Neurospora crassa is a heterothallic filamentous fungus with two mating types, mat a and mat A. Its mating involves differentiation of female reproductive structures (protoperithecia) and chemotropic growth of female-specific hyphae (trichogynes) towards a cell of the opposite mating type in a pheromone-mediated process. In this study, we characterize the pre-1 gene, encoding a predicted G-protein-coupled receptor with sequence similarity to fungal pheromone receptors. pre-1 is most highly expressed in mat A strains under mating conditions, but low levels can also be detected in mat a strains. Analysis of pre-1 deletion mutants showed that loss of pre-1 does not greatly affect vegetative growth, heterokaryon formation or male fertility in either mating type. Protoperithecia from Δpre-1 mat A mutants do not undergo fertilization; this defect largely stems from an inability of their trichogynes to recognize and fuse with mat a cells. Previous work has demonstrated that the Gα subunit, GNA-1, and the Gβ protein, GNB-1, are essential for female fertility in N. crassa. Trichogynes of Δgna-1 and Δgnb-1 mutants displayed severe defects in growth towards and fusion with male cells, similar to that of Δpre-1 mat A strains. However, the female sterility defect of the Δpre-1 mat A mutant could not be complemented by constitutive activation of gna-1, suggesting additional layers of regulation. We propose that PRE-1 is a pheromone receptor coupled to GNA-1 that is essential for the mating of mat A strains as females, consistent with a role in launching the pheromone response pathway in N. crassa.
If you can't find a tool you're looking for, please click the link at the top of the page to "Go to old article view". Alternatively, view our Knowledge Base articles for additional help. Your feedback is important to us, so please let us know if you have comments or ideas for improvement.
In heterothallic ascomycete fungi, mating occurs only between haploid cells of opposite mating type and is initiated by cell-specific pheromone and receptor combinations (Bölker and Kahmann, 1993; Kothe, 1999). The mating process has been studied extensively in the yeast Saccharomyces cerevisiae (Kurjan, 1993). In this organism, small diffusible peptide pheromones (α-factor and a-factor) are bound by their cognate seven-transmembrane helix receptors (Ste2p and Ste3p, respectively). These receptors are located on the cell surface and are coupled to a heterotrimeric G-protein complex consisting of Gα (Gpa1p) bound to GDP and Gβγ (Ste4p and Ste18p). Pheromone binding to the receptor causes exchange of GDP for GTP on Gα, which in turn releases Gβγ to propagate the signal and activate a mitogen-activated protein kinase (MAPK) signal transduction pathway (Ste11p, Ste7p, Fus3p). The MAPK pathway ultimately activates the transcription factor Ste12p, which promotes the expression of over 200 genes to mediate various mating-specific outputs (Roberts et al., 2000), including cell cycle arrest in G1, polarized morphogenesis, agglutination, cell fusion, karyogamy and adaptation to the pheromone signal (Sprague, 1991; Banuett, 1998; Dohlman, 2002).
In U. maydis, pheromones and receptors control conjugation tube formation in response to the presence of a mating partner as well as subsequent cell fusion and maintenance of the resulting dikaryon (Spellig et al., 1994; Kronstad and Staben, 1997). In the studied mushroom fungi, pheromones and pheromone receptors are not required for the initial fusion between mating partners, but are necessary for later events in the mating process, such as nuclear migration and clamp cell fusion, to form a dikaryotic mycelium. In contrast to yeasts, in which haploid cell fusion is followed immediately by nuclear fusion, the dikaryophase is maintained indefinitely in mushroom fungi until given proper environmental conditions, such as nutrient, light and temperature (Casselton and Olesnicky, 1998; Brown and Casselton, 2001). Under proper conditions, the dikaryotic mycelium develops into the multicellular fruit body or mushroom. Nuclear fusion eventually occurs in this structure and is immediately followed by meiosis and sporulation.
Neurospora crassa, a heterothallic filamentous fungus with two mating types, mat a and mat A, undergoes morphologically complex sexual differentiation, forming specialized cells for mating (Raju, 1992). Initiation of the sexual cycle requires nitrogen starvation, light and low temperature, and involves the formation of the multicellular protoperithecium (a female reproductive structure; Perkins and Barry, 1977). Specialized hyphae (trichogynes) emanating from the protoperithecium exhibit positive chemotropic growth towards male gametes (vegetative cells of the opposite mating type; Bistis, 1981). Subsequent to contact, the trichogyne recruits a single fertilizing nucleus from the male gamete to the basal body of the protoperithecium. Similar to basidiomycetes, fertilization is not immediately followed by nuclear fusion in N. crassa. Instead, the resulting dikaryon undergoes a developmental programme in which nuclei of opposite mating types recognize each other, migrate in pairs to the developing hook-shaped ascogenous hyphae (croziers or young asci) and divide mitotically in synchrony. Inside each developing ascus, the male and female nuclei fuse, and the diploid zygote nucleus immediately undergoes meiosis. Like mushroom fungi, the two meiotic divisions are followed by a post-meiotic mitosis. The resulting eight nuclei are sequestered into eight linearly ordered homokaryotic, haploid ascospores in an ascus. Numerous asci are enclosed in a single fruiting body or perithecium. Ascospores are ejected through a pore (ostiole) at the tip of the perithecium after maturation.
Previous work has implicated diffusible pheromones in the directional growth of N. crassa trichogynes (Bistis, 1981; 1983). Recently, two pheromone precursor genes, mfa-1 and ccg-4, have been identified in N. crassa (Bobrowicz et al., 2002; Kim et al., 2002). Study of mfa-1 null strains showed that the encoded peptide is essential for the attraction of trichogynes and the male fertility of mat a strains (Kim et al., 2002), and ccg-4 is hypothesized to provide the same function for mat A strains (Bobrowicz et al., 2002). The finding that N. crassa possesses pheromone precursor genes implies the presence of their cognate receptor genes.
