Abstract
We cloned the pheromone precursor genes of Podospora anserina in order to elucidate their role in the biology of this fungus. The mfp gene encodes a 24-amino-acid polypeptide finished by the CAAX motif, characteristic of fungal lipopeptide pheromone precursors similar to the a-factor precursor of Saccharomyces cerevisiae. The mfm gene encodes a 221-amino-acid polypeptide, which is related to the S. cerevisiae α-factor precursor and contains two 13-residue repeats assumed to correspond to the mature pheromone. We deleted the mfp and mfm coding sequence by gene replacement. The mutations specifically affect male fertility, without impairing female fertility and vegetative growth. The male defect is mating type specific: the mat+ Δmfp and mat− Δmfm mutants produce male cells inactive in fertilization whereas the mat− Δmfp and mat+ Δmfm mutants show normal male fertility. Genetic data indicate that both mfp and mfm are transcribed at a low level in mat+ and mat− vegetative hyphae. Northern-blot analysis shows that their transcription is induced by the mating types in microconidia (mfp by mat+ and mfm by mat−). We managed to cross Δmfp Δmfm strains of opposite mating type, by complementation and transient expression of the pheromone precursor gene to trigger fertilization. These crosses were fertile, demonstrating that once fertilization occurs, the pheromone precursor genes are unnecessary for the completion of the sexual cycle. Finally, we show that the constitutively transcribed gpd::mfm and gpd::mfp constructs are repressed at a posttranscriptional level by the noncognate mating type.
In heterothallic fungi, pheromones are major components in mating type determination and the initiation of sexual reproduction. Pheromone production and the corresponding cellular response have been investigated in detail in yeasts, providing frameworks for the study of pheromone signaling in other fungi (reviewed in reference 26). Typically, pheromones are secreted to attract mates of opposite mating type without cell-cell contact. They bind to receptors on the surface of the sexually compatible cells and initiate a signal transduction pathway involving a MAP kinase cascade, which induces morphological and physiological changes preparing cells for fusion (reviewed in reference 21).
In unicellular yeasts, cells of opposite mating type signal each other by a double pheromone-receptor interaction. Both gene pairs are present in each haploid genome, but each cell type expresses only one pheromone and the receptor specific for the other. Expression is directly regulated by the transcription factors encoded by the mating type genes. In filamentous ascomycetes, where the sexual process is more elaborate, recognition between compatible mating partners occurs between reproductive male and female structures. The existence of a hormonal mechanism was first shown by the ability of male cells to orientate the growth of trichogynes, the receptive hyphal filaments produced by the female organs (reviewed in reference 6). The attraction of trichogynes was only observed in intermating type combinations. That male cells producing a diffusible pheromone were chemo-attractive to trichogynes of opposite mating type suggested that the compatible female organs express the corresponding receptor. This assumption has recently been confirmed by the characterization of the pheromone receptor gene pre-1 in Neurospora crassa and the construction of knockout mutants (27). Δpre-1 mutants are female sterile in a mat A context because the chemotrophic growth of trichogynes towards mat a microconidia is abolished.
Genes encoding putative pheromone precursors have been characterized in several heterothallic filamentous ascomycetes: Cryphonectria parasitica (40), Magnaporthe grisea (36) and N. crassa (7). These genes encode two classes of polypeptides structurally similar to the a-factor and the α-factor precursors of Saccharomyces cerevisiae. The a-factor belongs to the CAAX family of hydrophobic peptide pheromones, and the α-factor is a 13-amino-acid oligopeptide which results from processing of a larger polypeptide containing several copies of the active tridecapeptide. Genes of both classes are present as a single copy in the haploid genome of filamentous fungi, except for the C. parasitica Mf2 gene, encoding a precursor of the a-factor-like pheromone, which is duplicated. In the three fungi, the pheromone precursor genes are transcribed in a mating type specific manner. From these observations it seemed clear that, in filamentous heterothallic ascomycetes, the primary role of pheromones is the control of initial recognition between male cells and female organs of opposite mating type.
Interestingly, both types of pheromone precursor gene have been identified in the homothallic ascomycete Sordaria macrospora (33). In this self-fertile fungus, the two pheromone precursor genes were shown to be transcribed in mycelium, but experimental evidence confirming that they are truly functional is lacking. The presence of pheromone precursor genes in homothallic fungi, in which initiation of sexual reproduction is not under mating type control, suggests that these genes might do more than mediate chemo-attraction between male and female reproductive structures. Evidence for extended functions of pheromones has already been provided in several fungi. In Schizosaccharomyces pombe, in contrast to S. cerevisiae, the pheromone signal is used not only for cell-cell recognition and fusion, but also for the initiation of meiosis; it induces the expression of mat1-Pm and mat1-Mm mating type genes, which activate transcription of mei3, the direct inducer of meiosis (39). In homobasidiomycetes, pheromone signaling regulates the postfusion events of nuclear migration and clamp cell fusion (reviewed in reference 8).
The function of pheromones was recently investigated in C. parasitica and N. crassa. In C. parasitica, a complete deletion of the Mf1-1 gene encoding the α-factor-like pheromone was obtained by site-directed recombination (38). Male cells produced by the Mf1-1 knockout mutant were defective in fertilization, but this strain was fully competent as a female parent and was not impaired in vegetative growth and asexual reproduction. In contrast, deletion of one of the two copies of the Mf2 gene encoding the a-factor like pheromone resulted in a pleiotropic phenotype: this strain produced barren perithecia when crossed as female with the wild-type strain as male, and its asexual reproduction was reduced (41). In N. crassa, mutants of the mfa-1 pheromone precursor gene, encoding the a-factor-like pheromone specific of the mat a mating type, were isolated by the repeat-induced point mutation (RIP) approach (28). The mfa-1 mutants, assumed to correspond to null mutants, were unable to attract mat A trichogynes and were thus sterile as male parent in sexual crosses. Surprisingly, these mutants were also affected in female fertility, perithecial development and vegetative growth. This phenotypic pleiotropy, and the fact that the mutant phenotype was observed in both mating type backgrounds, suggests that mfa-1 plays complex roles throughout perithecial development as well as in vegetative growth. These data suggest that pheromone genes may have additional functions beyond their role in fertilization.
We have addressed the biological role of the pheromone precursor genes in the filamentous ascomycete Podospora anserina, and have also explored regulation of pheromone expression by mating types. The structure of both mat+ and mat- idiomorphs has been determined, and the function of the four mating type genes (one in mat+, three in mat-) has been characterized (for a review see 14). Two of the genes encoding putative transcriptional factors, FPR1 (for fertilization plus regulator) in mat+ and FMR1 (for fertilization minus regulator) in mat-, were shown to control fertilization (17). In P. anserina, as in N. crassa (5), microconidia of one mating type can attract trichogynes from protoperithecia of opposite mating type (Bistis, personal communication). This chemotropic response suggested that mat+ and mat- microconidia produced diffusible pheromones whose expression could be controlled by FPR1 and FMR1, respectively.
The discovery that some fpr1 and fmr1 mutants had acquired the ability to self-fertilize and to fertilize a partner of the same mating type with low efficiency, while still able to cross with a partner of opposite mating type with high efficiency, has complicated this classic scheme (2). This observation suggests that the FPR1 and FMR1 genes not only activate their specific set of pheromone/receptor, but also repress the complementary pheromone/receptor set. The repressor function would be lost in some mutants, alleviating mating specificity. The molecular characterization of the pheromone precursors genes, the predicted targets of FPR1 and FMR1 regulators, is necessary to understand how these positive and negative controls operate. The manipulation of pheromone precursor genes may also permit the determination of necessity for postfertilization pheromone signaling in P. anserina.
During fertilization in P. anserina, as in other self-sterile filamentous ascomycetes, the male nucleus enters the trichogynial tip and travels to the primary ascogonial cell within the protoperithecium. The male and female nuclei do not fuse at this step. They proliferate in a coenocytic condition, and migrate, paired, to specialized dikaryotic cells, giving rise to ascogenous hyphae in which karyogamy, meiosis and formation of ascospores take place. The early events of the P. anserina sexual cycle are presented in (42). The accuracy of this intricate development relies on the proper pairing of one mat+ and one mat− nucleus at the transition from plurinucleate to dikaryotic cells. The mating type genes control the pairing between sexually compatible nuclei. The FPR1 gene confers mat+ nuclear identity, while the FMR1 and SMR2 genes confer mat− nuclear identity (1, 42). The molecular mechanism underlying this recognition has not been elucidated, but models implying the pheromone response pathway have been proposed (13, 16).
In this paper we present the characterization of the two pheromone precursor genes of P. anserina, cloned using their previously isolated homologs in N. crassa and C. parasitica when this work was initiated. They were named mfp and mfm for mating factor plus and mating factor minus, since their transcription is controlled by mat+ or mat− mating type. Deletion of mfp and mfm was found to exclusively impair male fertility in a mating type specific manner, without impairing female fertility and vegetative growth. We present data demonstrating that the mfp and mfm genes serve a unique role at fertilization and are not necessary afterwards. Finally, we have subtracted mfp and mfm from the control of the mat genes by replacing their 5′ non coding sequence by the constitutive glyceraldehyde-3-phosphate dehydrogenase (gpd) promoter. We found that expression of gpd::mfp was repressed posttranscriptionally by mat−, whereas expression of gpd::mfm was repressed posttranscriptionally by mat+.
MATERIALS AND METHODS
P. anserina biology and genetic methods.
P. anserina is a heterothallic filamentous ascomycete whose life cycle and general methods for genetic analysis have been described (22, 35) and are presented in http://cgdc3.igmors.u-psud.fr/Podospora/index.htm. The ascus of P. anserina contains 4 ascospores which develop around two nonsister nuclei after one postmeiotic mitosis; thus, they are dikaryotic (recently reviewed in reference 31). However, 2 to 5% of asci contain five ascospores among which two are smaller and uninucleate and give rise to homokaryotic mycelium. Tetrad analysis is routinely carried out on five-spored asci. All crosses and phenotypic analyses were performed on minimal synthetic agar medium.
Crosses were performed by spotting implants of mat+ and mat− strains on each half-plate of a petri dish. Since microconidia are not mobile and cannot fertilize distant female organs, sterile water was added to disperse the microconidia produced by the two parents and to promote fertilization over both parental mycelia. In spermatization assays, strains of opposite mating type were inoculated on different petri dishes. Microconidia were recovered from the strain acting as male by washing the surface of the mycelium with 1.5 ml of sterile water. The microconidial suspension was spread on the mycelium of the strain acting as female. The number of microconidia increased with the age of the male culture (106 at 7 days and 108 at 30 days on one petri dish). If necessary, fragmented mycelium can replace microconidia in spermatization assays: 1 cm2 of mycelium from a culture grown on a petri dish was put into a 1.5-ml tube containing 1 ml of water and fragmented with the FastPrep Cell Disrupter Instrument (Q-BIOgene, Illkirch, France) twice 40 s at 4 m/s.
