Genetics, Vol. 149, 915-926, June 1998, Copyright © 1998

The mod-A Suppressor of Nonallelic Heterokaryon Incompatibility in Podospora anserina Encodes a Proline-Rich Polypeptide Involved in Female Organ Formation

Christian Barreaua, Maya Iskandara, Gabriel Loubradoua, Véronique Levalloisa, and Joël Béguereta
a Institut de Biochimie et de Génétique Cellulaires, Centre National de la Recherche Scientifique, UPR 9026, 33077 Bordeaux, France

Corresponding author: Christian Barreau, Laboratoire de Génétique Moléculaire des Champignons Filamenteux, Institut de Biochimie et de Génétique Cellulaires, UPR CNRS 9026, 1, rue Camille Saint-Saëns, 33077 Bordeaux, France, christian.barreau{at}ibgc.u-bordeaux2.fr (E-mail).

Communicating editor: P. J. PUKKILA


*  ABSTRACT
*TOP
 *ABSTRACT
*MATERIALS AND METHODS
 *RESULTS
 *DISCUSSION
 *LITERATURE CITED

Vegetative incompatibility in fungi results from the control of heterokaryon formation by the genes present at het loci. Coexpression of antagonistic het genes in the same hyphae leads to a lethal process. In Podospora anserina, self-incompatible strains containing nonallelic incompatible genes in the same nucleus are inviable as the result of a growth arrest and a lytic process. Mutations in suppressor genes (mod genes) can restore the viability. These mod mutations also interfere with developmental processes, which suggests common steps between the incompatibility reaction and cellular differentiation. The mod-A locus, responsible for growth arrest in the self-incompatible strains, is also involved in the control of the development of female organs. The mod-A gene was isolated. An open reading frame 687 amino acids long was identified. The MOD-A-encoded polypeptide is rich in proline residues, which are clustered in a domain containing a motif that displays similarity to SH3-binding motifs, which are known to be involved in protein-protein interactions. Construction of a strain deleted for mod-A confirmed that the product of this gene involved in differentiation is a key regulator of growth arrest associated with vegetative incompatibility.

IN filamentous fungi, fusion between genetically distinct individuals is restricted as the result of somatic or vegetative incompatibility. Vegetative incompatibility is widespread in these organisms and is the consequence of the incapacity to form viable heterokaryotic cells following anastomosis between filaments of unlike genotype. The phenomenon is controlled by differences at specific loci named het loci, which have been genetically characterized in some filamentous fungi such as Neurospora crassa, Podospora anserina, Aspergillus nidulans, and Cryphonectria parasitica (for reviews see GLASS and KULDAU 1992 Down; BEGUERET et al. 1994). In P. anserina, the antagonist het genes may be alternate alleles at a single locus or nonallelic genes from independent loci (BERNET 1967 Down). In the latter case, different alleles can be observed at each nonallelic het locus. The formation of heterokaryons leads to compatible or incompatible genetic combinations of het genes. Heterokaryotic cells resulting from an incompatible combination of allelic genes (allelic incompatible interaction) or of nonallelic het genes (nonallelic interaction) undergo a lytic process leading to cell death. The accumulation of dying cells in the contact zone of two incompatible strains results in the formation of an abnormal contact named "barrage" (RIZET 1952 Down).

Three nonallelic systems—het-c/het-e, het-c/het-d, and het-r/het-v—have been characterized in P. anserina (BERNET 1967 Down, BERNET 1992 Down). Genetic crosses between strains bearing incompatible genes at these loci lead to offspring containing incompatible combinations of the het genes in the same nucleus. Uninucleated spores of this genotype can germinate and produce a homokaryotic mycelium, which, as the first consequence of incompatibility, stops growing a few hours after germination. The second consequence is a generalized lytic reaction of all the cells in the thallus, leading to the death of such self-incompatible (SI) strains (BERNET et al. 1973 Down). However, these strains can "escape" from self-incompatibility as seen by the emergence of growing sectors. The mutations most often occur in the het genes, leading to a neutral interaction. However, strains that display a new phenotype, called MSI (modified self-incompatible), have been identified (BELCOUR and BERNET 1969 Down; BERNET et al. 1973 Down). Although the mycelium has few aerial filaments, the apical growth of these strains is restored. However, the lytic defect is only partly suppressed. Most of the MSI strains result from modifier (mod) mutations occurring in the unlinked mod-A locus (BELCOUR and BERNET 1969 Down). All mutations identified at this locus are recessive. The mod-A1 mutation has no effect on allelic interactions but suppresses self-incompatibility resulting from coexpression of the het genes from the three nonallelic het-c/het-e, het-c/het-d, and het-r/het-v systems (BELCOUR and BERNET 1969 Down; BERNET et al. 1973 Down). No obvious phenotypical trait has been observed for mod-A1 mutants in a compatible background.

The MSI strains exhibit cold sensitivity (BERNET et al. 1973 Down), which can be suppressed by mutations in the independent mod-B locus, where the mod-B1 mutation was previously identified (BERNET 1971 Down). In MSI mod-A1 mod-B1 double mutant strains, lysis is suppressed (BERNET 1971 Down; LABARERE 1973). Although no mutant phenotype could be observed for the mod-B1 mutations in a compatible background, the mod-A1 mod-B1 double mutant exhibited a female sterility phenotype (BERNET 1971 Down; BOUCHERIE and BERNET 1974 Down). This phenotype was the first evidence for a link between genes involved in incompatibility and development. However, the level of sterility is heavily dependent on growth conditions. This suggests that mod-A and mod-B are not strictly required for the differentiation processes but that they become essential for the successful development of the female organs in unfavorable conditions (BOUCHERIE and BERNET 1980 Down; BERNET 1992 Down).

Incompatibility genes have been studied at the molecular level in N. crassa (GLASS et al. 1990 Down; PHILLEY and STABEN 1994 Down; SAUPE et al. 1996A Down; SAUPE et al. 1996B Down) and P. anserina (BEGUERET et al. 1994; SAUPE et al. 1995 Down; ESPAGNE et al. 1997 Down). Suppressors of heterokaryon incompatibility have been reported for N. crassa (NEWMEYER 1970 Down; JACOBSON 1992 Down; VELLANI et al. 1994 Down; ARGANOZA et al. 1994 Down), but they have not yet been characterized at the molecular level. The mechanism of the lethality resulting from the interaction between het genes is still unknown. In addition, its relation to developmental processes remains unclear. Molecular characterization of the mod-A1 and mod-B1 modifier genes that are able to suppress incompatible interactions resulting from the three nonallelic systems in P. anserina and are involved in growth control and in female organ differentiation may help explain the link between incompatibility and development. In this study, we report the isolation of the mod-A gene. No similarity with known proteins was observed for the mod-A-encoded polypeptide. However, a proline-rich region with a potential SH3-binding domain was observed, which may suggest an involvement in protein-protein interactions.