In this study, a pheromone receptor gene from N. crassa, pre-1, was mutated and phenotypes analysed. The results indicate that the pre-1 gene encodes a G-protein-coupled receptor (GPCR) containing seven predicted transmembrane helices, and that the PRE-1 protein is essential for the ability of mat A females to recognize and fuse with mat a male cells. This mating type-specific female sterility defect is consistent with a role for PRE-1 as a receptor in the pheromone response signal transduction pathway in N. crassa.
Identification of the N. crassa pre-1 gene
A putative pheromone receptor gene of N. crassa, pre-1, was identified by performing blast searches (Altschul et al., 1997) against the N. crassa genome database at the Whitehead Institute Center for Genome Research (WICGR; http://www-genome.wi.mit.edu/annotation/fungi/neurospora/) using amino acid sequences of several cloned fungal pheromone receptors, including C. cinereus Rcb3 (AF186385), Pneumocystis carinii Ste3b (AAG38548), Filobasidiella neoformans CPRa1p (AAK31936) and U. maydis Pra1 (U37795). A putative open reading frame (ORF), NCU00138.1, was found as a match to these proteins with scores of e-04 or less. NCU00138.1 belongs to the class of seven transmembrane G-protein-coupled receptors most similar to the Ste3p pheromone receptor from S. cerevisiae (reviewed by Kurjan, 1993), but contains an unusually extended C-terminal tail. There were no other predicted ORFs similar to NCU00138.1 in the N. crassa genome sequence, although a single predicted protein similar to the Ste2p pheromone receptor from S. cerevisiae is present in the genome (data not shown; see also Pöggeler and Kück, 2001). Automated gene annotation at the WICGR predicted the pre-1 ORF to be 2307 bp long (678 amino acids) with two introns of 194 and 76 bp.
In a recent study, Pöggeler and Kück (2001), performing Northern and reverse transcription-polymerase chain reaction (RT-PCR) analyses (based on the WICGR sequence), reported the pre-1 transcript as 1.4 kb and the ORF as much shorter than that included in NCU00138.1 (1265 bp, 402 amino acids long and lacking the extended C-terminal tail), with only one intron of 56 bp. In view of this discrepancy, we performed RT-PCR analysis of N. crassa mRNA using primers flanking each of the introns predicted for NCU00138.1 (see Experimental procedures and Table 1). Our results concur with those of Pöggeler and Kück (2001) regarding the pre-1 ORF structure (one intron and shorter ORF; data not shown). However, in contrast to the 1.4 kb transcript reported in their study, our results indicate that the pre-1 transcript is ≈ 3 kb (see below and Fig. 1C).
Table 1. Oligonucleotides used in this study.
pre-1 is preferentially expressed in mat A strains
Mating in N. crassa involves morphologically distinct male and female structures (the conidium and the trichogyne respectively). As pheromone gene expression is regulated developmentally and by mating type (Bobrowicz et al., 2002; Kim et al., 2002), we investigated the pattern of pre-1 expression during asexual and sexual development in both mating types. Northern analysis showed that the pre-1 gene is expressed as an ≈ 3 kb transcript, and the results from RT-PCR indicate that it contains a long 3′ untranslated region (Fig. 1C; data not shown). Consistent with a role as the mat A-specific pheromone receptor, pre-1 levels were high in sexual tissue (protoperithecia) of wild-type strain 74A. As N. crassa is not self-fertile, it was surprising that pre-1 message could also be detected in protoperithecia of wild-type strain 74a when analysing samples containing 50 µg of total RNA (Fig. 2A). However, the pre-1 mRNA was barely detected in 74A using samples containing 20 µg of total RNA (Fig. 2B), and densitometric analysis of multiple blots containing either amount of total RNA showed that levels of pre-1 in 74a were more than 100-fold lower than those in 74A (data not shown). These observations, in combination with the heterothallic nature of N. crassa, suggest that either the level of pre-1 in mat a cells is not high enough to promote self-mating and/or that pre-1 mRNA levels do not correlate with PRE-1 protein levels. Results from analysis of Δpre-1 phenotypes are consistent with these hypotheses (see below).
In strain 74A, pre-1 expression was much higher in sexual tissue than in fresh conidia or germlings (Fig. 2, lanes 1 and 2). Little pre-1 mRNA was detected in exponentially growing or starved vegetative tissues (Fig. 2, lanes 3–5). As the expression of mfa-1, the mat a pheromone precursor gene, increases greatly as perithecia develop (Kim et al., 2002), pre-1 levels were also examined in perithecia 3, 6 and 9 days after fertilization, representing early, middle and late perithecial development respectively. In contrast to mfa-1, pre-1 expression did not increase after fertilization and remained at similar levels to that in protoperithecia as perithecia developed (Fig. 2, 3- and 6-day-old perithecia).