To estimate fertilization efficiency, microconidia recovered from 8 day-old cultures were counted using a hemacytometer; 1 ml of microconidial suspension (or of serial dilutions) was spread on wild-type mycelia of opposite mating type, which had grown 4 days at light to induce formation of protoperithecia. Perithecia were counted 5 days after fertilization. For each strain, estimations were performed with several cultures grown on the same batch of medium.
Transformation.
Transformation with integrative plasmids was performed as previously described (32). About 5 μg of plasmid DNA were used in each assay. In transformation with the pFAC self-replicating plasmid, 100 to 500 ng of plasmid DNA digested with NotI to release telomeric ends were used in each assay. When necessary, hygromycin (Roche Diagnostics, Meylan, France) was added to the protoplast regeneration medium at a concentration of 100 μg/ml. Segregation of antibiotic resistance in the sexual crosses was scored on minimal medium containing 75 μg/ml hygromycin. In P. anserina, transforming DNA generally integrates at random ectopic chromosomal positions, resulting in a complex structure. To circumvent the potential phenotypic heterogeneity that may result from position effect, analysis was routinely performed on a set of independent transformants carrying the same transgene at different ectopic positions. Data obtained with primary transformants are routinely checked on purified transformants generated by crossing.
Biological assays with the synthetic peptide.
Hypothetical mat- mature pheromone peptide (QWCLRFVGQSCW) was synthesized by Sigma-Genosys. The peptide was dissolved with a small volume of dimethyl sulfoxide and diluted with sterile water just before the assay achieved to demonstrate its biological activity. A peptide solution (15 to 150 μg/ml) failed to induce formation of perithecia when spread on the surface of a mat+ culture acting as female. Fertilizing assays with microconidia from a mat− Δmfm strain suspended in a peptide solution and spread on the surface of a mat+ culture remained unsuccessful. Since the yeast a-cells respond to spatial gradients of α mating factor, assays were performed using gradients of the synthetic peptide: 15 to 600 μg of peptide were put on a filter disk placed on a mat+ female culture previously overlayed with mat− Δmfm microconidia but no fertilization was observed. We determined whether the synthetic peptide could inhibit the fertilizing activity of mat− microconidia, by adding the peptide to microconidial suspensions at a 15 to 150 μg/ml concentration before spreading these on mat+ cultures acting as female: no inhibitory effect was obtained.
Bacterial strains, cosmid, and plasmids.
Cosmids and plasmids were amplified in Escherichia coli DH5α (24). Plasmid pGEM-T (Promega, Charbonnières-les-Bains, France) was used for cloning of PCR-generated fragments. Plasmid pCB1004 (9) was used for subcloning P. anserina genomic DNA. It contains the hph gene conferring resistance to hygromycin. Plasmid pUL (13) is a derivative of pUC18 carrying the 2,191-bp HindIII-PstI fragment containing the leu1 gene from P. anserina (a kind gift from Béatrice Turcq). The pFAC self-replicating plasmid was described in (3).
Cloning of the mfp gene.
The mfp gene was cloned by PCR with degenerate primers. The sense primer (D5-mfa: 5′-CCCACAAACCATCAAMATGCCNTC-5′) and antisense primer (D3-mfa: 5′-CTCCAATAYYACATNACNACRCARTA-3′) were based on an alignment of pheromone genes shown in Fig. 1A and B. Primers (100 pmol each) were used in a 50-μl reaction mixture with 1 unit of Taq DNA polymerase (Q-BIOgene, Illkirch, France) and as a template P. anserina DNA (100 ng) or plasmid pB14A (1 ng) containing the mfa-1 gene of N. crassa (a kind gift from D. Bell-Pedersen and D. Ebbole) as a positive control. Cycling conditions began with a low-stringency cycle at 37°C. The transition from 37°C to 68°C was performed by a ramp of 0.3°C/mn. This was followed by 40 cycles at a 50°C. The P. anserina DNA produced a candidate 90 bp product that was cloned in the pGEM-T plasmid (Promega, Charbonnières-les-Bains, France). Three clones containing inserts of the expected size were sequenced. Sequence analysis revealed that the amplified DNA encodes a polypeptide with significant amino-acid homology to the N. crassa and C. parasitica pheromone precursor genes. One PCR product was used as a probe to screen a genomic cosmid library from P. anserina (20). Finally, one 1-kb BamHI-BamHI fragment was isolated and cloned in the pUC18 plasmid, giving rise to pucBBmfp.
FIG. 1.
mfp pheromone precursor gene of Podospora anserina. A) Alignment of the region encompassing the initiation codon of MF1-1 of M. grisea (36), ppg2 of Sordaria macrospora (Sm, accession no AJ259862), mfa-1 of N. crassa (Nc, AF397732), mf2-1 of C. parasitica (Cp, U92043) and mfp of P. anserina (Pa, AY829471). The last line shows the D5mfa primer used for PCR cloning of the pheromone precursor gene of P. anserina. A 5′ tail was added to the primer to increase the length of the PCR product. B) Alignment of the region of mfa-1 of N. crassa and mf2-1 of C. parasitica used to design the D3mfa primer. C) Alignment of the deduced mfp protein of P. anserina (Pa), mfa-1 of N. crassa (Nc, AAN03594), ppg2 of Sordaria macrospora (Sm, CAB96171), mf2-1 of C. parasitica (Cp, AAC39329), and MF1-1 of M. grisea (Mg) (36). The alignment was obtained with the Clustal W software and presented with the Boxshade program. Similar and identical amino acids are shown in gray and black, respectively.
Cloning of the mfm gene.
The mfm gene was cloned by heterologous hybridization with the ccg4 gene from N. crassa (7). The ccg4 coding sequence was amplified from plasmid pBP11 (a kind gift from D. Bell-Pedersen and D. Ebbole) with primers 5-ccg4 (5′-ATGAAGTTCACTCTCCCTCTT-3′) and 3-ccg4 (5′-GGAGTGAGCAGTGAGGATGG-3′). DNA probes prepared from the PCR product were hybridized at low stringency (32) with P. anserina genomic DNA digested by PstI and SalI. This Southern blot allowed us to identify a PstI-SalI fragment of a size between 1.6 and 2 kb, which gave a strong hybridization signal. One colony, containing a recombinant pUC18 with a 1.8-kb PstI-SalI insert was selected through hybridization from a subgenomic library. Sequencing confirmed that the insert contained a sequence similar to the ccg-4 gene. Finally, the 2.7-kb PstI-PstI fragment containing the entire putative pheromone precursor gene, which was called mfm (mating factor minus), was isolated from a P. anserina bacterial artificial chromosome library (37) and cloned in pCB1004 (9), yielding pCBMF-P3.
Deletion of the resident mfp gene.
Plasmid pucmfp::leu1 was prepared for the mfp knockout in P. anserina. The leu1-bearing PstI-HindIII cassette derived from the pUL plasmid (13) replaces the mfp coding sequence and is flanked by genomic sequences of 523 bp and 405 bp from the mfp locus. These sequences were obtained by PCR with plasmid pucBBmfp as a template. The mfp sequence upstream of the start codon was amplified using the reverse primer and oligonucleotide Pstdmfp (5′-AACTGCAGTTACTGAAGATGTTCAGAAG-3′), localized 21 nucleotides upstream of the mfp start codon. This 523-bp fragment was digested with BamHI and PstI. The mfp sequence downstream of the stop codon was amplified with the universal primer and oligonucleotide Hind3dmfp (5′-CCCAAGCTTGACCAACATGGTAT-3′), localized 2 nucleotides downstream of the TAA stop signal. The resulting 405-bp fragment was digested with HindIII and BamHI. The three fragments were introduced into pUC18.
Targeted mfp replacement was performed by transformation of mat+ leu1-1 protoplasts with the pucmfp::leu1 plasmid linearized with BamHI. A total of 511 Leu+ transformants were screened for replacement by the following PCR method. Twenty transformants were grown on a cellophane disk placed at the surface of minimal medium (four columns of five). After a 2-day growth a small band of mycelium was taken between two columns, which represented ten transformants. DNA was prepared from these pools of 10 transformants according to a procedure adapted from (29). The mycelium was put in a tube containing 200 μl of buffer (Tris, pH 8, EDTA 10 mM, SDS 2%), subjected twice to freezing and thawing, and extracted with phenol-chloroform pH 8. The aqueous phase is precipitated with an equal volume of propanol-2, centrifuged 10 min and the pellet resuspended in 40 μl of sterile distilled water. PCR is performed on 1 μl aliquots using DMF3D (5′-AAAGTGAGCCCGATTTGTTC-3′) and Leu2H (5′-ACAGCTATGACGTAGATTGG-3′) to screen for the right junction of the deletion. DNA from positive pools was tested with DMF5G (5′-TCTCTGACGGCAAAGACGCG-3′) and Leu1P (5′-GGCGTTCGATGACTGAATTT-3′) specific for the left junction.
The same assay was then performed on each transformant of the positive pool to identify putative candidates that were subjected to phenotypic and molecular analysis. The mat+ leu1-1 Δmfp(leu1+) null mutant, identified through PCR screening and assessed by Southern blot analysis, was crossed with the mat− leu1-1 strain. In leu1-1 homozygous crosses, segregation of the cassette-replaced locus could be easily followed, since it contained a leu1+ copy conferring prototrophy. Segregation of leucine auxotrophy and prototrophy indicated that the mfp locus is distal from the centromere (60% postreduction) and is not linked to the mating type locus. A mat+ leu1-1 Δmfp(leu1+) strain identified in the progeny was crossed with mat− leu1+ mfp+ strain to eliminate the leu1-1 mutation and to associate the mfp knockout with each mating type. The recovered strains are named mat+ Δmfp and mat- Δmfp.
The structure of the replaced Δmfp locus was confirmed by Southern blot analysis. Firstly, no signal was detected in hybridizing the BamHI-digested DNA of the Δmfp(leu1+) strain with the mfp coding sequence (data not shown). Second, on a blot probed with the 1-kb BamHI fragment encompassing the mfp gene, a 1-kb fragment was observed in the DNA extracted from the wild-type strain, while the DNA extracted from the putative Δmfp(leu1+) contained an expected 3.2-kb fragment resulting from the replacement of the mfp coding sequence by the leu1 cassette (data not shown). Third, this blot was deshybridized and probed with the BamHI-PstI fragment bearing leu1; the same 3.2-kb band was revealed confirming insertion of the leu1 cassette within the mfp genomic locus (data not shown).
Deletion of the resident mfm gene.
The pUCmfm::leu1 plasmid was constructed by placing the leu1 gene between two fragments corresponding to the 850 bp and 1170 bp sequences upstream and downstream of the mfm coding sequence. The 850 bp sequence was amplified from pCBMF-P3 by the universal primer and primer Asc-del-MFM (5′-TTGGCGCGCCGGTGGTTTTTGGTTGGGAAG-3′), and the PCR product was digested by PstI and AscI. The leu1 gene was amplified from pUL (13) by primers Mlu-Leu (5′-CCGACGCGTGCTTTGGGGAGTTCCAAGTC-3′) and Leu-Mlu (5′-CGGACGCGTGAATGACAGCTGCTATATATAC-3′) and the PCR product was digested by MluI. The 1,170-bp sequence was amplified from pCBMF-P3 by primers del-MFM-Asc (5′-GCGGCGCGCCGGTCAACTCGGCAACAGC-3′) and reverse primer, and the PCR product was digested by BamHI and AscI. These three fragments were ligated in pUC18 digested by PstI and BamHI to yield pUCmfm::leu1.