*  MATERIALS AND METHODS
*TOP
 *ABSTRACT
 *MATERIALS AND METHODS
*RESULTS
 *DISCUSSION
 *LITERATURE CITED

Fungal strains:
P. anserina is a pseudohomothallic ascomycete. Its life cycle and the general methods used for genetic analysis have been described elsewhere (RIZET and ENGELMAN 1949 Down; ESSER 1974 Down). The specific heterokaryon-compatibility pattern of a strain can be determined by confrontation on cornmeal agar medium with appropriate tester strains of known genotype. When two strains are incompatible, a dense and unpigmented mycelial pad called a barrage is formed in the region of contact between thalli, where the mycelia form heterokaryons (RIZET 1952 Down). Unless specified, the strains with the mod-A and mod-B mutations used in this study are all derived from the "F" geographical race (BERNET 1967 Down) bearing the het-c3 allele at the het-c locus. This gene is incompatible with the het-e2 allele from the antagonistic het-e locus (SAUPE et al. 1994 Down). In the crosses involving the mod-A1 mod-B1 double mutant, the strains were used as male. Fertility tests were performed with strains grown in the dark for 5 days at 26° on cornmeal agar medium or on synthetic medium (LOUBRADOU et al. 1996 Down). The plates were then transferred to the light; 3 days later, they were overlaid with a conidial suspension collected on a strain of the opposite mating type grown in the same conditions. The liquid was discarded, and the plates were checked after a few days for the development of perithecia.

Bacterial strain:
Escherichia coli DH5{alpha} [F- endA1 hsdR17 supE44 thi-1 recA1 gyrA96 relA1 ({phi}80d lacZ{Delta}M15); Bethesda Research Laboratories, Gaithersburg, MD] was used for bacterial transformation and plasmid propagation. Bacterial plasmids used in this study were constructed by standard recombinant DNA techniques.

Transformation of protoplasts:
Protoplasts were prepared and transformed as described previously (BERGES and BARREAU 1989 Down). For transformation, the pMOcosX (ORBACH 1994 Down) bearing the hygromycin resistance gene (hph) as a selectable marker was used as a vector. Transformants were selected on synthetic medium supplemented with hygromycin B at 100 µg/ml. When cotransformation experiments were performed, 5 µg of the transforming plasmid was mixed with 2 µg of pMOcosX. To screen for the deletion of the mod-A gene, a mod-B1 strain bearing a leu1-1 mutation was constructed to serve as a recipient strain. The cloned leu1 gene (B. TURCQ, unpublished results) was used as a selectable marker for transformation.

Nucleic acid isolation:
General methods for nucleic acid analysis and plasmid construction were standard procedures (SAMBROOK et al. 1989 Down). P. anserina DNA was prepared as described in JAVERZAT et al. 1993 Down or using the miniprep procedure described by LECELLIER and SILAR 1994 Down. The library construction has been described elsewhere (LOUBRADOU et al. 1996 Down).

DNA manipulations and reverse transcription-PCR:
Exonuclease III sequential deletions were performed using an ExoIII/Mung bean nuclease deletion kit (Stratagene, La Jolla, CA) as recommended by the supplier. Sequencing was performed using the dideoxynucleotide chain termination method (SANGER et al. 1977 Down) with the Sequenase reagent kit (version 2.0; United States Biochemical Corp., Cleveland) and [{alpha}-35S]dATP. The DNA sequence was established on both strands using clones sequentially deleted from both ends of the fragment and completed using specific synthetic primers derived from the available sequence.

Total RNAs and polyA+ RNAs were prepared as described (TURCQ and BEGUERET 1987 Down). Northern blots were probed with the 7.4-kbp ClaI fragment from pCC3 (see Figure 1). For slot-blot analysis, different strains were grown onto cornmeal solid medium overlaid with cellophane disks, and RNAs were extracted as described earlier, from the mycelium scraped from the cellophane disks. Decreasing amounts of RNA (5, 1, 0.2, and 0.04 µg, respectively) were blotted as described (SAMBROOK et al. 1989 Down) using a HYBRI-SLOT manifold (BRL Life Technologies). For reverse transcription-PCR (RT-PCR), the 3'RACE kit for rapid amplification of cDNA ends (BRL Life Technologies, Gaithersburg, MD), based on the system described by FROHMAN et al. 1988 Down, was used. The 37-nucleotide AP primer containing a (dT)17 tail at the 3' end was used to prime the cDNA synthesis with reverse transcriptase. For the first PCR, 1/20 of the cDNA reaction product was amplified using the AP primer and one of the sequence-specific synthetic primers. The second PCR was performed using 0.5 µl of the first PCR product as the source of DNA and two internal, sequence-specific primers. Primers used were as follows: OLI1 = 5'CGCCACGAGACCGCCAAC3'; OLI3 = 5'GAA CGAGATTGCAGAGACC3'; OLI5 = 5'CCGCGCTGTCACG TCG3'; OLI6 = 5'GCGAGTAGCCTGGAGTTGC3'; OLI7 = 5'GTCCATGGAGCGGGAGAGTCAAG3'; OLI8 = 5'GAAGCA CGAGGGTACACCAG3'; OLI10 = 5'GCATGCGGTAGGCGCT GTTGCGG3'; ROX = 5'GGCCCCGGGCGCCGTTGTC3'.