The observation that pre-1 levels were much greater in mat A wild-type strains than in the corresponding mat a wild type prompted us to examine the regulation of pre-1 by the mat A mating locus in more detail. The mat A locus contains three predicted ORFs, mat A-1, mat A-2 and mat A-3 (Glass et al., 1990; Ferreira et al., 1996). The MAT A-1 polypeptide is similar to the S. cerevisiae MATα1 mating-type protein and has been shown to confer mating-type specificity in N. crassa (Glass et al., 1990). The Am44 (Griffiths, 1982) strain contains a frameshift mutation at codon 54 in the mat A-1 gene (Glass et al., 1990). Although it forms conidia and protoperithecia, the Am44 mutant is greatly impaired in both male and female fertility (Griffiths, 1982) and does not produce the ccg-4 pheromone (Bistis, 1983; Bobrowicz et al., 2002). We compared levels of pre-1 mRNA in wild type and the Am44 sterile mutant using Northern analysis (Fig. 1C). No expression of pre-1 was observed in the Am44 strain (Fig. 1C, lane 2), consistent with regulation of pre-1 expression by the mat A locus in N. crassa.
Deletion of pre-1 does not affect vegetative growth and development
In order to determine roles for pre-1 during the asexual and sexual cycles in N. crassa, we generated Δpre-1 strains by targeted integration of a construct in which the entire amino acid coding region of pre-1 was replaced with a bacterial hygromycin B resistance gene (hph; Fig. 1A). Genomic DNA from hygromycin-resistant transformants was subjected to Southern analysis after digestion with SalI. Heterokaryotic strains containing both wild-type (corresponding to a 4.2 kb hybridizing fragment) and Δpre-1 (5.4 kb hybridizing fragment) nuclei were identified (Fig. 1B). Homokaryotic Δpre-1 strains were obtained by crossing heterokaryotic primary transformants to the wild type, with selection for growth of progeny on hygromycin B. Replacement of the gene in all nuclei was confirmed using Southern analysis (loss of 4.2 kb fragment and appearance of 5.4 kb fragment; Fig. 1B). Northern analysis of Δpre-1 mat A strains showed that they lack the pre-1 mRNA (Fig. 1C). Identical results were seen after Northern analysis of mRNA from Δpre-1 mat a strains (data not shown).
To verify that any phenotypes observed in Δpre-1 strains resulted from the deletion of the gene, the wild-type pre-1+ genomic fragment was targeted to the his-3 locus in a Δpre-1 strain background (strain (16)A his-3; Table 2). This rescued strain was selected by conferral of histidine prototrophy. Homokaryons were purified from initial heterokaryotic strains using microconidial isolation (Ebbole and Sachs, 1990). Single pre-1+ integration events in Δpre-1 + pre-1+ strains were confirmed by Southern analysis. The Δpre-1 recipient strain yielded a 9.5 kb hybridizing HindIII fragment, while integration of the wild-type pre-1+ yielded a 4.8 kb hybridizing fragment, because of the presence of a HindIII site in pre-1 (data not shown). The rescued strains exhibited wild-type phenotypes for all functions tested (data not shown).
Table 2. N. crassa strains.
FGSC, Fungal Genetics Stock Center, Kansas City, MO, USA.
Deletion of pre-1 did not result in any apparent vegetative defects in Δpre-1 strains of either mating type. Δpre-1 strains were similar to wild type when examined for apical extension rates on solid VM and sensitivity to elevated levels of NaCl, KCl and sorbitol (data not shown). Loss of pre-1 did not affect aspects of asexual sporulation, including aerial hyphae formation and conidiation.
An important vegetative trait of N. crassa is its ability to form a heterokaryon through hyphal fusion between two different strains (reviewed by Glass and Kaneko, 2003). Heterokaryon formation, like mating, is regulated by mating type in N. crassa (Griffiths and DeLange, 1978; Griffiths, 1982). However, in contrast to mating, both strains must be of the same mating type in order to form a heterokaryon. Because of the observed effect of mating type on pre-1 gene expression, we reasoned that Δpre-1 strains might be affected in their ability to form heterokaryons. Δpre-1 strains were made, each containing one of two different auxotrophic markers, his-3 or pan-2. Corresponding pre-1+ strains were used as controls. Mixtures containing conidia from strains with different auxotrophic markers were tested for their ability to form a colony when plated on medium lacking histidine and pantothenate. Δpre-1 his-3 and Δpre-1 pan-2 strains formed heterokaryons and conidiated as well as corresponding pre-1+his-3 and pre-1+pan-2 strains (data not shown). For both Δpre-1 and pre-1+ strains, mixtures containing pairs of the same mating type grew approximately five times faster than those with different mating types.
pre-1 is essential for chemotropic growth of mat A trichogynes towards and fusion with mat a cells
The fertility of an N. crassa strain is often tested by spotting conidia of one mating type directly on to cultures of another mating type cultured to produce protoperithecia and then assessing production of perithecia and ascospores (Davis and DeSerres, 1970). However, in order to follow early steps in the mating process, conidia of opposite mating type are applied at a distance from a protoperithecium. Microscopic observation is used to track growth of trichogyne tips towards, coiling around and fusion with conidia (Fig. 4A and B; Bistis, 1981), in a process mediated by diffusible pheromones (Bistis, 1983; Kim et al., 2002). If conidia of the same mating type or sterile mutants, such as am1, are used as females, trichogynes fail to display the above interactions (Bistis, 1981).
The expression pattern of pre-1, with high levels in conidia (males) and female structures (protoperithecia and perithecia), suggested a role in fertility in N. crassa. Therefore, Δpre-1 strains were examined for defects at various stages of sexual development. A function for pre-1 in female fertility was explored by culturing Δpre-1 strains on low-nitrogen SCM medium. Δpre-1 strains exhibited normal apical extension rates (data not shown) and formed fully differentiated protoperithecia with trichogynes on SCM (Fig. 3), although more protoperithecia were observed in Δpre-1 mat A mutants than in wild-type or Δpre-1 mat a strains (Fig. 3). At present, we do not know the reason for the increased number of protoperithecia in mat A versus mat a Δpre-1 mutants.