Targeted mfm replacement was performed by introducing the pUCmfm::leu1 plasmid, previously digested with PstI, to linearize the replaced genomic fragment, into mat- leu1-1 protoplasts. Among the 111 Leu+ transformants that were subjected to fertilization assays, six were found to exhibit male sterility. Replacement of the resident mfm coding sequence by the leu1 cassette was confirmed by PCR analysis of DNA from male sterile transformants with primer pairs specific of the left (D5mfm 5′-TCATCCACCACGCCGGGTCA-3′/Leu1P 5′-GGCGTTCGATGACTGAATTT-3′) and right (D3mfm 5′-GACACCCTCCGCTTTGTCCA-3′/Leu2H 5′-ACAGCTATGACGTAGATTGG-3′) junctions of the deletion. A Southern blot analysis was performed to determine the structure of the replaced mfm locus in the six candidates. Two of these displayed the expected structure at the mfm locus while the others carried rearrangements. Deletion was confirmed, since no band was obtained by hybridizing DNA extracted from putative knockout mutants with a probe corresponding to the mfm coding sequence (data not shown). When the blot was hybridized with a 2-kb PCR fragment overlapping the mfm coding sequence, the 2.7-kb PstI-PstI band corresponding to the resident mfm locus was replaced by a 4.9-kb fragment created by insertion of the leu1 cassette, as checked by further hybridization with a leu1 probe (data not shown). One of the two transformants with proper deletion was chosen for crossing and generation of mat+ Δmfm and mat− Δmfm strains.
Determination of fertilization efficiency of mat+ Δmfp and mat− Δmfm mutants.
Concentrated microconidial suspensions (>106/ml) were recovered from two week-old cultures and spread on wild-type mycelium of opposite mating type acting as female. To ascertain the origin of perithecia, we used as the female parent a strain carrying the 136 mutation, which results in nonpigmentation of the perithecial tissue (30). The cross between male mat+ Δmfp and female mat− 136 or between male mat− Δmfm and female mat+ 136 resulted in nonpigmented perithecia whereas false-positives due to contaminant hyphae resulted in pigmented wild-type perithecia.
Association of the Δmfp and Δmfm mutations.
The cross mat+ mfp+ Δmfm X mat− Δmfp mfm+ was performed to generate various combinations of mating type with functional and deleted pheromone genes. Twenty five-spored asci were analyzed. In each ascus, segregation of the mfp+/Δmfp and mfm+/Δmfm allelic pairs in the heterokaryotic binucleate spores was first deduced from the genotype of the small homokaryotic spores, which was established through their phenotype as male partner and/or by PCR analysis. The deduced genotype was then confirmed by genetic analysis of the progeny recovered from self-fertilization.
Construction of the pFACmfp and pFACmfm plasmids and preparation of perithecial cDNA.
pFACmfp was constructed by cloning the 1-kb BamHI fragment of pUCBBmfp at the BglII site of pFAC (3). pFACmfm was constructed by ligating the 2.7-kb BamHI-EcoRI insert of pCBMF-P3 with the pFAC digested with BglII, then blunt-ending the EcoRI and BglII site with the Klenow enzyme and ligating again. For perithecial RNA isolation, the Δmfp Δmfm strain was grown on 20 plates covered with sterile cheesecloth circles and fertilized with a conidial suspension of opposite mating type displaying pFAC-driven transient expression of the appropriate pheromone precursor gene. The perithecia were scraped 2 days after fertilization, freezed in liquid nitrogen and grounded with a mortar. RNA were extracted according to the acidic phenol method of Chomczynski and Sacchi (10) and precipitated with half a volume of isopropanol and half a volume of sodium citrate 0.8 M, sodium chloride 1.2 M. RNA were redissolved in water and purified on a RNeasy column (Qiagen, Courtaboeuf, France). If necessary, polyA RNA were isolated with Oligotex mRNA Kit (Qiagen). The mfm or mfp mRNA was searched in total RNA (1 microgramme) or polyA RNA (0.5 micrograms) with the Titan-One Tube RT-PCR Kit (Roche Diagnostics, Meylan, France) with MFD2 (5′-CTATGTCGCCAACCATATAC-3′)/MFG4 (5′-TCAAGCGGTGCACCGCCGAGGA-3′) and 3RFP (5′-CCGCATCCACTTCTGAACATCTTC-3′)/delDR (5′-AGCCGATTGATGTTGTCGTA-3′), respectively.
Construction of the gpd::mfp fusions.
Plasmid pCBgpd::mfp1, derived from pCB1004, contains a translational fusion between the glyceraldehyde-3-phosphate dehydrogenase (gpd) promoter from P. anserina and the mfp coding sequence. The 350 bp EcoRI-NcoI fragment, containing the translation start signal and the proximal part of the gpd 5′ noncoding sequence, was prepared from the PRP81-1 plasmid (34). A fragment corresponding to the mfp coding sequence followed by 408 bp at its 3′ end was amplified by PCR on pucBBmfp DNA with the universal primer and Ncomfp (5′-CATGCCATGGCTTCCACCACCGCTC-3′). The EcoRI-NcoI gpd fragment and the PCR product cleaved with NcoI and BamHI were ligated into the EcoRI/BamHI sites of plasmid PCB1004. Plasmid pCBgpd::mfp2 contains a promoter fusion between the gpd promoter and a fragment starting at the CAAA nucleotides preceeding the ATG initiation codon of mfp and including the 408 bp 3′ end. The gpd fragment was obtained by PCR amplification on the PRP81-1 plasmid with oligonucleotide 5GPD (5′-TTATGTCGCTCATGCCACCA-3′) and Bamgpd (5′-CGGGATTCTGTTGGTGAAGAGAGA-3′). The mfp fragment was obtained by PCR amplification of pucBBmfp with BHmfp (5′-CGGGATCCAAAATGCCTTCCACCACC-3′) and the universal primer. The gpd PCR product digested with EcoRI/BamHI and the mfp PCR product digested with BamHI were cloned into the EcoRI/BamHI sites of pCB1004. The two fusions were checked by DNA sequencing.
Construction of the gpd::mfm fusions.
The mfm coding sequence plus 438 bp at its 3′ end was amplified from pCBMF-P3 using primers Nco-MFM (5′-CATGCCATGGAGTTTTCCACCCCCCTC-3′) and 3-MFM-Pst (5′-CACACTGCAGGTCACACTTCACGTCAAG-3′), and the PCR product was digested by NcoI and PstI. This fragment was cloned with the 0.35-kb EcoRI-NcoI fragment containing the gpd promoter of P. anserina in pCB1004 digested by EcoRI and PstI. Fusion was checked by DNA sequencing.
Purification and genetic analysis of transformants carrying the gpd::mfm or gpd::mfp fusion.
The hygromycin-resistant transformants mat− mfp+ Δmfm (gpd::mfm-hph) and mat+ Δmfp mfm+ (gpd::mfp-hph) were crossed with a mfp+ mfm+ strain of opposite mating type. Fifteen five-spored asci were screened for hygromycin resistance and mating type. The presence of either Δmfm or mfm+ and either Δmfp or mfp+ allele in homokaryotic haploid spores was determined by fertility assays and/or PCR analysis. In the cross between mat+ Δmfp mfm+ (gpd::mfp1-8-hph) and mat− mfp+ mfm+, no mat+ male sterile progeny was recovered among 30 analyzed asci, which indicated that the gpd::mfp1-8 transgene cosegregated with Δmfp null mutation and was thus genetically linked to Δmfp. A PCR analysis was carried out with DNA from the transformant carrying gpd::mfp1-8 to determine the structure of the integration locus. Structure of the integrated gpd::mfp1-8 transgene was analyzed by PCR.
The following crosses were performed to introduce the gpd::mfm fusion into Δmat and mat+ mutant backgrounds. mat− Δmfm (gpd::mfm) × Δmat mfm+ (mat+) generated Δmat (gpd::mfm). mat− (gpd::mfm) × fpr1M112I 136 generated fpr1M112I (gpd::mfm). mat− (gpd::mfm) × fpr1::ura5 136 generated fpr1::ura5 (gpd::mfm). Each of the three crosses were performed with two transformants carrying the fusion integrated at two different positions.
The following crosses were performed to introduce the gpd::mfp1-8 transgene into Δmat and mat− mutant backgrounds. mat+ Δmfp (gpd::mfp1-8)× Δmat mfp+ (mat-) generated Δmat (gpd:mfp1-8). mat+ Δmfp (gpd::mfp1-8)× smr2::ura5 mfp+ (SMR1 SMR2-ble) generated smr2::ura5 Δmfp (gpd::mfp1-8). mat+ Δmfp (gpd::mfp1-8)× fmr1::ura5 mfp+ (FMR1-hph) generated fmr1::ura5 Δmfp (gpd::mfp1-8).
Nucleic acid extraction and analysis.
All common molecular biological manipulations were carried out according to standard methods. DNA and RNA from P. anserina were prepared as previously described (12, 15).
RESULTS
Cloning of the mfp pheromone precursor gene.
The mfp gene was cloned with PCR with degenerate primers. When the experiment was performed, the sequences of pheromone precursor genes were available in N. crassa and C. parasitica, two filamentous ascomycetes related to P. anserina. They encode the two types of pheromones that were extensively characterized in S. cerevisiae. The mfa-1 gene of N. crassa and the Mf2-1 and MF2-2 genes in C. parasitica were specifically expressed in the mat+-like mating type context (7, 40). They encode small polypeptides with a C-terminal CAAX motif structurally related to the precursor of the S. cerevisiae a-factor.
A couple of degenerate primers were designed from the alignment of the two 5′ UTRs and coding sequences (Fig. 1A and B). PCR amplifications of P. anserina genomic DNA allowed us to identify a 90-bp candidate band. After cloning and characterization of fragments prepared from this band, a sequence encoding a polypeptide with significant amino acid homology to the N. crassa and C. parasitica pheromone precursor was identified. This sequence was used as a probe to screen a genomic cosmid library from P. anserina and to identify a 1-kb BamHI restriction fragment (see Materials and Methods for details).
Sequence analysis (GenBank AY829470) indicated the presence of a putative gene encoding a 24-amino-acid polypeptide exhibiting high conservation with lipopeptide propheromones from ascomycete filamentous fungi (Fig. 1C). When the sequence encoding the polypeptide was used in hybridizing Southern blots of BamHI-digested DNA from P. anserina, a unique 1-kb fragment was revealed, indicating that the mfp gene was present as a unique copy in the genome (data not shown). Analysis of the 5′ noncoding region of mfp revealed a CAAAG sequence at position −223 upstream of the start codon, which corresponds to the pheromone response element binding the PRF1 HMG protein in Ustilago maydis (25). A potential TATA box is present at position −101. Strikingly, the region just upstream of the putative ATG initiation codon is strongly conserved in the lipopeptide pheromone precursor genes of P. anserina, N. crassa, Sordaria macrospora, and C. parasitica (Fig. 1A). Further analysis is necessary to determine if these sequences are functionally significant.
