, likely due to low FlbD protein levels in vivo. flbDC46D and flbDR47P strains showed delayed conidiation, again confirming the results in Fig. 4. Regarding sexual differentiation, flbDC46D and flbDR47P strains produced only naked ascospores without a peridium (Fig. 5B). In contrast, the flbDC46A strain was able to differentiate fully developed fruiting bodies, which produced viable ascospores just like those of the wild-type strain, also showing that the GFP tag did not interfere with FlbD's function in sexual development. Moreover, the flbDC46A strain produced larger numbers of cleistothecia than the wild-type strain, although the process occurred within the wild-type time frame (Fig. 5C). Our result indicate that amino acids C46 and R47 play an essential role in FlbD functions in both asexual and sexual differentiation and that for these roles, C46 can be replaced by A, which presumably does not affect FlbD binding to DNA.
We have previously shown that carbon starvation during liquid culture induces the differentiation of minimal conidiophores, while nitrogen starvation induces the formation of fully differentiated conidiophores (63). We found that under these conditions, fluffy mutants such as tmpA, flbB, flbC, flbD, and flbE mutants conidiated well under carbon starvation conditions but failed to conidiate under nitrogen starvation conditions (not shown). Having shown that the flbDC46A allele is fully functional in asexual and sexual differentiation, we asked whether it was also functional for nitrogen starvation-induced conidiation. The results in Fig. 6 show that under nitrogen starvation conditions, the wild-type and flbD::gfp strains formed fully developed conidiophores, in sharp contrast to the ΔflbD and flbDC46A mutants, which failed to differentiate any conidiophore structures. These results clearly indicate that the C46A mutant cannot replace wild-type FlbD regulation of conidiation induced by nitrogen starvation and that FlbD plays multiple roles during A. nidulans development.
flbDC46A strains are impaired in conidiation induced by nitrogen starvation. Mycelia from strains CJAQ51 (wild type), CJAQ52 (ΔflbD), TJAQ48 (flbD::gfp), and TJAQ51 (flbDC46A::gfp) grown for 18 h in liquid culture were shifted to the same medium lacking nitrate (−N). After 15 h, samples were fixed and processed for scanning electron microscopy. White bars indicate the magnification.
DISCUSSION
In this report, we have shown that fluF is an allele of the flbD gene, whose only known role was to activate brlA, a gene essential for asexual sporulation. Indeed, FlbD is a Myb TF whose expression depends on the bZIP TF FlbB, and both FlbD and FlbB are jointly required to activate brlA expression (21). In this study, we found that FlbD is also essential for the differentiation of the peridium, a tissue that constitutes the fruiting body external wall protecting the sexual spores. Furthermore, we showed that this role is specific for FlbD, as fluffy genes flbB, flbC, flbE, and tmpA are not required for peridium development. This is particularly important in the case of FlbB, as it indicates that in contrast to what occurs during conidiation, FlbB is not necessary for flbD expression during sexual differentiation, and that FlbD does not require interaction with FlbB to carry out its functions in peridial development.
Recently, it was found that when fluffy mutants affected in flbC or flbE were induced to undergo sexual development, they showed increased cleistothecium formation and decreased conidiation, suggesting that these genes repress sexual development and maintain the balance between asexual and sexual differentiation (33, 34). Whether or not this relates to FlbD's role in sexual development requires further investigation.
In the plant pathogen Gibberella zeae, the inactivation of MTY1, a different Myb transcription factor highly conserved in the ascomycetes, results in decreased female fertility and perithecium production and, in some cases, the production of immature ascospores (41). This suggests that different Myb transcription factors might be involved in different aspects of fungal sexual development. In plants, members of the Myb family regulate flavonoid biosynthesis, cell fate and identity, the cell cycle, and responses to biotic and abiotic stresses. In these roles, Myb TFs have been described as coactivators or positive regulators, sometimes carrying out partially redundant functions (15, 19, 37).
Myb TFs have been described as targets for redox regulation through reduction/oxidation of the cysteine C46 conserved in FlbD and in many other R1R2R3 and R2R3 family members from plants, fungi, and animals. In fact, the oxidation of this cysteine impairs DNA binding in vitro. Our results show that in FlbD, C46S and C46D substitutions impair function during asexual and sexual development. In contrast, a C46A substitution resulted in wild-type conidiation and fully developed fruiting bodies. Presumably, FlbDC46A can bind its DNA targets constitutively or with wild-type affinity (Fig. 4A and B and 5), as the same replacement in the Myb TF Bas1p resulted in a functional TF in vitro and in vivo (22, 49, 52). Since flbB is necessary to express flbD during conidiation but not during sexual development, putative constitutive DNA binding by FlbDC46A would not interfere with FlbD's normal function during conidiation but could explain the larger numbers of fully differentiated fruiting bodies than those of a wild-type strain.