View larger version (20K):
In this window
In a new window
Download PPT slide
  		Figure 1. —Map of the cloned DNA fragment containing the mod-A gene. (A) Restriction map of the ClaI insert of pCC3. Restriction sites: BamHI (B), ClaI (C), EcoRV (EV), MluI (Ml), NruI (Nr), and SmaI (Sm). Each of the horizontal lines above the map represents a subclone issued from the Exonuclease III deletion experiment. The right margin shows the percentage of fertile transformants obtained with pCC3 and with each of these subclones. The grey box delimits the NruI fragment containing the mod-A gene. The mod-A ORF is schematized by the thick line bordered by the ATG and the polyadenylation site. The 55-bp intron is represented by the small box. The arrows represent the synthetic oligonucleotides used in the mRNA characterization (1 = OLI1; 3 = OLI3; 5 = OLI5; 6 = OLI6; 7 = OLI7; 8 = OLI8; 10 = OLI10; and AP = adapter primer). (B) Restriction map of the {Delta}mod-A(leu1) construct used to generate the strain deleted for the mod-A gene. The left part of the pCC3 insert, delimited by the ClaI and MluI restriction sites, was first deleted to generate a subclone in which mod-A was contained in a XbaI-ClaI fragment. The SmaI fragment containing the ATG and 3/4 of the MOD-A NH2-terminus was replaced by a 2.2-kbp blunt-ended HindIII-PstI fragment containing the P. anserina leu1 gene. Xb, XbaI. Arrows indicate the orientation of the genes.

The PCR products were made blunt-end by treatment with Klenow polymerase cloned in the EcoRV site of the pBluescript SK+ vector and sequenced as above.


*  RESULTS
*TOP
 *ABSTRACT
 *MATERIALS AND METHODS
 *RESULTS
*DISCUSSION
 *LITERATURE CITED

Cloning of the mod-A gene:
A library of 3000 clones (LOUBRADOU et al. 1996 Down) constructed in the pMOcosX cosmid (ORBACH 1994 Down) from DNA containing the wild-type mod-A gene was used to transform protoplasts of the het-c3 het-e3 mod-A1 mod-B1 strain. Hygromycin-resistant transformants were tested for their capacity to develop female organs. Three transformants that displayed perithecia after fertilization and were indistinguishable from the wild-type control were obtained. DNA was prepared from the fertile transformants and used either for in vitro packaging or for transformation of competent DH5{alpha} E. coli cells. The cosmid clones obtained after transfection following packaging were all identical in restriction pattern, but none of them were able to complement the defect in fertility of the het-c3 het-e3 mod-A1 mod-B1 strain. One cosmid, named pR4-5, obtained after direct transformation of E. coli, produced fertile transformants upon transformation of the recipient strain. ClaI restriction of pR4-5 DNA resulted in only four insert fragments, which were cloned and assayed for ability to confer the fertile phenotype. The pCC3 plasmid containing a 7.5-kbp genomic ClaI fragment retained this ability. The map of this insert is shown in Figure 1A.

To demonstrate that the cloned fragment was able to complement the mod-A1 phenotype in incompatibility, transformants obtained with pR4-5 cosmid were subjected to genetic and molecular analyses. Four transformants were purified by crossing with a het-c3 het-e3 mod-A1 mod-B1 strain isogenic to the recipient strain except for the mating type. For each strain, one homocaryotic, hygromycin-resistant strain was selected among the progeny (purified pR4-5 transformant) and then crossed with a het-c2 het-e1 mod-A1 MSI strain (het-c2 and het-e1 are incompatible genes); offspring were then analyzed (Table 1). SI and MSI strains were observed in the expected proportions, demonstrating that the cloned fragment was not only able to complement the sterile phenotype but also restored a wild-type behavior (i.e., self-incompatibility) identical to the wild-type mod-A allele.


 
View this table:
In this window
In a new window

  		Table 1. Crosses performed to characterize the cloned mod-A gene

Then, it was shown that the cloned gene was mod-A and not an extragenic suppressor of mod-A1. It has been previously observed that plasmids containing large inserts such as cosmids integrate with a high frequency at the resident locus (PICARD et al. 1987 Down). To locate the integration site of the cosmid, four independent pR4-5 transformants were crossed with a het-c2 het-e2 strain containing the wild-type mod-A gene. For one transformant, MSI strains were observed among the progeny, showing that integration occurred outside the mod-A locus; but for the three other transformants, no MSI strains were obtained, and SI strains were observed in 1/4 of the progeny—as expected for integration of the complementing gene at the mod-A locus. For at least two of these transformants, it was confirmed by Southern blot analysis that transformation occurred by integration in the genomic counterpart of the cloned fragment (result not shown), thus demonstrating that this fragment contained the mod-A gene.

Eight fertile strains issued from transformation with pCC3 were also crossed with a het-c2 het-e2 strain. As demonstrated by the presence of MSI strains in the progeny, six of the eight strains resulted from integration of pCC3 at an ectopic locus. DNA blot analysis confirmed this result. This excluded the possibility that fertile transformants might arise by reconstitution of a functional gene as the result of integration of the plasmid containing an incomplete wild-type mod-A gene in the mod-A1 mutant copy. This result therefore demonstrates that the mod-A gene contained in the cloned CC3 fragment is functional.

Analysis of the mod-A gene:
Exonuclease III deletions were performed starting from the two ends of the insert of pCC3. Clones corresponding to progressive deletions of about 1000 bp were used for cotransformation experiments with pMOcosX, and the proportion of fertile transformants was determined. This allowed us to locate the complementing gene in a region corresponding to about three kbp. The pNRU plasmid, a subclone containing a 2828-bp NruI fragment (boxed in grey in Figure 1A), was able to complement the mod-A1 defect.

The insert of pNRU was sequenced on both strands, and an open reading frame (ORF) 687 amino acids long (nucleotides 438–2557) interrupted by a potential 55-bp intron (nucleotides 1510–1565) was identified (Figure 1 and Figure 2). Complementation of the mod-A1 defect is lost when pNRU is restricted with BamHI or with EcoRV—which cut, respectively, at the beginning and at the end of the ORF (Figure 1). These data suggest that this ORF encodes the putative MOD-A protein.



View larger version (108K):
In this window
In a new window
Download PPT slide
  		Figure 2. —Nucleotide sequence and ORF of the mod-A gene. The numbers on the left refer to the nucleotide sequence, and those on the right refer to the amino acid sequence. The asterisk represents the stop codon. The elements of the promoter are lightly underscored. The intron sequence is heavily underscored, with italicized nucleotides corresponding to the conserved splice junctions and lariat sequences. The proline-rich regions are boxed, and the region showing alignment with the yeast verprolin repeated motif resembling SH3-binding domains is boxed in grey. The accession number is AF025289.