Cloning of the mfm pheromone precursor gene.
The mfm gene was cloned by hybridization with the N. crassa clock-controlled gene 4 (ccg-4) encoding the putative matA-specific pheromone precursor (6). When used as a probe, a DNA fragment of the ccg-4 gene coding region consistently detected a 1.6- to 2-kb fragment on Southern blots of total genomic DNA from P. anserina. Hybridization of a subgenomic library isolated a clone containing part of a candidate gene, which was used to seek the entire corresponding locus in the P. anserina BAC library (see Materials and Methods for details). Finally, subcloning isolated a 2.7-kb PstI-PstI fragment containing the sought mfm open reading frame (GenBank AY829471): this encodes a 221-amino-acid polypeptide exhibiting high conservation with polypeptides encoded by ccg-4 in N. crassa and ppg1 in Sordaria macrospora (Fig. 2).
FIG. 2.
Alignment of the deduced proteins of mfm (Pa) and mfA-1 of N. crassa (Nc, accession Q01301) and ppg1 of Sordaria macrospora (Sm, CAB 96172). For methods, see the legend to Fig. 1. Oligopeptide repeats are boxed. The two repeats present in the mfm gene product are numbered.
As is the case for its orthologs, the mfm coding region shows structural features similar to the α-factor precursors encoded by the Mfα1 and Mfα2 genes in S. cerevisiae. The α-factor precursors are polypeptides of 165 and 120 amino acids which contain, respectively, four and two copies of the 13-amino-acid mature pheromone and must undergo several proteolytic cleavages before the mature biologically active pheromone is secreted from the cell. Within the 221-amino-acid polypeptide deduced from the mfm coding sequence there are two repeats of a dodecapeptide, which may correspond to the active pheromone. This putative pheromone is well conserved, with the undecapeptide present at five copies in the ccg-4 and ppg1 coding region (Fig. 2).
Mating type-specific expression of mfp and mfm.
In the heterothallic filamentous ascomycete C. parasitica (40), M. grisea (36) and N. crassa (7), it was shown that pheromone precursor genes are expressed in a mating type specific manner. In P. anserina, Northern blots were first performed with RNA extracted from vegetative mycelium of mat+ and mat− wild-type strains. Hybridization with mfm and mfp sequences gave no signal, suggesting that expression of these genes in vegetative mycelium is very low (data not shown). Since it is technically difficult to obtain large amounts of purified microconidia in P. anserina, we took advantage of the incA mutant (4), which produced 1,000 times more microconidia than the wild-type strain (23). Northern analysis of total RNA from incA vegetative mycelium of both mating types confirmed that mfp is only transcribed in mat+ whereas mfm is only transcribed in mat− strains (Fig. 3). To confirm regulation of the mfp and mfm genes by the mating type locus, RNA from an incA Δmat strain devoid of mating type sequence was analyzed. Neither the mfp nor the mfm transcript could be detected in this strain (data not shown). The mfp and mfm transcripts were estimated at 0.9-kb and 1.1-kb long, respectively.
FIG. 3.
Mating type-specific transcription of pheromone precursor gene. The Northern blot contains total RNA (10 and 90 μg/ml) extracted from mat− incA and mat+ incA strains grown in minimal liquid medium. The blot was probed sequentially with PCR fragments specific for mfp, mfm, and AS1. It was stripped between each hybridization. AS1, encoding the S12 ribosomal protein (19), was used as a loading control. The sizes of rRNA genes are given on the right.
Construction of mfp and mfm null mutants.
To study the function of the pheromone precursor genes, we generated mfp and mfm null mutants by gene replacement. Plasmids containing a null allele of mfp and mfm marked with a leu1+ cassette were introduced into a leu1-1 mat+ and leu1-1 mat− recipient, respectively. The recovered leucine prototrophic transformants were screened for replacement of each resident gene by the interrupted transgene. A molecular screen based on a PCR assay was used to search for mfp deletion. Specific primers were used which gave an amplicon only if the resident mfp locus had been replaced by the interrupted transgenic copy: one primer was localized in the genomic sequence flanking the mfp sequence cloned and manipulated in the plasmid, and the second primer was localized at one extremity of the leu1 cassette. Fast screening was performed on DNA from pools of ten transformants (see Materials and Methods).
A total of 511 transformants were tested and yielded two candidates which were subjected to Southern blot analysis as described in Materials and Methods. One strain with a replacement of the mfp coding sequence with the leu1 cassette was named Δmfp and used for further analyses. As was expected for a pheromone gene mutant, the mat+ Δmfp strain was found to be sterile as a male partner (see below). Consequently, this phenotype was used for the identification of the mfm knockout in the mat− strains. Six sterile male candidates were found out of 111 Leu+ transformants and subjected to Southern blot analysis (Materials and Methods). Two of these displayed the expected structure at the mfm locus, while the others carried rearrangements. One of the two positive transformants called Δmfm was subjected to detailed genetic analysis.
Phenotype of mfp and mfm null mutants.
The Δmfp and Δmfm mutations were associated with mat+ and mat− mating types by crosses (Materials and Methods) and the different strains were examined phenotypically. They did not differ from the wild-type strain at the level of germination, vegetative growth, mycelial morphology, and microconidial production (data not shown). They were subjected to mating tests as shown in Fig. 4. Such tests define the ability of a strain to mate as a male or a female partner. It can be seen that perithecia are present on the mat− Δmfm mycelium and absent on the mat+ wild-type mycelium. This observation indicates that the microconidia produced by the wild-type mat+ mycelium have fertilized the mat− Δmfm female organs, whereas the microconidia produced by the mat− Δmfm mycelium have not fertilized the mat+ female organs. The mat− Δmfm mutant that produced as many microconidia as the wild-type strain is thus inactive as male partner while it is active as female partner.
FIG. 4.
mat− Δmfm mutant displays complete male sterility and normal female fertility. Several implants of each parent were inoculated on one half-plate. After 4 days of growth, microconidia were dispersed over the surface of the cultures by adding 1 ml of sterile water and gently shaking the plate. The small black dots correspond to mature perithecia. In the mat+ × mat− cross, mature perithecia are present on both mat+ and mat− mycelia. In the mat+ × mat− Δmfm cross, mature perithecia are present only on the mat− Δmfm parent, indicating that it is female fertile but male sterile. In some cases, strong lines of mature perithecia are visible where the mat+ and mat− strains make contact.
By contrast, no fertility defect was observed when the Δmfm mutation was associated with mat+ (data not shown). The symmetric situation was obtained with the Δmfp mutation: microconidia produced by the mat+ Δmfp mutant were inactive in fertilization, whereas microconidia produced by the mat− Δmfp mutant were active. The Δmfp mutation did not alter female fertility. To more precisely determine the extent of the male defect in mat− Δmfm and mat+ Δmfp mutants, fertilization assays were performed with highly concentrated microconidial suspensions (106 to 108/ml) as described in Materials and Methods. Perithecia were never observed on wild-type cultures acting as a female.
Complementation assays were achieved to confirm that the male defect was due to the Δmfm and Δmfp null mutations. The mat− Δmfm mutant was transformed with the pCBMF-P3 plasmid carrying a wild-type copy of mfm and the hph gene conferring resistance to hygromycin (Hygr), whereas the mat+ Δmfp mutant was cotransformed with the pucBBmfp plasmid carrying a wild-type copy of mfp and pCB1004 carrying hph. In both transformation assays most of the Hygr transformants and cotransformants subjected to mating tests (19 of 20 and 24 of 40, respectively) were found to be efficient as male partner.
We then examined the ability of N. crassa pheromone precursors to complement the P. anserina null mutants. The mat+ Δmfp mutant was cotransformed with plasmids pBP14a containing mfa-1 from N. crassa and pCB1004 conferring resistance to hygromycin. Among the ten Hygr transformants analyzed, seven were found to be fertile as male partner in crosses with a mat− strain. Further analysis performed on two purified transformants showed that the mutant microconidia carrying the N. crassa mfa-1 gene were as efficient in crosses as the wild-type microconidia. The mat− Δmfm mutant was cotransformed with plasmids pCB1004 and pBP11 containing the ccg-4 gene from N. crassa. None of the 20 transformants analyzed could fertilize a mat+ strain.
In conclusion, deletion of the sequences encoding the pheromone precursors in P. anserina confers a unique phenotype in the appropriate mating type, disruption of fertilizing ability of microconidia.
Characterization of null mutants lacking both pheromone precursor genes.
The fully fertile mat− Δmfp and mat+ Δmfm strains were crossed to generate various combinations of mating type with functional and knockout pheromone genes. Due to the particular programming of ascus development in P. anserina (see Materials and Methods), the cross produced binucleate ascospores heterokaryotic for mating type, which were either homokaryotic or heterokaryotic for each Δmfm/mfm+ and Δmfp/mfp+ allelic pair. Consequently, all the associations presented in Table 1 could be recovered from the cross between mat− Δmfp and mat+ Δmfm strains (see Materials and Methods for details).
TABLE 1.
Fertilizing activity of heterokaryons carrying different combinations of mfm+/Δmfm, mfp+/Δmfp and mat+/mat− pairs of alleles
| Line no. | Genotype of nuclei in heterokaryotic cultures issued from binucleate ascospores | Fertilizationa
|
||
|---|---|---|---|---|
| mat− female partner | mat+ female partner | Self | ||
| 1 | mat+ mfp+mfm+/mat− mfp+mfm+ | + | + | + |
| 2 | mat+ mfp+mfm+/mat− Δmfp mfm+ | + | + | + |
| 3 | mat+ Δ mfp mfm+/mat− mfp+mfm+ | − | + | + |
| 4 | mat+ mfp+ Δmfm/mat− mfp+mfm+ | + | + | + |
| 5 | mat+ mfp+mfm+/mat− mfp+ Δmfm | + | − | + |
| 6 | mat+ mfp+ Δmfm/mat− Δmfp mfm+ | + | + | + |
| 7 | mat+ Δmfp mfm+/mat− mfp+ Δmfm | − | − | +/− |
| 8 | mat+ Δmfp Δmfm/mat− mfp+ Δmfm | − | − | +/− |
| 9 | mat+ Δmfp mfm+/mat− Δmfp Δmfm | − | − | +/− |
| 10 | mat+ Δmfp Δmfm/mat− Δmfp Δmfm | − | − | − |
+, optimal fertilization efficiency (≤5,000 perithecia produced on a petri dish); +/−, 1,000 to 2,000 perithecia produced on a petri dish; −, no perithecium produced.
We examined whether the male sterility caused by each null mutation could be complemented by the corresponding wild-type allele associated with the opposite mating type. The heterokaryotic strains were tested for their capacity to self-fertilize and to fertilize a mat+ and a mat− female partner. As reported in Table 1 line 10, the heterokaryotic strain mat− Δmfp Δmfm/mat+ Δmfp Δmfm devoid of pheromone precursor genes was unable to self-fertilize or to fertilize a mat+ and a mat− strain, although its female organs were functional, since they could be fertilized by wild-type microconidia of either mating type. The data reported in lines 3 and 7 demonstrated that the mat+ Δmfp microconidia produced by heterokaryotic mycelium containing mat− mfp+ nuclei were unable to fertilize mat− female organs. Similarly, the fertilizing activity of mat− Δmfm microconidia was not restored by the presence of mat+ mfm nuclei in the same mycelium (lines 5 and 9).