ROS produced by the NADPH oxidase NoxA are required for fruiting body development in A. nidulans and other fungi (4, 8, 35, 42). ΔnoxA and ΔflbD ΔnoxA mutants were arrested at a similar early stage of development (cleistothecia initial) (Fig. 2C). This indicates that ROS are required for both peridium and ascospore development and that these stages are separable differentiation processes, with FlbD playing a role only in peridium formation. Although flbD mRNA levels were very low during sexual development, slightly increased levels were detected at the time of cleistothecium production (not shown). This and the fact that alcA-driven expression of flbD was enough for normal conidiation but not for peridium differentiation under conditions of drastically reduced expression (see Fig. S6 in the supplemental material) suggest that high and/or localized flbD mRNA levels are needed to develop the peridial tissue. Further research is needed to determine if localized NoxA-generated ROS can regulate FlbD function during sexual development.
Besides air contact, nutritional stress such as carbon or nitrogen starvation induces conidiation (63, 64). We have found that when fluffy mutants affected in the flbB, flbC, flbD, flbE, or tmpA gene (64) are starved for nitrogen, they fail to fully induce the brlA gene and to conidiate. Surprisingly, the flbDC46A allele, which was fully functional in sexual differentiation and conidiation induced by an air interface, was not able to induce conidiation under nitrogen starvation conditions in submerged cultures. Under these conditions, FlbD might perform additional functions in nitrogen signaling or utilization. In Schizosaccharomyces pombe, the Myb-type DNA binding protein Reb1 regulates G1 cell cycle arrest and sexual differentiation in response to nitrogen starvation (55), while in the unicellular red alga Cyanidioschyzon merolae CmMYB1, an R2R3-type Myb TF activates expression of key nitrogen assimilation genes in response to nitrogen status (28).
ACKNOWLEDGMENTS
This work was funded by grants CB-2005-01-49667 and 153256 from CONACYT, by DFG-CONACYT Germany-Mexico collaboration grant 75306 from CONACYT, and by grant IN209211-2 from PAPIIT-UNAM. Jenny Arratia-Quijada was supported by a scholarship from CONACyT and DGEP-UNAM.
We thank Mercedes Tamame for helpful discussions and for the original fluF1 strain, Steve Osmani for a ts allele of sonA, and Adriana Hernández for a ΔnoxA strain. We also thank the Molecular Biology and Microscopy Units of IFC-UNAM for technical support.
FOOTNOTES
- Received 27 March 2012.
- Accepted 6 July 2012.
- Accepted manuscript posted online 13 July 2012.
Supplemental material for this article may be found at http://dx.doi.org/10.1128/EC.00101-12.
- Copyright © 2012, American Society for Microbiology. All Rights Reserved.
REFERENCES
- 1.↵
- 2.↵
- 3.↵
- 4.↵
- 5.↵
- 6.↵
- 7.↵
- 8.↵
- 9.↵
- 10.↵
- 11.↵
- 12.↵
- 13.↵
- 14.↵
- 15.↵
- 16.↵
- 17.↵
- 18.↵
- 19.↵
- 20.↵
- 21.↵
- 22.↵
- 23.↵
- 24.↵
- 25.↵
- 26.↵
- 27.↵
- 28.↵
- 29.↵
- 30.↵
- 31.↵
- 32.↵
- 33.↵
- 34.↵
- 35.↵
- 36.↵
- 37.↵
- 38.↵
- 39.↵
- 40.↵
- 41.↵
- 42.↵
- 43.↵
- 44.↵
- 45.↵
- 46.↵
- 47.↵
- 48.↵
- 49.↵
- 50.↵
- 51.↵
- 52.↵
- 53.↵
- 54.↵
- 55.↵
- 56.↵
- 57.↵
- 58.↵
- 59.↵
- 60.↵
- 61.↵
- 62.↵
- 63.↵
- 64.↵
- 65.↵
- 66.↵
- 67.↵
- 68.↵
- 69.↵
- 70.↵
- 71.↵
- 72.↵
- 73.↵
- 74.↵
- 75.↵
- 76.↵
- 77.↵
- 78.↵
- 79.↵
- 80.↵
- 81.↵
- 82.↵
- 83.↵