Probably because the amount of the mRNA is insufficient, no transcripts corresponding to this ORF could be detected by Northern blotting. The mRNA transcribed from the mod-A gene was identified using an approach based on the 3'RACE procedure (FROHMAN et al. 1988 Down). The polyA+ mRNAs from the wild-type strain were used to synthesize cDNAs using the adapter primer (AP) designed to hybridize with polyA tails. After a first PCR with either the AP-OLI3 oligonucleotide pair or the AP-OLI5 pair (Figure 1), no specific band could be identified by ethidium bromide staining. When a Southern blot of the PCR products was probed with the NruI fragment, one 1200-bp band (PCR with AP-OLI3 pair) and one 700-bp band (PCR with AP-OLI5 pair) were identified. No PCR product was detected in a control experiment where the reverse transcriptase had been omitted in the reverse transcription step. In contrast, an abundant specific PCR product was detected by ethidium bromide staining when using the same cDNA preparation as above and a pair of primers made up of AP and ROX, a primer specific for the het-c gene (SAUPE et al. 1994 Down). From the size of the bands detected in the Southern blot, it could be deduced that the 3'-OH end of this potential mRNA lies about 100–150 bases downstream from the TAG stop codon located at position 2555 (Figure 2). A second round of PCR was performed using as substrate an aliquot of the first PCR product issued from the AP-OLI3 pair and the OLI3-OLI6 or OLI5-OLI8 pairs as primers. For the OLI5-OLI8 pair, a fragment of the expected size (328 bp) was obtained. For the OLI3-OLI6 pair, which amplifies a region spanning a potential intron, the size of the fragment obtained was consistent with the size expected for a spliced mRNA (357 bp). Sequencing of this fragment showed that it corresponds to the sequence of the ORF in which the 55-bp intron (Figure 2) had been spliced at the potential junction sites detected by sequence analysis.

In order to confirm that an mRNA was transcribed from the region corresponding to the 5' part of the potential ORF, an RT reaction with polyA+ mRNAs as substrate was performed using the mRNA-specific oligonucleotide OLI6 as primer. A first PCR was then performed with the pair OLI1-OLI6. Again, no specific band could be detected, but when the product of the first PCR was used as substrate for a second PCR with the OLI7-OLI10 pair as primer, it was possible to amplify a fragment of the expected size in which the intron has been removed.

The detection by RT-PCR of a cDNA corresponding to the whole potential mod-A ORF allowed us to demonstrate that this ORF is expressed. The 2828-bp NruI fragment (boxed in grey in Figure 1) containing this ORF will henceforth be referred to as the mod-A gene, and the polypeptide deduced from this ORF as the MOD-A putative protein.

MOD-A begins at position 438 (Figure 2), and the sequence around the initiation codon (CAACATGGA) follows the preference observed in fungi (i.e., an A or a purine at position -3) and has eight of the nine conserved nucleotides when compared with the fungal consensus sequence (BALLANCE 1986 Down). Typical transcription signals can be detected in the 5' upstream region. A potential TATA box is found at position 257 in a pyrimidine-rich region, 180 bp upstream from the ATG, and a CAAT box lies 20 bp upstream of this TATA box. In addition to these conventional promoter elements, three blocks containing 10–12 contiguous G or C are found 32, 100, and 256 bp upstream from the ATG.

The ORF is interrupted by a small intron. The splice junctions, or sites for lariat formation, are in good agreement with the consensus splicing sites 5'-G(A/T)GTA(G)A(C)GT(C/A)T(C) and A(G)C(T)AGC(G) 3', and with the lariat sequence TGCTAAT/C found for P. anserina introns (our unpublished results) or more generally for filamentous fungi (BALLANCE 1986 Down). At the end of the ORF, the ATAAA sequence is present 130 bases downstream from the stop codon. This sequence may—according to the results of the RT-PCR, and in agreement with the situation generally found in eukaryotes—correspond to the polyadenylation signal.

The 687-amino-acid MOD-A predicted polypeptide is rich in leucine (9.3%) and proline (7.4%). Analysis of the polypeptide sequence did not give any information concerning the potential function of this protein, except that most proline residues are clustered in a short, proline-rich (30% proline) region (boxed in Figure 2). Neither Infobiogen nor National Center for Biotechnology Information (NCBI) BLAST (ALTSCHUL et al. 1990 Down) database searches with MOD-A revealed significant homology with any known protein, if one excludes the alignments resulting from the C-terminal proline-rich region, which give relatively high scores with many proline-rich proteins. When BLAST was performed with the 60-amino-acid C-terminal proline-rich domain, with filtering to eliminate low-complexity regions (ALTSCHUL et al. 1994 Down), a similarity with the yeast Saccharomyces cerevisiae proline-rich verprolin (DONNELLY et al. 1993 Down) was observed. Careful inspection of the protein alignments shows that, in fact, the MOD-A proline-rich motif contains a domain (boxed in grey in Figure 2) exhibiting significant similarity with various verprolin proline-rich repeated motifs. These motifs have been reported to display similarities to sequences of SH3-binding domains (DONNELLY et al. 1993 Down). The MOD-A domain contains proline at positions matching the conserved prolines of the SH3 ligand consensus sequence X P Ø P P X P, with Ø representing hydrophobic residues and the two prolines in bold representing conserved residues that have been shown to play a critical role for binding to SH3 domains (REN et al. 1993 Down; YU et al. 1994 Down; COHEN et al. 1995 Down).