In crosses to wild-type males, Δpre-1 mat A strains were completely unable to develop perithecia (Fig. 3), while Δpre-1 mat a strains underwent normal sexual development (Fig. 3). Protoperithecia of Δpre-1 mat A strains did not show any indication of fertilization, such as enlargement or darkening, after application of conidia. When examined microscopically, trichogynes of Δpre-1 mat A strains did not display the directional growth seen in wild type (Fig. 4D and E). Similar to observations from crosses of two same mating-type wild-type strains, the tips of Δpre-1 mat A trichogynes grew in random directions without bending or coiling (Fig. 4D) and did not fuse with 74a conidia, even during incidental direct contact (Fig. 4E). In contrast, trichogynes of Δpre-1 mat a strains showed normal chemotropic attraction to 74A conidia (Fig. 4C). This result is consistent with a mating type-specific defect in Δpre-1 mat A strains. Reintroduction of pre-1+in trans into a Δpre-1 mat A strain restored female fertility (Fig. 3) and directional growth of trichogynes (Fig. 4F).
In contrast to the results for female fertility, both mating types of Δpre-1 strains were as fertile as males. When conidia of Δpre-1 strains were spotted directly onto a wild-type female of opposite mating type, perithecia developed and produced viable ascospores, similar to wild-type to wild-type crosses (data not shown). Homozygous crosses between Δpre-1 strains were able to produce ascospores if Δpre-1 mat A strains were used as the male parent to fertilize Δpre-1 mat a females (data not shown). Presumably, Δpre-1 mat a strains are female-fertile because they contain the mat a pheromone receptor, PRE-2 (Pöggeler and Kück, 2001), which recognizes the mat A pheromone, CCG-4 (Bobrowicz et al., 2002).
Pheromone precursor genes in S. cerevisiae are transcriptionally regulated by the pheromone response pathway (Roberts et al., 2000). This prompted us to examine expression of the two pheromone precursor genes, ccg-4 (the mat A pheromone precursor gene) and mfa-1 (the mat a pheromone precursor gene) in sexually differentiated tissues of Δpre-1 strains of corresponding mating type. Compared with wild type, expression of ccg-4 and mfa-1 was significantly reduced in Δpre-1 mat A(Fig. 5A) and mat a strains (Fig. 5B) respectively. Thus, loss of the PRE-1 pheromone receptor affects the expression of both peptide pheromones in the appropriate mating type.
The mating responses of trichogynes require G protein subunits
Heterotrimeric G proteins play essential roles during sexual development in fungi, where they act as signal transducers that couple cell surface receptors to cytoplasmic effectors. In S. cerevisiae, Gpa1p acts as a negative regulator of the active Gβγ heterodimer during pheromone signalling (for a recent review, see Dohlman, 2002). The Gα Gpa1p from S. pombe works in concert with the small GTP-binding protein Ras1 as positive regulators of pheromone signal transduction (reviewed by Davey, 1998). In the fungal pathogens Magnaporthe grisea, U. maydis and C. parasitica, the Gα subunits, MAGB, Gpa3 and Cpg-1, are required for female fertility (Gao and Nuss, 1996; Liu and Dean, 1997; Regenfelder et al., 1997).
In N. crassa, it has been hypothesized that GNA-1 and, in a minor capacity, GNA-2 activate the pheromone response pathway (Ivey et al., 1996; Baasiri et al., 1997; Kays et al., 2000) and that GNB-1 regulates female fertility indirectly by modulating GNA-1/GNA-2 protein levels (Yang et al., 2002). To examine possible involvement of the G-protein subunits in the mating response pathway in N. crassa, the trichogyne responses of Δgna-1, Δgna-1 Δgna-2, Δgna-2, Δgna-3, Δgna-2 Δgna-3 and Δgnb-1 strains (Table 3) were monitored.
Table 3. Summary of mating responses.
Trichogyne chemotropism and growth
Wild type Δpre-1 mat a Δgna-2 Δgna-3
Yes; Trichogynes fully differentiated
Yes; Directional growth towards and tight coiling around male cell
Δpre-1 mat A Δgna-1Δgna-2 Δgnb-1
Yes; Trichogynes fully differentiated
No; Both directional growth and coiling absent
Yes; Trichogynes fully differentiated
No; Slight bending observed on occasion
Yes; Trichogynes fully differentiated
Yes; Loose coiling observed
Δpre-1 mat A gna-1*
Yes; Short trichogynes
In Δgna-1 mutants, trichogynes did not display the directional growth seen in wild type; instead, they grew in random directions and did not fuse with male cells even when in direct contact (Fig. 4G; Table 3). On occasion, a trichogyne was observed to bend around a male cell in close proximity; however, the trichogyne was unable to circle around or fuse with the male cell, and there was no subsequent enlargement or darkening of protoperithecia (data not shown). Trichogynes of a Δgna-1 Δgna-2 strain displayed defects similar to those seen in Δpre-1 mat A strains (Fig. 4H; Table 3), with growth in random directions and no occasional bending around conidia, as observed in a Δgna-1 strains. These data suggest that GNA-1, in concert with a pheromone receptor, is essential for directional growth and fusion of trichogynes during the pheromone response, and that GNA-2 plays a more minor, compensatory role in this process.