Surprisingly, although the heterokaryotic strains reported in lines 7, 8, and 9 produced inactive microconidia, these strains were able to self-fertilize. Self-fertilization was efficient since 1,500 to 2,000 perithecia were produced on a petri dish while a wild-type heterokaryotic strain (mat+ mfp+ mfm+/mat− mfp+ mfm+) produced 5,000 perithecia in the same experimental conditions. Perithecia issued from self-fertilization developed normally and gave abundant progeny displaying conventional segregation of the genetic markers (data not shown).
Self-fertilization occurring in the absence of functional microconidia could be attributable to hyphae which played the role of fertilizing cells and thus substituted for deficient microconidia. Self-fertilization observed on the heterokaryotic mat+ Δmfp Δmfm/mat− mfp+ Δmfm culture must be due to the mat+-specific pheromone, since the mat- specific mfm gene was absent in both parents. This result demonstrated that the mfp gene was transcribed in mat− nuclei. Similarly, self-fertilization observed on mat− Δmfp Δmfm/mat+ Δmfp mfm+ cultures demonstrated that the mfm gene was transcribed in mat+ nuclei. This conclusion is in contrast with data from Northern analysis: the mfp and mfm transcripts were not found in the nonspecific mating type context, although conditions were suitable for their detection, since RNAs were extracted from vegetative mycelium containing both hyphae and microconidia. We can thus postulate either that the transcription level was below the detection threshold or that the transcripts were unstable in the mycelium. Contrary to what was observed for heterokaryotic hyphae, the presence of nuclei containing the mfp+ or mfm+ gene associated with the nonspecific mating type did not rescue the mutant microconidia bearing the corresponding null allele, which remained inactive in fertilization. The data of this analysis show that although mfp and mfm are regulated by the mating types, they are constitutively transcribed at a low level in vegetative mycelium whatever its mating type.
Are pheromone precursor genes involved after fertilization?
If involvement of pheromone precursor genes in fertilization could be easily established by the phenotype of null mutants, it was more complicated to determine if these genes play a role in development of the fertilized female organ. The conceptually simple experiment of a cross involving homozygous null mfm and mfp mutants is not possible, since the absence of pheromones totally prohibits fertilization.
We initially tried to induce fertilization with a chemically synthesized QWCLRFVGQSCW polypeptide, whose sequence was deduced from the two dodecapeptides present in mfm sequence and assumed to correspond to the active pheromone. In S. cerevisiae, synthetic α-factor was shown to be biologically active in eliciting the initial biological responses of the pheromone signaling pathway in a cells (11). In P. anserina, we have not been successful in determining any significant activity of the synthetic peptide (for details see Materials and Methods). In particular the peptide pure or mixed to microconidia from a mat− Δmfm mutant was unable to induce formation of perithecia on a mat+ culture acting as a female.
We then attempted transient expression of pheromone precursor in Δmfp Δmfm double mutants by cloning mfp and mfm into the self-replicating pFAC1 vector (see Materials and Methods). This mitotically unstable plasmid (3) could be maintained vegetatively on selective medium containing hygromycin but was not transmitted through the cross. Consequently, the pheromone precursor gene should be lost once fertilization was achieved. If pheromone precursor genes are necessary after fertilization, perithecial development should be interrupted or at least severely altered.
To test this idea, pFACmfm and pFACmfp were introduced into mat− Δmfp Δmfm and mat+ Δmfp Δmfm recipients, respectively. In each assay, 10 Hygr transformants were subjected to genetic analysis. Plasmid instability was first assessed: when transformants were grown at least 5 days on medium without hygromycin, implants taken on the growth front and transferred on hygromycin medium were unable to regenerate, confirming loss of the plasmid. Fertility assays were then performed by growing each transformant separately on hygromycin medium and recovering microconidia to fertilize a Δmfp Δmfm strain of the opposite mating type. Fertilization was efficient, as evidenced by production of thousands of perithecia. The perithecia developed normally and gave an abundant progeny.
A sample of 20 asci was recovered from the progeny of three transformants bearing pFACmfm and three transformants bearing pFACmfp. Their genetic analysis confirmed the presence of Δmfp and Δmfm mutations and Mendelian segregation of mat+ and mat− mating types. All ascospores were sensitive to hygromycin (Hygs), which showed that the pFAC plasmid was not transmitted. However, the nonintegrative pFACmfp and pFACmfm plasmids must be present in the nucleus of microconidia to promote production of the pheromone, and it is not known when they were lost. This prompted us to examine expression of the pFAC-carried mfm gene in sexually differentiated female organs of a mat+ Δmfp Δmfm strain fertilized with microconidia recovered from a mat− Δmfp Δmfm transformant containing the pFACmfm plasmid.
Polyadenylated RNAs were extracted from 2-day-old perithecia and cDNA was produced as described in Materials and Methods. The mfm cDNA was not detected, while cDNA of SMR2, a mat− gene that is specifically transcribed in developing perithecia (13), was detected (data not shown). Therefore, homozygous crosses between Δmfp Δmfm strains undergo normal sexual development in the absence of postfertilization expression of the mfm precursor gene. A similar analysis was carried out on perithecia issued from the symmetric cross involving the pFACmfp plasmid. In that case, the mfp cDNA was detected, as it was detected when polyadenylated RNAs were extracted from wild-type perithecia issued from fertilization of a mat− culture by mat+ microconidia, which made the assay inconclusive. This interesting observation will be reexamined in the discussion.
A third strategy to reveal a possible function of pheromones after fertilization was based on complementation observed in heterokaryotic strains, presented in lines 8 and 9 of Table 1. Self-fertility of the mat+ Δmfp Δmfm/mat− mfp+ Δmfm heterokaryon indicated that the hyphae could act as a male donor of mat+ Δmfp Δmfm nuclei to mat− mfp+ Δmfm female organs present on the same culture. We took advantage of this complementation to fertilize a mat− Δmfp Δmfm culture used as female with fragmented mycelium from a mat+ Δmfp Δmfm/mat− mfp+ Δmfm heterokaryon (Materials and Methods and Fig. 5). Perithecia which developed normally and produced asci were obtained. Genetic analysis confirmed that these asci were issued from the expected cross, mat+ Δmfp Δmfm as male with mat− Δmfp Δmfm as female.
FIG. 5.
Schematic representation of the assay based on complementation within heterokaryotic fertilizing hyphae.
We can propose the interpretation illustrated in Fig. 5: production of the mat+ pheromone from the transcript synthesized in a mat− nucleus may allow attraction of a mat− Δmfp Δmfm trichogyne, which can fuse to the heterokaryotic mycelial fragments acting as fertilizing elements. The mat+ Δmfp Δmfm nucleus enters the trichogyne and triggers development of a fertile perithecium. Similarly, the reciprocal cross of a mat+ Δmfp Δmfm strain fertilized with fragmented mycelia from the mat− Δmfp Δmfm/mat+ Δmfp mfm+ heterokaryon was fertile. Therefore, if the fertilization defect is overcome, two mat+ and mat− nuclei lacking both mfp and mfm coding sequence can successfully go through the sexual process. These data, which confirm the data from the transient expression assay, demonstrate that the pheromones are not necessary after fertilization.
Deregulation of the mfp and mfm gene.
Replacement of the pheromone precursor gene 5′ noncoding sequence by the promoter of the glyceraldehyde-3-phosphate dehydrogenase (gpd) gene from P. anserina was expected to alleviate the requirement for mating type transcriptional control and to allow constitutive transcription. The transcriptionally deregulated pheromone genes would be useful to investigate the effect of this deregulation and, eventually, to find evidence for a posttranscriptional control.
The gpd::mfm and gpd::mfp fusions constructed in a plasmid carrying the hph gene conferring resistance to hygromycin (Materials and Methods) were first introduced into the relevant null mutant to determine if they complemented male sterility. In transformation of the mat− mfp+ Δmfm recipient, 20 Hygr transformants carrying the gpd::mfm fusion were analyzed. They were crossed as male partner with a mat+ wild-type strain in standard conditions. Nine transformants were found to be good male maters, indicating efficient complementation of the Δmfm null mutant by the gpd::mfm fusion. The restoration of male fertility by the gpd::mfm fusion was confirmed in the four transformants purified by sexual crosses. Standard fertility tests do not permit effective measurement of fertilization efficiency of microconidia, since more than 10,000 microconidia are produced by the strain used as male partner, while no more than 2,000 to 3,000 perithecia can develop on the female partner. An additional assay was carried out with a diluted microconidial suspension from one purified Δmfm (gpd::mfm) transformant (see Materials and Methods). Microconidia from the transformant were as efficient in fertilization as the wild-type microconidia (Table 2).
TABLE 2.
Complementation of the male defect of mat− Δmfm, mat+ Δmfp, and Δmat mutants by the gpd::mfm and gpd::mfp fusions
| Genotype of male parent | % of microconidia active in fertilizationa |
|---|---|
| mat− mfm+ | 27.5 ± 5 |
| mat− Δmfm | 0 |
| mat− Δmfm (gpd::mfm) | 21 ± 2.6 |
| Δmat (gpd::mfm) | 7* |
| mat+ mfp+ | 43* |
| mat+ Δmfp | 0 |
| mat+ Δmfp (gpd::mfp1-8) | 38.5 ± 0.7 |
| mat+ Δmfp (gpd::mfp1-2) | 1.1-1.2* |
| mat+ Δmfp (gpd::mfp2-4) | 0.7-1.6* |
| mat+ Δmfp (gpd::mfp2-6) | 0.9-2.7* |
| Δmat (gpd::mfp1-8) | 3* |
| Δmat (gpd::mfp1-2) | 0.02* |
| Δmat (gpd::mfp2-6) | 0.02* |
Calculated as number of perithecia produced on the wild-type culture used as the female/number of microconidia recovered from the male parent and used in fertilization (see Materials and Methods) times 100. Each value is the mean of three assays except *, values from one or two assays.
Two gpd::mfp fusions were successively constructed and tested (Materials and Methods). The gpd::mfp1 fusion encodes a polypeptide with an amino acid change: the proline at position 2 is now an alanine. The gpd::mfp2 fusion is a true promoter fusion, which does not change the coding sequence or the initiation codon context of the mfp coding sequence. The mat+ Δmfp mfm+ recipient was transformed with plasmids containing the gpd::mfp1 fusion. Twenty-three Hygr transformants were analyzed, among which 21 were found to be male-fertile in crosses with the mat− wild-type strain. Fertility was nevertheless low, since only 10 to 200 perithecia were observed on the female tester (several thousands in the control wild-type cross). Only one transformant (named gpd::mfp1-8) had the same male efficiency as the wild-type strain.