The mod-A gene product is involved in differentiation of protoperithecia:
The mod-A1 mutation does not confer a mutant phenotype in a compatible background, but when it is associated with mod-B1, female sterility is observed due to a defect in the formation of protoperithecia (BOUCHERIE and BERNET 1974 Down). This sterility is conditional and also greatly depends on the alleles present at the het-c and het-e loci (BOUCHERIE and BERNET 1980 Down). The phenotype of the het-c2 het-e2 strain used in this study was analyzed in different genetic combinations for mod-A and mod-B genes. For the mod-A1 strain compared with wild type, no difference is observed after exposure to the light, which induces the formation of protoperithecia (Figure 3A). As for mod-A1, the mod-B1 mutation alone does not alter fertility. The density of protoperithecia is similar for the two mod-A1 and mod-B1 strains and comparable to that of the wild-type strain (114, 148, and 168 perithecia per cm2, respectively). However, when mod-A1 is associated with mod-B1, the mycelium is less dense, with very few aerial filaments, and the density of protoperithecia decreases (57 per cm2). In addition, these female organs are defective. Except for the trichogyne, which looks normal, they lack pili, which correspond to the nearby filaments that later form the envelope of the perithecium (Figure 3B). Such protoperithecia cannot form perithecia when fertilized with microconidia (Figure 4A). This defect is restored by transformation with the mod-A gene. The consequence of the defect on the development of female organs is also illustrated in Figure 4B. This figure shows the formation of perithecia resulting from cross-fertilization in the contact zone between het-c2 het-e2 strains of opposite mating types bearing different combinations at mod-A and mod-B loci and grown on minimal medium. Abundant perithecia are formed between the two wild-type strains. The crosses between mod-A1 and mod-B1 or between mod-A1 and wild-type strains are fertile. However, on this medium, it can be observed that the density of perithecia is clearly lower on the mycelium of the mod-A1 strain and that the perithecia are smaller. The density of perithecia in the cross between the mod-B1 and wild-type strain is comparable to that observed between two wild-type strains. In contrast, between the two mod-A1 mod-B1 strains of opposite mating type, no perithecia are formed, illustrating the sterile phenotype. It can also be observed that in the crosses between mod-A1 mod-B1 strains and any other strain, the perithecia never develop on the thallus of the mod-A1 mod-B1 strain.



View larger version (53K):
In this window
In a new window
Download PPT slide
  		Figure 3. —Defect in protoperithecia formation resulting from the mod-A1 mutation associated with the mod-B1 mutation. Mycelia were grown on solid cornmeal medium for 5 days and exposed to the light to induce the development of protoperithecia. (A) View of the mycelia at x50 magnification. (B) Microscopic observation of defective protoperithecia of mod-A1 mod-B1 strain compared with those of the wild-type strain. Magnification is x200.



View larger version (55K):
In this window
In a new window
Download PPT slide
  		Figure 4. —Fertility of different strains with various combinations of the mod-A gene. (A) View of the mycelia and perithecia (magnification is x4) of strains grown on cornmeal medium and fertilized with microconidia as described in MATERIALS AND METHODS. Tmod-A+mod-B1 is a fertile strain is-sued from transformation of the mod-A1 mod-B1 strain with the pNRU plasmid containing the wild-type mod-A gene. T{Delta}mod-A mod-B1 is a strain deleted for the mod-A gene. (B) Cross-fertilizations. The mycelia were grown on synthetic solid medium until confrontation and exposed to the light until full development of perithecia resulting from fertilization in the contact zone between two thalli. Magnification is x2.

The mod-A gene product is not essential, and its deletion confers the suppressor phenotype:
It has been previously shown that the mod-A1 mutation can be suppressed by a nonsense suppressor (BEGUERET 1973 Down). This suggests that it is the loss of MOD-A function, rather than its alteration, that restores the growth of self-incompatible strains. To confirm this hypothesis, we constructed a strain deleted for the mod-A gene. The rationale was that the mod-A1 gene is not essential for growth and that the deletion of this gene in a strain containing the mod-B1 mutation will lead to female sterility. The {Delta}mod-A(leu1) construct shown in Figure 1B was used for direct gene replacement. The major part of the mod-A ORF, including the ATG, was replaced by the P. anserina leu1 gene. The XbaI-ClaI DNA fragment containing the deleted mod-A gene was used to transform a leu1-1 het-c3 het-e3 mod-B1 recipient strain. Prototrophic transformants were tested for the development of female organs. Among 1200 transformants screened, 50 showed very low fertility or total sterility. Twenty-five of these transformants were used as males in a cross with the leu1-1 het-c3 het-e3 mod-B1 recipient strain. Five strains in which female sterility segregated with the leu1 marker, as expected for a gene replacement at the mod-A locus (Table 2), were selected for further analysis. Southern analysis (results not shown) confirmed a correct gene replacement by the deleted construct in two of these strains, T19{Delta}mod-A and T448{Delta}mod-A. The properties of the deletion mutants were compared to those of the mod-A1 mutant. Deletion of the mod-A gene conferred sterility only in association with the mod-B1 mutation. The phenotype of these strains was indiscernible from the mod-A1 mod-B1 mutant phenotype (Figure 4A). To analyze the effect of the deletion in an incompatible interaction, a cross with a het-c2 het-e2 wild-type strain was performed. In the progeny, SI and MSI strains were obtained in the expected proportions (Table 2). The MSI strains displayed the typical mod-A1 MSI phenotype, confirming definitely that the cloned gene is mod-A and that it is the absence of this gene product that restores the capacity of SI strains to grow. The strains deleted for the mod-A gene did not display an obvious phenotype in a compatible wild-type background.


 
View this table:
In this window
In a new window

  		Table 2. Crosses performed to characterize the mod-A deleted strain

Constitutive overexpression of mod-A has no detectable effect:
We investigated whether the constitutive overexpression of the mod-A gene in a compatible background would affect protoperithecia growth and/or formation. The ORF between BamHI and NruI restriction sites (see Figure 1) was cloned downstream from the Aspergillus nidulans GPD constitutive promoter (PUNT et al. 1987 Down) to give the pGPD-MODA plasmid. This plasmid was used with the pMOcosX vector to cotransform the het-c3 het-e3 mod-A1 mod-B1 strain. Transformation with pGPD-MODA produces the same proportion of fertile transformants as pNRU, the plasmid containing the mod-A gene with its own promoter. This result demonstrates that the construct is functional in complementing the deficiency of the mod-A1 mutation. Using slot-blot analysis of total RNAs from the wild type, it was possible to detect a faint signal with the mod-A gene as probe. As a control, no signal was observed with RNA preparations of the mod-A deleted strains. A significantly stronger signal was observed in RNAs prepared from two pGPD-MODA transformed strains (Figure 5). Compared with the control het-c3 het-e3 mod-B1 strain or to the wild type, no differences in the growth rate and development of perithecia of these strains could be observed.



View larger version (81K):
In this window
In a new window
Download PPT slide
  		Figure 5. —Slot-blot analysis of total RNAs from strains with modified expression of the mod-A gene. Total RNAs from the different strains were prepared and blotted as described in MATERIALS AND METHODS, using the mod-A gene as probe. 1, wild-type strain (WT); 2 and 3, RNAs from T19{Delta}mod-A and T448{Delta}mod-A deleted strains; 4 and 5, RNAs from two pGPD-MODA transformants.