Trichogynes of Δgna-2 and Δgna-3 strains grew specifically towards, coiled around and fused with male cells as well as wild type (Fig. 4I and J respectively; Table 3). Δgna-3 trichogynes exhibited branching while coiling around conidia. Trichogynes of a Δgna-2 Δgna-3 double mutant grew towards, coiled around and fused with male cells. However, their pattern of coiling around male cells was slightly different from that of wild type. Δgna-2 Δgna-3 mutants formed loose trichogyne lariats around conidia and, on many occasions, the sides (instead of tips) of trichogynes initiated the encirclement (Fig. 4K; Table 3). Like Δgna-3 mutants, Δgna-2 Δgna-3 strain trichogynes formed branched structures during encirclement of conidia. These results suggest that GNA-3 and GNA-2 are involved in the mating process, possibly in localizing the PRE-1 protein or other signalling components at the tips of trichogynes. Finally, trichogynes of the Δgnb-1 strain lacking the Gβ subunit displayed severe trichogyne defects similar to those of Δpre-1 mat A strains (Fig. 4L; Table 3).
To examine whether deletion of pre-1 affects G-protein levels during sexual development, we performed Western analysis using antisera to N. crassa Gα and Gβ proteins (Ivey et al., 1996; Baasiri et al., 1997; Kays et al., 2000; Yang et al., 2002). The results show that GNA-1, GNA-2, GNA-3 and GNB-1 proteins are present at the same levels in Δ pre-1 mat A strains as in mat A or mat a wild-type strains (Fig. 6). Thus, the defects in trichogyne growth and coiling of Δpre-1 mutants cannot be explained by effects on levels of G-protein subunits, including GNA-1 and GNB-1.
Constitutive activation of GNA-1 alone does not restore female fertility in a Δpre-1 background
Studies have shown that mutational activation of GNA-3-related Gα subunits of fungi can suppress the defects of Gα-coupled receptor null mutants (Xue et al., 1998; Welton and Hoffman, 2000). For example, constitutive activation of S. cerevisiae Gpa2p through the use of a predicted GTPase-deficient allele (R273A) completely suppresses the pseudohyphal defect of mutants lacking its coupled receptor, Gpr1p (Xue et al., 1998). In S. pombe, an activated Gpa2 protein (R176H) also fully reverses the phenotypes resulting from loss of its coupled receptor, Git3, including delayed germination and starvation-independent mating (Welton and Hoffman, 2000).
No studies have reported the consequences of constitutive activation of Gαi-related fungal Gα subunits in the absence of their coupled receptors. As a group, these proteins are required for normal sexual fertility in diverse ascomycete filamentous fungi (Bölker, 1996). As Δgna-1 and Δpre-1 mutants exhibit similar sexual defects, we investigated whether activation of GNA-1 could reinstate female fertility in the absence of pheromone signalling in N. crassa. This was done by targeting a constitutively activated presumed GTPase-deficient allele of the Gα subunit gna-1 (gna-1Q204L) to the his-3 locus of a Δpre-1 mat A strain (Table 2). The genotype of transformants was verified using Southern analysis, in which integration of the gna-1Q204L allele resulted in a 6.2 kb hybridizing fragment as a result of the presence of HindIII sites in gna-1 (Yang and Borkovich, 1999). Δpre-1 mat A strains carrying the dominant-activated gna-1 allele showed a great reduction in protoperithecial formation (Fig. 3), similar to that described previously for Δgna-1 + gna-1Q204L strains (Fig. 3; Yang and Borkovich, 1999). However, the Δpre-1+ gna-1Q204L strain was female-sterile, indicating that activation of the GNA-1 protein alone cannot restore female fertility in a strain lacking the appropriate pheromone receptor.
The pre-1 gene of N. crassa encodes a pheromone receptor similar to the hydrophobic peptide-binding Ste3p GPCR from S. cerevisiae. In this study, an essential role for pre-1 in female fertility in N. crassa has been demonstrated through microscopic observation of trichogyne responses towards male cells and analysis of mature fertilized fruiting body formation. Although not required for differentiation of female reproductive structures, including protoperithecia and their specialized hyphae (trichogynes) during sexual development in either mating type, PRE-1 is essential for trichogynes from mat A strains to recognize, grow specifically towards, coil around and fuse with mat a cells. Furthermore, the observed failure of Δpre-1 mat A trichogynes to fuse with male cells even when in direct contact is in direct contrast to the ability of Δpre-1 hyphae to form heterokaryons by complementation during vegetative growth. This suggests that PRE-1 is not necessary for vegetative cell fusion events but, instead, is needed for the normal functioning of specialized female reproductive structures in Δpre-1 mat A strains. Δpre-1 strains are male-fertile in the standard fertility assay. However, the great reduction in expression of the two pheromone precursor genes in sexually differentiated tissues of Δpre-1 strains suggests that these mutants may be less competitive than males in nature under situations in which they must search for a mating partner over a distance.
Although homozygous crosses between Δpre-1 strains undergo normal sexual development, producing viable ascospores, the high expression of pre-1 in developing perithecia suggests a role for pheromone receptors after fertilization. Such a scenario may operate for the major mating type ORFs (encoded by mat a and mat A-1), as accumulating evidence implicates their involvement not only in opposite mating type recognition, but also in regulation of post-fertilization events in N. crassa (Chang and Staben, 1994; Saupe et al., 1996).
This study provides evidence for mating type-independent effects on gene expression with regard to pre-1. First, although Δpre-1 mat a strains do not exhibit obvious phenotypic defects, pre-1 mRNA is expressed in sexually differentiated tissues of mat a strains (albeit at reduced levels relative to mat A strains). Secondly, although pre-1 is dispensable for fertility in mat a strains, loss of pre-1 affects the absolute level of expression for both pheromone genes in the appropriate mating type. These results suggest that the low level of pre-1 mRNA observed in mat a cells may play a role in regulating pheromone gene expression. This could result from a pheromone-independent signalling activity of the pheromone receptor, which leads to activation of a common transcription factor that affects the expression of both pheromones, while maintaining mating-type specificity. However, the ramifications of such a mechanism for growth or sexual fertility are not obvious in our phenotypic assays.