The same experiment was then performed with the gpd::mfp2 construct. Among the 26 Hygr transformants subjected to fertility tests, 6 were sterile and 20 displayed poor male fertility. Two transformants carrying the gpd::mfp1 fusion (gpd::mfp1-2 and gpd::mfp1-8) and two transformants carrying the gpd::mfp2 fusion (gpd::mfp2-4 and gpd::mfp2-6) were purified, and their fertilization efficiency was determined. The data reported in Table 2 confirmed the observations performed on primary transformants: the gpd::mfp fusion only partially complemented the fertilization defect of mat+ Δmfp microconidia except for the gpd::mfp1-8 transformant. The phenotypic homogeneity of all the analyzed transformants indicated that the low activity of the transgene was a property of the transgene itself rather than a cis effect of sequences adjacent to the ectopically integrated transgenes. Genetic analysis confirmed that the fusion had integrated ectopically at different genomic locations in gpd::mfp1-2, gpd::mfp2-4, and gpd::mfp2-6 transformants. In contrast, the gpd::mfp1-8 fusion integrated at the resident Δmfp locus.
The genomic structure resulting from the transgene integration was investigated by PCR analysis. First, a PCR was performed with a primer located in the gpd promoter at position −149 relative to the initiation codon and a primer located at position +418 within the mfp 3′ sequence. A product of the expected size was obtained, which demonstrated that the fusion had integrated without rearrangement of the fusion construct (data not shown). Second, primer pairs specific for the 5′ and 3′ junctions of mfp deletion were used. Only the sequence corresponding to the 5′ junction could be amplified with DNA from the transformant, whereas signals corresponding to 5′ and 3′ junctions were amplified with DNA from the mat+ Δmfp strain, used as a control. This analysis suggested that the fusion had integrated within the sequence downstream of the leu1 cassette in the gpd::mfp1-8 transformant.
Northern analysis was carried out with RNA extracted from the mat+ (gpd::mfp1-2) mycelium. Upon hybridization with a mfp sequence, a 0.5-kb transcript was detected, which was further identified as the gpd::mfp transcript, since it also hybridized with a gpd probe (data not shown). On RNA from the mat+ gpd::mfp1-8 mycelium, two bands were revealed with a mfp probe, a 0.5-kb and an additional 0.9-kb band. The two bands displayed similar intensity and were also revealed by a gpd probe. In both the gpd::mfp1-2 and gpd::mfp1-8 transformants, the presence of the incA mutation was not necessary for the detection of the gpd::mfp transcripts, which demonstrated efficient constitutive transcription of the fusion in mycelium.
Once the ability of the fusions to complement the knockout mutants was confirmed (Table 2), Δmat (gpd::mfm) and Δmat (gpd::mfp) strains were generated by sexual crosses with available Δmat strains carrying an ectopic copy of the mat+ or mat− sequence (Materials and Methods). Two strains carrying the gpd::mfm transgene at different integration sites were tested, both of which had recovered the ability to fertilize a mat+ female partner. Fertilization assays performed with one of the Δmat (gpd::mfm) strains showed that 7% of the microconidia were active in fertilization (Table 2). Characterization of Δmat (gpd::mfp) strains showed that the gpd::mfp fusion was able to restore male mating ability as a mat+ partner. This activity was nevertheless very low (Table 2). In Δmat (gpd::mfp1-2) and Δmat (gpd::mfp2-6) strains, less than 1 microconidium in 5,000 was efficient in fertilization of a mat− strain. On the other hand, the Δmat (gpd::mfp1-8) strain, in which fusion was integrated at the resident Δmfp locus, produced 3% active microconidia.
Finally, we examined the fertility of perithecia issued from fertilization with Δmat (gpd::mfm) or Δmat (gpd::mfp1-8) microconidia. Concentrated microconidial suspensions recovered from Δmat (gpd::mfm) and Δmat (gpd::mfp1-8) cultures were used to fertilize wild-type mat+ and mat− female cultures, respectively, and produced about 5,000 perithecia. The mat+ female tester contained an ectopic SMR1 copy, since this gene is absolutely required to obtain a progeny (2). The perithecia enlarged but remained smaller than the wild-type perithecia and produced rare ascospores: less than 1,000 scattered ascospores and no asci were ejected from perithecia, whereas in the same conditions a wild-type cross produced more than 100,000 asci. Fifty ascospores from the progeny of each cross were analyzed. They were all Hygr, indicating the presence of the plasmid that carried the fusion, and they exclusively displayed the mating activity of the fusion. These data showed that all ascospores had the genotype of the male parent, Δmat (gpd::mfm) or Δmat (gpd::mfp1-8), depending on the cross.
A more extensive analysis was performed by using as a female parent a strain containing the 136 mutation that prevents full spore pigmentation (136 ascospores are green instead of black) and allows easy discrimination between uniparental and biparental progeny (42). Only black ascospores were recovered, which thus contained nuclei of male origin. By multiplying the fertilization assays, we observed biparental asci at an extremely low frequency. Estimations indicated that about 1 perithecium out of 20,000 produced biparental progeny.
In summary, the data show that constitutive activation of the pheromone precursor genes rescues the male mating defect of a mating type-deficient strain and allows production of few progeny of paternal origin when the Δmat rescued strain is crossed as a male to a wild-type strain of opposite mating type as a female.
Characterization of the mat+ (gpd::mfm) and mat− (gpd::mfp) strains.
Since the gpd::mfm and gpd::mfp fusions were found to promote constitutive transcription of pheromone and to confer mating ability in a strain devoid of mating type, they were expected to induce selfing when introduced in a strain displaying the complementary mating type. Contrary to this prediction, the mat+ (gpd::mfm) and mat− (gpd::mfp) strains generated as described in Materials and Methods were found to be self-sterile (Table 3). Moreover, mat+ (gpd::mfm) microconidia were not active as the mat− partner, and similarly mat− (gpd::mfp1-8) microconidia were not active as the mat+ partner. This analysis showed that both fusions lost their activity when they were associated with the noncognate mating type. Although inactive, both fusions were nevertheless transcribed, as evidenced by Northern analysis (not shown) performed on RNA from mat+, mat−, and Δmat strains carrying a gpd::mfm fusion or the gpd::mfp1-8 fusion (in this last case, the 0.5-kb and 0.9-kb transcripts described above were detected). These data indicate that repression of gpd::mfm and gpd::mfp expression by mat+ and mat− acts at a posttranscriptional step.
TABLE 3.
Activity of the gpd::mfp and gpd::mfm fusions associated with wild-type and mutated mat− and mat+ mating typesa
| Genotype | Self-fertilization | Fertilization efficiency (no. of perithecia)
|
|
|---|---|---|---|
| mat− female culture | mat+ female culture | ||
| mat+ | 0 | >5,000 | 0 |
| mat− | 0 | 0 | >5,000 |
| Δmat | 0 | 0 | 0 |
| Δmat (gpd::mfm)b | 0 | 0 | >5,000 |
| mat+ (gpd::mfm)b | 0 | >5,000 | 0 |
| fpr1::ura5c | 30-100 | >5,000 | 20-40 |
| fpr1M112Ic | 30-100 | >5,000 | <10 |
| fpr1::ura5 (gpd::mfm)c | >5,000 | >5,000 | >5,000 |
| fpr1M112I (gpd::mfm)c | >5,000 | >5,000 | >5,000 |
| Δmat (gpd::mfp1-8) | 0 | >5,000 | 0 |
| mat− (gpd::mfp1-8) | 0 | 0 | >5,000 |
| fmr1::ura5c | 50-60 | 50-60 | >5,000 |
| fmr1::ura5 (gpd::mfp1-8) | >5,000 | >5,000 | >5,000 |
| smr2::ura5c | <10 | 0 | >5,000 |
| smr2::ura5 (gpd::mfp1-8) | 100 | 20-40 | >5,000 |
Fertilization assays were carried out with at least 106 microconidia to get the maximum number of perithecia on the female culture even with the Δmat (gpd::mfm) and Δmat (gpd::mfp1-8) strains, which only produced 7 and 3% active microconidia, respectively. (Table 2). Three assays were carried out for each strain.
Two transformants carrying the fusion at different genomic locations were analyzed.
Mutations in mating type genes were previously shown to confer the ability to self-fertilize and to fertilize a strain of the same mating type at very low efficiency while keeping the ability to fertilize normally a strain of the opposite mating type (2).
To identify which of the mat genes was responsible for this repression, the gpd::mfp1-8 construct was associated by crosses with fmr1::ura5 and smr2::ura5 mutations (Materials and Methods), and the generated strains were tested for fertility. The third mat gene, SMR1, was not tested, since it is not involved in fertilization (1). The data presented in Table 3 showed that the fmr1::ura5 mutation totally alleviated repression, which identified FMR1 as the gene mainly involved in the posttranscriptional repression of the gpd::mfp transgene. It can be seen that the presence of the smr2::ura5 mutation permitted residual fertilization ability. This indicates that full repression of the gpd::mfp1-8 expression by FMR1 requires a functional SMR2 gene. The mat+ idiomorph contains a unique functional gene, FPR1, flanked by a DNA sequence with no apparent function (18). To confirm the involvement of FPR1 in repression of gpd::mfm expression, we determined the effect of two alleles, missense (fpr1M112I) and a disruption (fpr1::ura5) (Table 3). The data demonstrated that both mutations alleviated repression (Table 3).
In conclusion, the study of transformants displaying constitutive expression of the pheromone precursor genes has revealed that these genes are repressed posttranscriptionally by the noncognate mating type.
DISCUSSION
In P. anserina, mating types control a recognition process between male and female cells during fertilization, and between paternal and maternal nuclei during development of the fertilized female organs. To investigate the role of the pheromone response pathway in these recognition processes and more generally in the physiology of the fungus, we have cloned the pheromone precursor genes, taking advantage of the previous characterization of the homologous genes in N. crassa and C. parasitica. Two genes, mfp (mating factor plus) and mfm (mating factor minus), have been identified; they are single-copy genes in P. anserina. DNA sequence analysis indicated that the mfp gene encodes a putative 24-amino-acid polypeptide displaying high similarity with the polypeptides encoded by the homologous genes in N. crassa and Sordaria macrospora (Fig. 1) and ending with CAAX, a prenylation signal common to all known fungal lipopeptide pheromones. The mfm gene encodes a putative polypeptide of 221 amino acids containing two repeats of a dodecapeptide that could correspond to the mature pheromone, since putative protease-processing sites border it (Fig. 2). However, we have not been able to demonstrate that the corresponding peptide as chemically synthesized was biologically active.
Deletion of either of the pheromone precursor genes leads exclusively to a mating type-specific male sexual defect without impairing other functions.
Null mfp and mfm mutants were created by targeted gene replacement. The mat+ Δmfp and mat− Δmfm null mutants produced microconidia deficient in fertilization of female organs of the opposite mating type. Male sterility was shown to be rescued by introducing a wild-type transgenic copy of mfp or mfm, respectively. In contrast, male fertility was not affected in mat− Δmfp and mat+ Δmfm strains, and deletion of either or both pheromone precursor genes did not affect growth, mycelial morphology, differentiation and function of female organs, or development of fertilized female organs. This lack of phenotype other than male sterility is in contrast with the pleiotropic phenotype reported for the presumptive mfa-1 null mutants, the gene homologous to mfp in N. crassa (28).