*  DISCUSSION
*TOP
 *ABSTRACT
 *MATERIALS AND METHODS
 *RESULTS
 *DISCUSSION
*LITERATURE CITED

In P. anserina, mutations occurring in the mod-A locus are able to restore the growth of SI strains resulting from the combination of antagonistic incompatible genes from the three nonallelic het-c/het-e, het-c/het-d, and het-r/het-v systems but do not suppress allelic interactions (BELCOUR and BERNET 1969 Down; BERNET et al. 1973 Down). In the MSI strains, cell death is delayed but not suppressed, and lytic activities associated with cell death are not affected by mod-A mutations. These observations suggest that mod-A does not interfere with the cell death process, although this hypothesis cannot be fully excluded. The mod-A1 mutation was reported to interfere with the development of protoperithecia when associated with mutations in the mod-B locus (BOUCHERIE and BERNET 1974 Down). In an attempt to understand how the product of nonallelic incompatibility genes may interfere with a gene that is involved in differentiation, we isolated the mod-A gene.

A genomic fragment was cloned by complementation of the sterility defect of the mod-A1 mod-B1 strain. It was shown that the cloned gene also complemented the mod-A1 mutation in the incompatibility reaction. It was confirmed by gene replacement that the cloned gene was the mod-A gene. The mod-A deleted strains are identical to the wild-type strain in a compatible background but exhibit female sterility when the deletion is associated with the mod-B1 mutation. This result is not surprising, since some mod-A mutations previously characterized as nonsense mutations (BEGUERET 1973 Down) conferred the same phenotype. As expected, the deletion of the mod-A gene also confers the MSI phenotype in an incompatible interaction. These results demonstrate (1) that the two functions, growth arrest in incompatible interaction and protoperithecia development, are controlled by mod-A and (2) that both the suppressor effect and the defect in fertility result from the absence of the mod-A gene product.

The ORF corresponding to mod-A was identified. The mRNA could not be detected by Northern blot in polyA+ mRNA preparations. As mod-A is always active in self-incompatibility, it should be expressed constitutively, at least at a low level. However, mod-A is also expected to act during female organ development in the fungal cycle. Its expression might then be regulated developmentally. However, increased expression at the moment of female organ differentiation was not detected (result not shown), and overexpression using a constitutive promoter produces fertile strains with the same phenotype as transformants obtained with the wild-type mod-A gene, suggesting that growth and differentiation are not affected by variation in the level of expression of the mod-A gene.

The predicted MOD-A protein deduced from the ORF sequence has no significant similarity to any known protein in the databases. However, sequence alignments with the yeast verprolin could be found for a small region rich in proline. The alignments involved a region in MOD-A that aligns with the various proline repeats of verprolin, reported to exhibit similarity to SH3-binding motifs (DONNELLY et al. 1993 Down). The MOD-A domain contains a potential motif with proline at positions matching the conserved prolines of the SH3 ligand consensus sequence X P Ø P P X P (YU et al. 1994 Down). Such short proline-rich motifs have been characterized by their ability to bind SH3 domains of molecules acting in signaling processes such as protein kinases, phosphatases, and Ras-controlling proteins and are more generally involved in protein-protein interactions (KOCH et al. 1991 Down; COHEN et al. 1995 Down). Recently, different observations suggested a role for proteins containing SH3 domains and SH3 ligand motifs in fungal morphogenesis. In the yeast S. cerevisiae, the defect in verprolin leads to large cells with distorted morphology (DONNELLY et al. 1993 Down), and the BEM1 gene containing SH3 domain motifs plays a role in the polarized growth during budding and mating (CHENEVERT et al. 1992 Down). A putative SH3 ligand motif has also been observed in the predicted myp1 gene product of the dimorphic fungus Ustilago maydis, which was shown to be required for full expression of the mycelial phenotype during the transition from budding cells to filamentous growth (GIASSON and KRONSTAD 1995 Down). A gene required for nuclear migration and hyphal morphogenesis in N. crassa has recently been reported to contain proline-rich regions with potential SH3-binding domains (VIERULA and MAIS 1997 Down). This gene seems to play an important role in early events involved in the conidiation process.

Several recent results in our laboratory suggest an interconnection between incompatibility and signal transduction in P. anserina. Nonallelic genes of the het-c/het-e incompatibility system were cloned. The het-c gene encodes a small protein displaying similarity to a glycolipid transferase protein, and its deletion drastically reduces the ascospore production (SAUPE et al. 1994 Down). The het-e gene codes for a large polypeptide with two domains, a functional GTP-binding motif and a domain with WD40 repeats, which suggests that this polypeptide can be involved in protein-protein interactions (SAUPE et al. 1995 Down; ESPAGNE et al. 1997 Down). It was then shown that an additional copy of the adenylate cyclase gene relieves developmental defects resulting from mutations in a gene involved in vegetative incompatibility (LOUBRADOU et al. 1996 Down). More recently, the mod-E gene, in which mutations suppress some developmental defects, was shown to partly restore the growth of het-r/het-v SI strains. This mod-E gene (LOUBRADOU et al. 1997 Down) was characterized as a member of the HSP90 protein family, which is implicated in many cellular processes involving signal transduction (LINDQUIST and CRAIG 1988 Down). The presence of potential SH3-binding motifs in the MOD-A protein may also suggest a potential role in signal transduction during female organ formation. This signal may transfer the information required to control differentiation of protoperithecia from quiescent mycelial cells, where fruiting bodies form. However, the effect of the absence of a signal can be observed only when mod-B1 mutation is present. This suggests a dual action of these two genes in the same pathway or involvement in parallel pathways. An inhibition of apical growth might be a prerequisite to successfully differentiate a female organ from quiescent cells. A tempting interpretation has been proposed by BERNET 1992 Down. As reported previously, the level of sterility of the mod-A1 mod-B1 strain is conditional, depending on the temperature, the medium, or the age of the mycelia (BOUCHERIE and BERNET 1980 Down). This might suggest for MOD-A a role in particularly limiting conditions—such as total exhaustion of nutrients, for example. To mobilize all the residual growing potential for the formation of fruiting bodies, a growth arrest might be coupled with a partial lysis of hyphal cells in the region surrounding the protoperithecia. This lysis, which is thought to generate the nutrients necessary for the achievement of female organ development, would be controlled by the mod-B gene. Without these two control steps—which are presumably interconnected—the development aborts, leading to defective protoperithecia. Arguing in favor of this hypothesis are the facts that sterility of mod-A1 mod-B1 strains can be fully suppressed by addition of various amino acids to the medium (BOUCHERIE and BERNET 1980 Down) and that sterility is never observed on complete medium.