The importance of heterotrimeric G proteins during sexual development has been demonstrated previously in N. crassa (Ivey et al., 1996; Baasiri et al., 1997; Yang et al., 2002). In this study, we show that the mating responses of trichogynes from Δgna-1, Δgna-1 Δgna-2 and Δgnb-1 strains are defective and that their protoperithecia show no indication of fertilization, similar to Δpre-1 mat A strains. However, levels of the GNA-1, GNA-2, GNA-3 and GNB-1 proteins are not affected in the absence of PRE-1. Thus, in accordance with previous results, these data further support a mechanism in which GNA-1 is coupled to a pheromone receptor and transduces the pheromone signal, whereas GNB-1 functions indirectly in the pathway by modulating levels of free Gα proteins. An involvement of gna-3 in sexual development was observed previously only during homozygous crosses; the resulting perithecia are smaller than those of wild type and lack beaks (Kays et al., 2000). The trichogyne defects of a Δgna-2 Δgna-3 double mutant observed in this study indicate that GNA-3 does affect the mating process during heterozygous crosses (perhaps by modulating cAMP levels), and that GNA-2 plays a compensatory role for GNA-3 in mating.
Mutational activation of GNA-1 does not restore fertility in a Δpre-1 mat A background, even though GNA-1 appears to be coupled to the PRE-1 pheromone receptor in mat A strains. This could result from at least two different scenarios. First, it is likely that normal chemotropism requires cycles of activation and deactivation of GNA-1. For example, in Dictyostelium discoideum and numerous bacterial species, normal chemotaxis requires adaptation to chemoattractant levels (reviewed by Bourret and Stock, 2002; Iijima et al., 2002). In S. cerevisiae, the regulator of G protein signalling (RGS) protein Sst2p plays a key role in signal desensitization by stimulating the GTPase activity of GTP-bound Gpa1p. In so doing, Sst2p activity functions as a built-in feedback mechanism for limiting signalling to a restricted time window (Dohlman, 2002).
Secondly, GNA-1 may require other protein(s) to transduce the pheromone signal. Similar to N. crassa, mating in S. pombe requires pheromones and nitrogen starvation. The S. pombe pheromone response pathway MAPK cascade is dually controlled by Ras1 and Gpa1 (reviewed by Lengeler et al., 2000). Analysis of the WICGR genome database reveals that N. crassa possesses two true ras genes (previously identified as NC-ras and NC-ras2/smco7; Altschuler et al., 1990; Kana-uchi et al., 1997). NC-RAS shows the highest sequence identity to S. pombe Ras1, but its cellular role(s) has not been described for N. crassa.
In S. cerevisiae, it has been suggested that different levels of the same pheromone signal direct mating through three distinct patterns of cell division and morphology (Segall, 1993; Erdman and Snyder, 2001). At very low levels, pheromones induce filament formation in cells, allowing the search for a mating partner. At moderate levels of pheromone, cells are arrested in G1 and form long enlarged projections, approaching potential mates by tracking pheromone gradients. At very high levels of pheromone, cells form multiple sequential projections in an attempt to contact a nearby partner and eventually to fuse. Our data show that N. crassa is capable of different patterns of trichogyne growth towards and around male cells (compare wild type and Δgna-2 Δgna-3 mutants) that are completely absent in mutants deleted for pre-1 (in mat A strains). We propose that at least some of these complex patterns of pheromone responses may also be regulated by pheromone gradients.
The MAPK cascade is highly conserved in eukaryotes and mediates various signals generated at the cell surface to the nucleus (Widmann et al., 1999). Several MAPK cascade components have now been identified in the N. crassa genome sequence (data not shown; Galagan et al., 2003). NRC-1 is similar to the S. cerevisiae Ste11p and S. pombe Byr2 MAPKKKs (Kothe and Free, 1998), and MAK-2 is a homologue of the S. cerevisiae Fus3p and S. pombe Spk1 MAPKs. Mutation of nrc-1 or mak-2 results in loss of protoperithecia development and female sterility (Kothe and Free, 1998; D. Ebbole, personal communication). Based on the observed mutant phenotypes and by analogy with the studied yeasts, we would predict that NRC-1 and MAK-2 function as the MAPKKK and MAPK proteins in the pheromone response pathway of N. crassa. However, this hypothesis awaits verification, as we have not been able to reveal a difference in MAK-2 phosphorylation levels before and after fertilization using an antibody that recognizes the phosphorylated form of MAK-2 (H. Kim and K. A. Borkovich, unpublished observations).
PRE-1 is the predicted receptor for the hydrophobic pheromone, MFA-1. Mutation of the mfa-1 gene results in numerous defects, including delayed conidial germination and reduced vegetative growth rate, drastically reduced protoperithecial formation and highly abnormal perithecial development during homozygous crosses (Kim et al., 2002). However, the absence of defects in vegetative growth or protoperithecial formation in strains mutated for its cognate receptor suggests that MFA-1 may interact with another receptor for modulation of these traits. Analysis of the N. crassa genome sequence reveals 10 GPCRs, only two of which are predicted pheromone receptors (Galagan et al., 2003). It is possible that MFA-1 serves as a ligand for another GPCR in order to regulate vegetative functions in N. crassa.