If the mat a-specific male sexual defect of these mutants confirmed the role of mfa-1 in promoting fertilization, the associated female and vegetative growth defects observed in mat a and mat A backgrounds suggested a broader function. The mfa-1 mutants exhibited reduced vegetative growth on minimal medium, differentiated 1,000 times fewer protoperithecia than a wild-type strain, and produced only a small number of viable ascospores in crosses with the wild-type acting as the male. When two mfa-1 mutant strains were crossed, the rare perithecia that developed were barren. This phenotype has led the authors to propose that the MFA1 pheromone could play the role of a conglutinin allowing maternal vegetative hyphae adjacent to female organs to adhere to one another.
In P. anserina, the absence of a female defect in mutants devoid of genes encoding propheromones totally excludes the involvement of either pheromone in formation of the perithecial wall tissues. Similarly, the absence of a vegetative phenotype in these mutants excludes the involvement of pheromones in filamentous growth. It must be noted that the mfp null mutant of P. anserina was screened molecularly for deletion of the mfp coding sequence. In contrast, N. crassa mfa-1 mutants were obtained by the RIP approach and selected because of poor vegetative growth (28). The fact that another gene has been altered during this process is not totally excluded, since complementation experiments with a cosmid did not allow transformants to recover a wild-type vegetative phenotype. Alternatively, even if all RIP mutations are really restricted to mfa-1 sequence, they affect both the coding sequence and the 3′ UTR of the gene.
Without determining the phenotype resulting from mutations restricted to the coding sequence of mfa-1, it is difficult to interpret functional differences between mfa-1 in N. crassa and mfp in P. anserina. Nevertheless, preliminary unpublished data from the Podospora Genome Project achieved in collaboration with the French National Sequencing Center (Genoscope) indicate that these two genes share several interesting characteristics. Like mfa-1, the mfp transcript has a long 3′ UTR region of 750 nucleotides. Analyses of unsubtracted cDNA libraries show that mfp clones account for 0.9% of the perithecial library (90 of 10,000 isolates), indicating that mfp is well expressed during female organ development, as it has been observed for mfa-1. However, in contrast to mfa-1, the mfp transcript was not found in mycelial cDNA library obtained on minimal media (0 in 17,000 isolates). Further investigation is necessary to assess the functional significance of the 3′-UTR of mfp and mfa-1. For the moment, it is clear that the primary role of mfp in P. anserina and mfa-1 in N. crassa is to allow fertilization in a mating type-specific manner. The genes are functional orthologs, as shown by the ability of mfa-1 to complement the male defect of the mat+ Δmfp mutant of P. anserina. Efficient complementation indicates that the precursor encoded by mfa-1 is correctly processed in P. anserina. In contrast, the introduction of the N. crassa ccg-4 gene encoding the homolog of mfm failed to complement the mfm deletion mutant. We have not determined at what level ccg4 expression is inefficient in P. anserina.
Transcription of each pheromone precursor gene that operates at a basal level in vegetative hyphae is specifically activated by the cognate mating type in microconidia.
The mat+-specific male defect of the Δmfp null mutant and the mat-specific male defect of the Δmfm null mutant indicated that each pheromone precursor gene is expressed in a mating type-dependent manner. In agreement with that observation, the mfp transcript was only detected in mat+ strains, whereas the mfm transcript was only detected in mat− strains. Transcripts were detected by Northern blots only in RNAs extracted from hyperconidial strains, which demonstrated that transcription of pheromone precursor genes was activated in microconidia. Fertilization by fragmented mycelium is currently used in P. anserina and provides evidence that the mfm and mfp genes are expressed in mat− and mat+ mycelium, respectively (23). Failure to detect mfm and mfp transcripts by Northern blot in RNA extracted from wild-type mycelia indicated that this expression occurs at a very low level in vegetative cells. However, the self-fertilization of mat+ Δmfp Δmfm/mat− mfp+ Δmfm and mat+ Δmfp mfm+/mat− Δmfp Δmfm heterokaryons (Table 1) indicated that the mfp and mfm genes were also transcribed in mat− and mat+ vegetative nuclei, respectively.
We do not know whether the posttranscription steps required for the expression of the mat+ pheromone is mediated by the genes present in the mat− nucleus or in neighboring mat+ nuclei. A similar question remains unresolved as to the expression of the mat− pheromone. In contrast to vegetative hyphae, both mat+ and mat− microconidia produced by the same heterokaryons were inactive in fertilization. This different behavior of hyphae and microconidia may result from the fact that vegetative hyphae are plurinucleate whereas microconidia are uninucleate. Consequently, the pheromone defect of a microconidium cannot be complemented by a neighboring nucleus. In conclusion, our data suggest that both pheromone precursor genes are expressed at a low level in vegetative hyphae and allow them to act as fertilizing elements. In microconidia, each mating type activates transcription of its cognate pheromone precursor gene, i.e., FPR1 activates transcription of mfp in mat+ microconidia, while FMR1 activates transcription of mfm in mat− microconidia.
Pheromone precursor genes are not required for development of the fertilized female organs.
We addressed the question of whether pheromones were also required for postfertilization events by examining fertility of crosses between mat+ Δmfp Δmfm and mat− Δmfp Δmfm mutants. Complementation of fertilization deficiency was achieved either by introducing into the strain acting as male the pFAC unstable self-replicating plasmid carrying the pheromone precursor gene relevant to the mating type of the strain or more conclusively by taking advantage of internuclear complementation observed within heterokaryotic hyphae. Active fertilizing elements (microconidia or fragmented mycelia) able to transmit Δmfp Δmfm nuclei to Δmfp Δmfm female organs of the opposite mating type were recovered in both assays. If the pheromone precursor genes were required after fertilization, development of the fertilized protoperithecia should have been severely impaired. On the contrary, the crosses were found to be fully fertile, which demonstrates that perithecial development does not require the presence of the mfm and mfp genes. The pheromone precursor coding sequences have a role only at fertilization: they allow the male element to be sensed by the female organ. Consequently, the present study invalidates definitively the hypothesis by which recognition between mat+ and mat− nuclei during the formation of dikaryotic ascogenous hyphae would be mediated through a pheromone-receptor interaction (13, 16).
Expression of mfp and mfm is repressed by mat− and mat+, respectively.
To subtract the mfp and mfm genes from mating type control, we replaced their 5′ noncoding sequence with the constitutive gpd promoter from P. anserina. The activity of the fusions was inferred from efficient complementation of the male defect in the Δmfp and Δmfm null mutants (Table 2). Each fusion restored male fertility when it was introduced in the Δmat mating-deficient mutant (Table 2). This shows that the gpd::mfp and gpd::mfm transgenes are transcribed and expressed constitutively. Mating types are thus unnecessary for propheromone maturation and pheromone secretion. The gpd::mfm fusion was found to confer higher fertilizing activity than the gpd::mfp fusion, except for one transformant in which the gpd::mfp fusion had integrated at the Δmfp locus (gpd::mfp1-8).
We suppose that the integration of gpd::mfp1-8 occurred near a sequence required for the efficient expression of this gene. As the integration occurred in the sequence downstream of the resident mfp coding sequence, we can predict that this regulatory sequence belongs to the 3′ noncoding region of mfp. Mutations in the 3′ UTR of the mfa-1 transcript of N. crassa also result in a partial defect in male fertility which can be attributed to a decrease in pheromone expression (28). Both results suggest that the mfp and mfa-1 genes contain a regulatory element in the 3′ noncoding region that is required for optimal expression of the genes.
Analysis of two mat+ (gpd::mfm) strains and of the mat− (gpd::mfp1-8) strain showed that the fusions did not confer corresponding male fertility when associated with the noncognate mating type: the mat+ (gpd::mfm) and mat− (gpd::mfp1-8) strains were not able to self-fertilize or to fertilize a tester strain of similar mating type (Table 3). However, both fusions were expressed in the Δmat genetic background. These results imply that the expression of each fusion is repressed at the posttranscriptional level by the opposite mating type. In agreement with this assumption, the transcripts of the gpd::mfp1-8 and gpd::mfm fusions were detected in mat+, mat−, and Δmat strains (data not shown). By examining the activity of the gpd::mfp and gpd::mfm fusions in strains carrying mutations in the mat− and mat+ mating type genes, we provided evidence that the mat− FMR1 and SMR2 genes repressed expression of the gpd::mfp fusion, while the unique mat+ FPR1 gene repressed expression of the gpd::mfm fusion (Table 3). The fpr1 and fmr1 mutants used in our study are altered in their C-terminal end (Table 3). This part of the protein, which is dispensable for fertilization control (18), is thus required to repress expression of gpd::mfp and gpd::mfm fusions.
Our data help to specify the model previously proposed to explain that fpr1, fmr1, and smr2 mutants were able of self-fertilizing and of fertilizing a strain of the same mating type at very low efficiency (2). According to this model, the wild-type FPR1 protein would act not only as an activator of mat+ fertilization genes but also as a direct or indirect negative regulator of mat− fertilization genes. Reciprocally, FMR1, an activator of mat− fertilization genes, would also act as a repressor of mat+ fertilization genes. Although SMR2 was not required for fertilizing a mat+ partner (17), it would act as a repressor of mat+ fertilization genes.
We can now propose the regulation scheme presented in Fig. 6 for the pheromone precursor genes, the first fertilization genes studied in P. anserina. Two points can be emphasized. First, repression by the FPR1 and FMR1 proteins occurs posttranscriptionally. Second, optimal posttranscriptional repression of mfp by FMR1 requires a functional SMR2 gene, since expression of the gpd::mfp1-8 fusion was not totally repressed when SMR2 was mutated. The fact that SMR2 is necessary to optimize repression of mfp by FMR1 may explain the finding that an smr2::ura5 mutant self-fertilizes and fertilizes a mat− strain at a low efficiency (2).
FIG. 6.
Model for regulation of pheromone precursor gene expression by mating type genes. Activation is symbolized by an arrow and repression by a bar.
The regulation model proposed in Fig. 6 relies mainly on genetic data and must now be ascertained by molecular analysis. In particular, the binding of FPR1 to the 5′ noncoding region of mfp and that of FMR1 to the 5′ noncoding region of mfm has not been proven at the molecular level. The recent availability of the P. anserina genome opens a new field of investigation. For example, we will be able through BLAST searches to look for homologs of all genes involved in maturation and production of active pheromones known in other fungi. All these genes are possible targets of mating type-controlled repression described in this paper. Comparison of their transcriptional pattern in mat+ and mat− genetic backgrounds should provide useful information if repression operates at a transcriptional level.
The pheromone precursor genes are the first target genes of mating proteins characterized in P. anserina. They are involved in the cell-cell recognition allowing fertilization. The data reported in this paper demonstrate conclusively that the other event controlled by the mating types, recognition between mat+ and mat− nuclei before formation of dikaryotic ascogenous hyphae, does not rely on a pheromone signaling pathway. The two genes necessary for fertilization, FPR1 and FMR1, in association with SMR2 control internuclear recognition, probably by activating a new set of target genes. We need to identify these target genes in order to understand the molecular mechanisms underlying the recognition between nuclei of opposite mating types. Possible molecular approaches will be facilitated by the availability of the complete DNA sequence of P. anserina.