The question remains how these control steps can be linked to three nonallelic het systems. First, these het genes are probably redundant (HET-E and HET-D polypeptides are very similar; unpublished result) and can be presumed to act in the same pathway. Second, the observation that the level of female sterility of the mod-A1 mod-B1 strains is strongly dependent on the combination of compatible het genes at het-c, het-e, het-d, het-r, and het-v loci (BOUCHERIE and BERNET 1980 Down) led BERNET 1992 Down to propose a potential role for the nonallelic het genes in the control of female organ development. It can be speculated that this control involves MOD-A. Then, in case of an incompatible interaction of nonallelic het gene products, an abnormal signal leads to a cellular catastrophe resulting from the deregulation of the two pathways controlling growth and cell lysis, respectively. Mutations in mod-A suppress the effect of this signal transmission, thus allowing the growth, and dominant mutations in mod-B relieve the downstream process of cell lysis.

The functionality of the potential SH3 ligand of the MOD-A polypeptide is now under investigation, and its potential interaction with het-e and/or het-c gene products will be assessed. Determination of the exact function of HET-C and HET-E in the fungus and cloning of the mod-B gene will help to explain the mechanism connecting incompatibility and female organ differentiation.


*  ACKNOWLEDGMENTS

We warmly thank JEAN BERNET for helpful discussions. We acknowledge the technical assistance of VERONIQUE ROQUES-DUFLO. This work was supported by a grant from the Centre National de la Recherche Scientifique. G.L. was supported by a fellowship from the Association pour la Recherche contre le Cancer.

Manuscript received October 21, 1997; Accepted for publication February 26, 1998.


*  LITERATURE CITED
*TOP
 *ABSTRACT
 *MATERIALS AND METHODS
 *RESULTS
 *DISCUSSION
 *LITERATURE CITED

ALTSCHUL, S. F., W. GISH, W. MILLER, E. W MYERS, and D. J. LIPMAN, 1990  Basis local alignment search tool. J. Mol. Biol. 215:403-410[Medline].

ALTSCHUL, S. F., M. S. BOGUSKI, W. GISH, and J. C. WOOTTON, 1994  Issues in searching molecular sequence databases. Nat. Genet. 6:119-129[Medline].

ARGANOZA, M. T., J. OHRNBERGER, J. MIN, and R. A. AKINS, 1994  Suppressor mutants of Neurospora crassa that tolerate allelic differences at single or at multiple heterokaryon incompatibility loci. Genetics 137:731-742[Abstract].

BALLANCE, D. J., 1986  Sequences important for gene expression in filamentous fungi. Yeast 2:229-236[Medline].

BÉGUERET, J., 1973 L'incompatibilité cellulaire chez le champignon Podospora anserina: Étude physiologique de la réaction. Thesis, Université de Bordeaux 2.

BÉGUERET, J., B. TURCQ, and C. CLAVÉ, 1994  Vegetative incompatibility in filamentous fungi: het genes begin to talk. Trends Genet. 10:441-446[Medline].

BELCOUR, L. and J. BERNET, 1969  Sur la mise en évidence d'un gène dont la mutation supprime spécifiquement certaines manifestations d'incompatibilité chez le Podospora anserina.. C. R. Acad. Sci. Paris 269:712-714.

BERGÈS, T. and C. BARREAU, 1989  Heat-shock at an elevated temperature improves transformation efficiency of protoplasts from Podospora anserina.. J. Gen. Microbiol. 135:601-604[Medline].

BERNET, J., 1967  Systèmes d'incompatibilité chez le Podospora anserina.. C. R. Acad. Sci. Paris 265:1536-1539.

BERNET, J., 1971  Sur un cas de suppression de l'incompatibilité chez le champignon Podospora anserina.. C. R. Acad. Sci. Paris 273:1120-1122.

BERNET, J., 1992  In Podospora anserina, protoplasmic incompatibility genes are involved in cell death control via multiple gene interactions. Heredity 68:79-87.

BERNET, J., J. BÉGUERET, and J. LABARÈRE, 1973  Incompatibility in the fungus Podospora anserina. Are the mutations abolishing the incompatibility reaction ribosomal mutations? Mol. Gen. Genet. 124:35-50[Medline].

BOUCHERIE, H. and J. BERNET, 1974  Protoplasmic incompatibility and female organ formation in Podospora anserina: Properties of mutations abolishing both processes. Mol. Gen. Genet. 135:163-174[Medline].

BOUCHERIE, H. and J. BERNET, 1980  Protoplasmic incompatibility in Podospora anserina: a possible function for incompatibility genes. Genetics 96:399-411[Abstract/Free Full Text].

CHENEVERT, J., K. CORRADO, A. BENDER, J. PRINGLE, and I. HERSKOWITZ, 1992  A yeast gene (BEM1) necessary for cell polarization whose product contains two SH3 domains. Nature 356:77-79[Medline].

COHEN, G. B., R. REN, and D. BALTIMORE, 1995  Modular binding domains in signal transduction proteins. Cell 80:237-248[Medline].

DONNELLY, S. F. H., M. J. POCKLINGTON, D. PALLOTTA, and E. ORR, 1993  A proline-rich protein, verprolin, involved in cytoskeletal organization and cellular growth in the yeast Saccharomyces cerevisiae.. Mol. Microbiol. 10:585-596[Medline].

ESPAGNE, E., P. BALHADÈRE, J. BÉGUERET, and B. TURCQ, 1997  Reactivity in vegetative incompatibility of the HET-E protein of the fungus Podospora anserina is dependent on GTP-binding activity and WD40 repeated domain. Mol. Gen. Genet. 256:620-627[Medline].

ESSER, K., 1974 Podospora anserina, pp. 531–551 in Handbook of Genetics, edited by R. C. KING. Plenum Press, New York.

FROHMAN, M. A., M. K. DUSH, and G. R. MARTIN, 1988  Rapid production of full-length cDNAs from rare transcripts: Amplification using a single gene-specific oligonucleotide primer. Proc. Natl. Acad. Sci. USA 85:8998-9002[Abstract/Free Full Text].

GIASSON, L. and J. W. KRONSTAD, 1995  Mutations in the myp1 gene of Ustilago maydis attenuate mycelial growth and virulence. Genetics 141:491-501[Abstract].