We have shown that PRE-1 plays a pivotal role in mating in N. crassa, modulating trichogynal growth towards and around male cells in mat A strains. In contrast, PRE-1 is not required for vegetative growth and development or differentiation of unfertilized female reproductive structures (protoperithecia). These results suggest that PRE-1 is essential for mating type-specific cell recognition during mating, consistent with a role in recognition of extracellular MFA-1, followed by transduction of the pheromone signal by GNA-1 and/or GNB-1. Further work will elucidate the specific roles of GNA-1 and GNB-1 in this cascade, as well as the identity and localization of PRE-1 and downstream signalling components in N. crassa.
Strains, media and growth conditions
All plasmids were maintained in Escherichia coli strain DH5α (Hanahan, 1983). N. crassa strains used in this study are listed in Table 2. N. crassa strains were cultured on Vogel's minimal medium (VM; Vogel, 1956) or sorbose-containing medium (FIGS; to facilitate colony formation on plates; Davis and DeSerres, 1970) for vegetative growth and on synthetic crossing medium (SCM; Westergaard and Mitchell, 1947) to induce production of protoperithecia. Hygromycin B was used at 200 µg ml−1 in media as indicated. When needed, l-histidine (100 µg ml−1), pantothenic acid (10 µg ml−1) and pyridoxine-HCl (10 µg ml−1) were used to supplement auxotrophic strains.
Five- to 7-day-old conidia were used to inoculate all cultures. For submerged cultures, conidia were inoculated into liquid VM at a final concentration of 3 × 106 cells ml−1. Cultures were incubated in the dark at 30°C with shaking for 5, 16 or 24 h. For cultures growing on solid medium, 1 µl of a conidial suspension was inoculated on to the centre of VM or SCM plates. For growth of cultures for RNA or protein extraction, plates were overlain with cellophane (Bio-Rad Laboratories). VM plates were grown in the dark at 30°C for 3 days, while SCM plates were grown for 6 days at 25°C under constant light. For perithecial tissues, wild-type strain 74A or 74a was grown on SCM and fertilized with a conidial suspension of opposite mating type, 74a or 74A respectively. Three, 6 or 9 days after fertilization, perithecia were scraped from the plates and frozen immediately in liquid nitrogen until used for isolation of protein or RNA. Microconidia used as males in trichogyne attraction assays were isolated as described previously (Bistis, 1981).
Gene replacement and rescue constructs
Plasmid pCSN44 contains the dominant drug resistance marker, E. coli hygromycin B phosphotransferase (hph), under the control of the Aspergillus nidulans trpC promoter, which is expressed in N. crassa (Staben et al., 1989). In the pre-1 gene replacement construct, pHK4 (Fig. 1), the 2.3 kb pre-1 region (spanning from 49 bp upstream of the translational start codon to 1111 bp downstream of the stop codon) was replaced with the 1.4 kb BamHI–SalI fragment from pCSN44 (containing the hph gene and the A. nidulans promoter). pHK4 also contains 4 kb of 5′ and 3.4 kb of 3′ flanking DNA, which were amplified from a genomic cosmid clone (G21A6; Orbach, 1994) using primers A and B for 5′ flanking DNA and primers C and D for 3′ flanking DNA in polymerase chain reactions (PCRs; Table 1). To facilitate the cloning process, an EcoRI, SmaI or SalI restriction enzyme site was introduced into each primer.
The pre-1+ complementation construct, pHK6, was made by inserting a 4.65 kb wild-type pre-1+ fragment (Fig. 1) into the PmeI and XbaI sites of the his-3 locus targeting vector pRAUW122 (Aramayo, 1996). The insert was obtained by amplification from the G21A6 cosmid in PCRs using primers E and F, which contain introduced SmaI and SpeI sites respectively (Table 1).
Strain construction and Southern analysis
pHK4 was linearized using XhoI and electroporated into N. crassa wild-type strain 74A as described previously (Ivey et al., 1996). Hygromycin-resistant transformants were isolated by plating electroporated cells on FIGS medium containing hygromycin. N. crassa genomic DNA was isolated from transformants using the Puregene DNA kit (Gentra Systems) according to the manufacturer's instruction.
Δpre-1::hph+ gene replacement heterokaryons were identified using Southern analysis. All genomic DNAs were digested with SalI, and blots were probed with the 1.6 kb AatII fragment from pHK1 (Fig. 1A). Probes were prepared using the random priming method (Feinberg and Vogelstein, 1983), with the Promega oligolabelling kit. The heterokaryotic strains were subsequently crossed to the wild-type strain 74a, and progeny were isolated after plating on hygromycin-containing FIGS medium. Homokaryotic Δpre-1::hph strains were verified by Southern analysis (Fig. 1B). The homokaryotic mutant strains were crossed to pan-2 or his-3 strains of opposite mating type in order to obtain auxotrophic Δpre-1 strains. Progeny from the crosses were plated on FIGS containing hygromycin and the necessary supplement, and were subsequently tested for pantothenic acid or histidine requirements.
To complement the Δpre-1 mutation, pHK6 was electroporated into the Δpre-1 mutant strain (16)A his-3, and transformants were selected on minimal medium. To examine the effect of Gα subunit activation in the absence of PRE-1, the (16)A his-3 strain was electroporated with the plasmid containing a predicted GTPase-deficient activated gna-1 allele (gna-1Q204L), pQY21, with selection on minimal medium (Yang and Borkovich, 1999). Heterokaryotic transformants that contained the desired integrated DNA were identified by Southern analysis, using the 3.6 kb HindIII and BamHI fragment from pRAUW122 as a probe (data not shown). All genomic DNAs were digested with HindIII. Homokaryons containing nuclei with a single copy of the targeted vector at the his-3 locus were purified using microconidia (Ebbole and Sachs, 1990) and verified by Southern analysis (data not shown).