Acknowledgments
We are very grateful to Dan Ebbole and Deb Bell-Pedersen for giving plasmids pB14a and pBP11 and for sharing the sequence data of mfa-1 and ccg-4 with us in 1999, well prior to publication. We thank Michelle Dequard Chablat and Corinne Jamet-Vierny for providing a cosmid library of P. anserina DNA and Southern blots of DNA pools from this library, Christian Barreau for the pFAC plasmid, and H.-D. Osiewacz for the pRP81-1 plasmid. We are grateful to all the members of the RSTD group for helpful comments on the manuscript.
This research was supported by a grant from the Action Concertée Incitative of the French Ministère de la Recherche.
REFERENCES
- 1.Arnaise, S., R. Debuchy, and M. Picard. 1997. What is a bona fide mating type gene? Internuclear complementation of mat mutants in Podospora anserina. Mol. Gen. Genet. 256:169-178. [DOI] [PubMed] [Google Scholar]
- 2.Arnaise, S., D. Zickler, S. Le Bilcot, C. Poisier, and R. Debuchy. 2001. Mutations in mating type genes of the heterothallic fungus Podospora anserina lead to self-fertility. Genetics 159:545-556. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Barreau, C., M. Iskandar, B. Turcq, and J. P. Javerzat. 1998. Use of a linear plasmid containing telomeres as an efficient vector for direct cloning in the filamentous fungus Podospora anserina. Fungal Genet. Biol. 25:22-30. [DOI] [PubMed] [Google Scholar]
- 4.Bernet, J. 1986. Podospora anserina mutations reducing cell survival following nutrient exhaustion. Curr. Microbiol. 14:133-136. [Google Scholar]
- 5.Bistis, G. N. 1981. Chemotropic interactions between trichogynes and conidia of opposite mating type in Neurospora crassa. Mycologia 73:959-975. [Google Scholar]
- 6.Bistis, G. N. 1998. Physiological heterothallism and Sexuality in euascomycetes: a partial history. Fungal Genet. Biol. 23:213-222. [DOI] [PubMed] [Google Scholar]
- 7.Bobrowicz, P., R. Pawlak, A. Correa, D. Bell-Pedersen, and D. J. Ebbole. 2002. The Neurospora crassa pheromone precursor genes are regulated by the mating type locus and the circadian clock. Mol. Microbiol. 45:795-804. [DOI] [PubMed] [Google Scholar]
- 8.Brown, A. J., and L. A. Casselton. 2001. Mating in mushrooms: increasing the chances but prolonging the affair. Trends Genet. 17:393-400. [DOI] [PubMed] [Google Scholar]
- 9.Carroll, A. M., J. A. Sweigard, and B. Valent. 1994. Improved vectors for selecting resistance to hygromycin. Fungal Genet. Newsl. 41:22. [Google Scholar]
- 10.Chomczynski, P., and N. Sacchi. 1987. Single-step method of RNA isolation by acid guanidinium thiocyanate-phenol-chloroform extraction. Anal. Biochem. 162:156-159. [DOI] [PubMed] [Google Scholar]
- 11.Ciejek, E., J. Thorner, and M. Geier. 1977. Solid phase peptide synthesis of alpha-factor, a yeast mating pheromone. Biochem. Biophys. Res. Commun. 78:952-961. [DOI] [PubMed] [Google Scholar]
- 12.Coppin, E. 2002. The fle1 gene encoding a C2H2 zinc finger protein co-ordinates male and female sexual differentiation in Podospora anserina. Mol. Microbiol. 43:1255-1268. [DOI] [PubMed] [Google Scholar]
- 13.Coppin, E., and R. Debuchy. 2000. Coexpression of the mating type genes involved in internuclear recognition is lethal in Podospora anserina. Genetics 155:657-669. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Coppin, E., R. Debuchy, S. Arnaise, and M. Picard. 1997. Mating types and sexual development in filamentous ascomycetes. Microbiol. Mol. Biol. Rev. 61:411-428. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Coppin-Raynal, E., M. Picard, and S. Arnaise. 1989. Transformation by integration in Podospora anserina. III. Replacement of a chromosome segment by a two-step process. Mol. Gen. Genet. 219:270-276. [DOI] [PubMed] [Google Scholar]
- 16.Debuchy, R. 1999. Internuclear recognition: a possible connection between euascomycetes and homobasidiomycetes. Fungal Genet. Biol. 27:218-223. [DOI] [PubMed] [Google Scholar]
- 17.Debuchy, R., S. Arnaise, and G. Lecellier. 1993. The mat− allele of Podospora anserina contains three regulatory genes required for the development of fertilized female organs. Mol. Gen. Genet. 241:667-673. [DOI] [PubMed] [Google Scholar]
- 18.Debuchy, R., and E. Coppin. 1992. The mating types of Podospora anserina: functional analysis and sequence of the fertilization domains. Mol. Gen. Genet. 233:113-121. [DOI] [PubMed] [Google Scholar]
- 19.Dequard, C. M., and C. H. Sellem. 1994. The S12 ribosomal protein of Podospora anserina belongs to the S19 bacterial family and controls the mitochondrial genome integrity through cytoplasmic translation. J. Biol. Chem. 269:14951-14956. [PubMed] [Google Scholar]
- 20.Dequard-Chablat, M., and C. Alland. 2002. Two copies of mthmg1, encoding a novel mitochondrial HMG-like protein, delay accumulation of mitochondrial DNA deletions in Podospora anserina. Eukaryot. Cell. 1:503-513. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Elion, E. A. 2000. Pheromone response, mating and cell biology. Curr. Opin. Microbiol. 3:573-581. [DOI] [PubMed] [Google Scholar]
- 22.Esser, K. 1974. Podospora anserina, p. 531-551. In R. C. King (ed.), Handbook of genetics, vol. 1. Plenum, New York, N.Y. [Google Scholar]
- 23.Graia, F., V. Berteaux-Lecellier, D. Zickler, and M. Picard. 2000. ami1, an orthologue of the Aspergillus nidulans apsA gene, is involved in nuclear migration events throughout the life cycle of Podospora anserina. Genetics 155:633-646. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Hanahan, D. 1983. Studies on transformation of Escherichia coli with plasmids. J. Mol. Biol. 166:557-580. [DOI] [PubMed] [Google Scholar]
- 25.Hartmann, H. A., R. Kahmann, and M. Bolker. 1996. The pheromone response factor coordinates filamentous growth and pathogenicity in Ustilago maydis. EMBO J. 15:1632-1641. [PMC free article] [PubMed] [Google Scholar]
- 26.Hirsch, J. P., and F. R. Cross. 1992. Pheromone response in yeast. Bioessays 14:367-373. [DOI] [PubMed] [Google Scholar]
- 27.Kim, H., and K. A. Borkovich. 2004. A pheromone receptor gene, pre-1, is essential for mating type-specific directional growth and fusion of trichogynes and female fertility in Neurospora crassa. Mol. Microbiol. 52:1781-1798. [DOI] [PubMed] [Google Scholar]
- 28.Kim, H., R. L. Metzenberg, and M. A. Nelson. 2002. Multiple functions of mfa-1, a putative pheromone precursor gene of Neurospora crassa. Eukaryot. Cell. 1:987-999. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Lecellier, G., and P. Silar. 1994. Rapid methods for nucleic acids extraction from Petri dish-grown mycelia. Curr. Genet. 25:122-123. [DOI] [PubMed] [Google Scholar]
- 30.Marcou, D., and M. Picard. 1967. On the structure of a complex locus in Podospora anserina: the relations between the genetic map and the complementation map. C R Acad. Sci. Hebd. Seances Acad. Sci. D 265:1962-1965. [PubMed] [Google Scholar]
- 31.Perkins, D. D. 2003. A fratricidal fungal prion. Proc. Natl. Acad. Sci. USA 100:6292-6294. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Picard, M., R. Debuchy, and E. Coppin. 1991. Cloning the mating types of the heterothallic fungus Podospora anserina: developmental features of haploid transformants carrying both mating types. Genetics 128:539-547. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Poggeler, S. 2000. Two pheromone precursor genes are transcriptionally expressed in the homothallic ascomycete Sordaria macrospora. Curr. Genet. 37:403-411. [DOI] [PubMed] [Google Scholar]
- 34.Ridder, R., and H. D. Osiewacz. 1992. Sequence analysis of the gene coding for glyceraldehyde-3-phosphate dehydrogenase (gpd) of Podospora anserina: use of homologous regulatory sequences to improve transformation efficiency. Curr. Genet. 21:207-213. [DOI] [PubMed] [Google Scholar]
- 35.Rizet, G., and C. Engelmann. 1949. Contribution à l'étude génétique d'un ascomycète tétrasporé: Podospora. (Ces) Rehm. Rev. Cytol. 11:201-304. [Google Scholar]
- 36.Shen, W. C., P. Bobrowicz, and D. J. Ebbole. 1999. Isolation of pheromone precursor genes of Magnaporthe grisea. Fungal Genet. Biol. 27:253-263. [DOI] [PubMed] [Google Scholar]
- 37.Silar, P., C. Barreau, R. Debuchy, S. Kicka, B. Turcq, A. Sainsard-Chanet, C. H. Sellem, A. Billault, L. Cattolico, S. Duprat, and J. Weissenbach. 2003. Characterization of the genomic organization of the region bordering the centromere of chromosome V of Podospora anserina by direct sequencing. Fungal Genet. Biol. 39:250-263. [DOI] [PubMed] [Google Scholar]
- 38.Turina, M., A. Prodi, and N. K. Alfen. 2003. Role of the Mf1-1 pheromone precursor gene of the filamentous ascomycete Cryphonectria parasitica. Fungal Genet. Biol. 40:242-251. [DOI] [PubMed] [Google Scholar]
- 39.Willer, M., L. Hoffmann, U. Styrkarsdottir, R. Egel, J. Davey, and O. Nielsen. 1995. Two-step activation of meiosis by the mat1 locus in Schizosaccharomyces pombe. Mol. Cell. Biol. 15:4964-4970. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Zhang, L., R. A. Baasiri, and N. K. Van Alfen. 1998. Viral repression of fungal pheromone precursor gene expression. Mol. Cell. Biol. 18:953-959. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.Zhang, L., A. C. Churchill, P. Kazmierczak, D. H. Kim, and N. K. Van Alfen. 1993. Hypovirulence-associated traits induced by a mycovirus of Cryphonectria parasitica are mimicked by targeted inactivation of a host gene. Mol. Cell. Biol. 13:7782-7792. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42.Zickler, D., S. Arnaise, E. Coppin, R. Debuchy, and M. Picard. 1995. Altered mating type identity in the fungus Podospora anserina leads to selfish nuclei, uniparental progeny, and haploid meiosis. Genetics 140:493-503. [DOI] [PMC free article] [PubMed] [Google Scholar]