GLASS, N. L., J. GROTELUESCHEN, and R. L. METZENBERG, 1990  Neurospora crassa A mating-type region. Proc. Natl. Acad. Sci. USA 87:4912-4916[Abstract/Free Full Text].

GLASS, N. L. and G. A. KULDAU, 1992  Mating type and vegetative incompatibility in filamentous ascomycetes. Annu. Rev. Phytopathol. 30:201-224.

JACOBSON, D. J., 1992  Control of mating type heterokaryon incompatibility by the tol gene in Neurospora crassa and N. tetrasperma.. Genome 35:347-353[Medline].

JAVERZAT, J. P., V. BHATTACHERJEE, and C. BARREAU, 1993  Isolation of telomeric DNA from the filamentous fungus Podospora anserina and construction of a self-replicating linear plasmid showing high transformation frequency. Nucleic Acids Res. 21:497-504[Abstract/Free Full Text].

KOCH, C. A., D. ANDERSON, M. F. MORAN, C. ELLIS, and T. PAWSON, 1991  SH2 and SH3 domains: Elements that control interaction of cytoplasmic signaling proteins. Science 252:668-674[Abstract/Free Full Text].

LABARÈRE, J., 1973  Propriétés d'un système d'incompatibilité chez le champignon Podospora anserina et intérêt de ce système pour l'étude de l'incompatibilité. C. R. Acad. Sci. Paris 27:1301-1304.

LECELLIER, G. and P. SILAR, 1994  Rapid methods for nucleic acids extraction from petri-dish-grown mycelia. Curr. Genet. 25:122-123[Medline].

LINDQUIST, S. and E. A. CRAIG, 1988  The heat shock proteins. Annu. Rev. Genet. 22:631-677[Medline].

LOUBRADOU, G., J. BÉGUERET, and B. TURCQ, 1996  An additional copy of the adenylate cyclase gene relieves developmental defects produced by a mutation in a gene involved in vegetative incompatibility in Podospora anserina.. Gene 170:119-123[Medline].

LOUBRADOU, G., J. BÉGUERET, and B. TURCQ, 1997  A mutation in an HSP90 gene affects the sexual cycle and suppresses vegetative incompatibility in the fungus Podospora anserina.. Genetics 147:581-588[Abstract].

NEWMEYER, D., 1970  A suppressor of the heterokaryon-incompatibility associated with mating type in Neurospora crassa.. Can. J. Genet. Cytol. 12:914-926[Medline].

ORBACH, M. J., 1994  A cosmid with a HyR marker for fungal library construction and screening. Gene 150:159-162[Medline].

PHILLEY, M. L. and C. STABEN, 1994  Functional analyses of the Neurospora crassa MT a-1 mating-type polypeptide. Genetics 137:715-722[Abstract].

PICARD, M., R. DEBUCHY, P. JULIEN, and Y. BRYGOO, 1987  Transformation by integration in Podospora anserina. II. Targeting of the resident locus with cosmids and instability of the transformants. Mol. Gen. Genet. 210:129-134.

PUNT, P. J., R. P. OLIVER, M. A. DINGUEMAUSE, P. H. POUWELS, and C. A. M. J. J. VAN DEN HONDEL, 1987  Transformation of Aspergillus nidulans based on the hygromycin B resistance marker of Escherichia coli.. Gene 56:117-124[Medline].

REN, R., B. J. MAYER, P. CICCHETTI, and D. BALTIMORE, 1993  Identification of a ten-amino acid proline-rich SH3 binding site. Science 259:1157-1161[Abstract/Free Full Text].

RIZET, G., 1952  Les phénomènes de barrage chez Podospora anserina: Analyse génétique des barrages entre les souches s et S. Rev. Cytol. Biol. Vég. 11:201-304.

RIZET, G. and C. ENGELMAN, 1949  Contribution à l'étude génétique d'un ascomycète tétrasporé: Podospora anserina.. Rev. Cytol. Biol. Vég. 11:201-304.

SAMBROOK, J., E. F. FRITSCH and T. MANIATIS, 1989 Molecular Cloning: A Laboratory Manual. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY.

SANGER, F., S. NICKLEN, and A. R. COULSON, 1977  DNA sequencing with chain terminating inhibitors. Proc. Natl. Acad. Sci. USA 74:5463-5467[Abstract/Free Full Text].

SAUPE, S., C. DESCAMPS, B. TURCQ, and J. BÉGUERET, 1994  Inactivation of the Podospora anserina vegetative incompatibility locus het-c, whose product resembles a glycolipid transfer protein, drastically impairs ascospore production. Proc. Natl. Acad. Sci. USA 91:5927-5931[Abstract/Free Full Text].

SAUPE, S., B. TURCQ, and J. BÉGUERET, 1995  A gene responsible for vegetative incompatibility in the fungus Podospora anserina encodes a protein with a GTP-binding motif and Gß homologous domain. Gene 162:135-139[Medline].

SAUPE, S., L. STENBERG, K. T. SHIU, A. J. GRIFFITHS, and N. L. GLASS, 1996a  The molecular nature of mutations in the mt A-1 gene of the Neurospora crassa A idiomorph and their relation to mating-type function. Mol. Gen. Genet 250:115-122[Medline].

SAUPE, S., G. A. KULDAU, M. L. SMITH, and N. L. GLASS, 1996b  The product of the het-C heterokaryon incompatibility gene of Neurospora crassa has characteristics of a glycin-rich cell wall protein. Genetics 143:1589-1600[Abstract].

TURCQ, B. and J. BÉGUERET, 1987  The ura5 gene of the filamentous fungus Podospora anserina: Nucleotide sequence and expression in transformed strains. Gene 53:201-209[Medline].

VELLANI, T. S., A. J. GRIFFITHS, and N. L. GLASS, 1994  New mutations that suppress mating-type vegetative incompatibility in Neurospora crassa.. Genome 37:249-255[Medline].

VIERULA, P. J. and J. M. MAIS, 1997  A gene required for nuclear migration in Neurospora crassa codes for a protein with cysteine-rich, LIM/RING-like domains. Mol. Microbiol. 27:331-340.

YU, H., J. K. CHEN, S. FENG, D. C. DALGARNO, and A. W. BRAUER et al., 1994  Structural basis for the binding of proline-rich peptides to SH3 domains. Cell 76:933-945[Medline].