Northern, RT-PCR and Western analysis
Total RNA was isolated using the Purescript RNA isolation kit (Gentra Systems) according to the manufacturer's instructions. Northern analysis of samples containing 20–50 µg of total RNA was performed as described previously (Tsui et al., 1994). Δpre-1 gene replacement mutants were used as negative controls for the presence of the pre-1 transcript. The first exon (0.64 kb) of the pre-1 coding sequence was amplified by PCR from the G21A6 cosmid using primers G and H (Table 1), labelled and then used as a probe to detect expression of the pre-1 gene. Levels of pre-1 mRNA were quantified by densitometry of autoradiograms using a UVP BioChemi imaging system according to the manufacturer's recommendations.
RT-PCR was performed using 1 µg of total RNA with the Access RT-PCR system (Promega) according to the manufacturer's recommendations. Primers were designed to examine the existence and/or length of two introns annotated by the WICGR Automated Gene Caller (Fig. 1A). Primer pair 1 and 2 was used to investigate the first intron, while primer pairs 1 and 3 and 4 and 5 were used to examine the second intron at different positions (Table 1; Fig. 1A). The 2.3 kb region deleted in Δpre-1 mutants was amplified from a cosmid using primers G and I (Table 1) and used directly as a probe after purification.
Extraction of the plasma membrane fraction, determination of protein concentration, gel electrophoresis and blotting for Western analysis were carried out essentially as described previously (Turner and Borkovich, 1993; Ivey, Hodge et al., 1996). GNA-1, GNA-2 and GNB-1 antisera were used at dilutions of 1:5000, while GNA-3 antisera was used at a dilution of 1:1000 (Ivey et al., 1996; Baasiri et al., 1997; Kays et al., 2000; Yang et al., 2002). A goat anti-rabbit IgG (heavy plus light chain) conjugated with horseradish peroxidase (Bio-Rad) was used as the secondary antibody at a 1:10 000 dilution. Detection was performed using the enhanced chemiluminesence method (Pierce), as described by the manufacturer, with visualization using an Epi Chemi II darkroom (UVP BioImaging Systems). A duplicate gel was stained with Coomassie blue and destained to verify equal protein loading as described previously (Sambrook and Russell, 2001)
Growth rate, conidiation, hyperosmotic sensitivity and heterokaryon formation tests. Apical extension rates of the pre-1 strains were measured on VM plates incubated in the dark at 25°C, 30°C or 37°C, and on SCM plates cultured under constant light at 25°C, as described previously (Ivey et al., 1996). Morphology of submerged cultures and growth in standing liquid cultures were examined as described previously (Kays et al., 2000). For analysis of hyperosmotic sensitivity, strains were incubated on VM plates supplemented with 0.75 M NaCl, 0.75 M KCl or 1.5 M sorbitol at room temperature in constant light (Ivey et al., 1996). Heterokaryon formation was assessed by testing the ability of two wild-type or Δpre-1 strains with different auxotrophic markers (his-3 or pan-2) to undergo hyphal fusion and form a colony (Davis and DeSerres, 1970). Conidial suspensions of the two strains were spotted on the centre of VM plates. These plates were incubated at 30°C and monitored for production of a colony.
Crosses and fertility tests. Crosses between various N. crassa strains were conducted using standard techniques (Davis and DeSerres, 1970), placing dilute conidial suspensions directly on to cultures grown on SCM. Strains fl a and fl A were used for testing mating types of progeny obtained from such crosses. To examine the role of the pre-1 gene in sexual development, the pre-1 mutant strains were crossed to wild type as either female (protoperithecial) or male (fertilizing) parent to detect dominant mating-specific defects. The Δpre-1 strains were crossed to sibling mutant strains in order to determine whether pre-1 acts as a recessive mutation affecting sexual development (Nelson and Metzenberg, 1992). The formation of protoperithecia and development of perithecia were examined microscopically using a SZX9 stereomicroscope with an ACH 1× objective lens, and images were photographed using a PM-C35B camera (Olympus America).
Trichogyne assay. Chemotropic interactions between trichogynes and conidia of opposite mating type were examined as described previously (Bistis, 1981), with minor modifications. Small blocks of 6-day-old SCM plate cultures were placed in the centre of 2% water agar (BBL Select agar; Becton Dickinson Microbiology Systems) plates, incubated for 4 days at 25°C under constant light in a humid chamber, followed by 3–5 days growth after transfer to normal humidity. Thin blocks of sterile 2% water agar were then placed on regions of the 7- to 9-day-old cultures, each covering a few to several protoperithecia. Microconidia isolated from wild-type strains 74A or 74a were subsequently applied on the top of the thin blocks using capillary pipettes. Microconidia were used as fertilizing agents as they germinate more slowly than macroconidia (Maheshwari, 1999), but are equally efficient as male mating partners (Bistis, 1981; 1983). Trichogyne orientation and growth were monitored and photographed using a BX41 fluorescent microscope with UM Plan Fluorite objective lenses and a PM-C35B camera (Olympus America). Observations were made at 10–15 h after application of microconidia and at 24 h intervals thereafter.
We thank George Bistis and David Jacobson for advice on trichogyne assays; Louise Glass and David Jacobson for advice regarding heterokaryon formation tests; Daniel Ebbole for the mak-2 strain and for sharing results before publication; and Louise Glass and Amita Pandey for the MAK-2 phosphorylation protocol. We thank members of the Borkovich laboratory for many helpful discussions. This research was supported by Public Health Service grant GM48626 from the National Institutes of Health (to K.A.B.).