Structures of the Mating-Type Loci of Cordyceps takaomontana

Abstract

Nucleotide sequences of the mating-type loci MAT1-1 and MAT1-2 of Cordyceps takaomontana were determined, which is the first such report for the clavicipitaceous fungi. MAT1-1 contains two mating-type genes, MAT1-1-1 and MAT1-1-2, but MAT1-1-3 could not be found. On the other hand, MAT1-2 has MAT1-2-1. A pseudogene of MAT1-1-1 is located next to MAT1-2.

Entomopathogenic fungus Paecilomyces tenuipes is an anamorph of Cordyceps takaomontana (Ascomycota: Pyrenomycetes: Hypocreales [Clavicipitales]: Clavicipitaceae [15]). We use the holomorphic name C. takaomontana for P. tenuipes in this report. C. takaomontana has been used as a traditional medicine in Korea. Clavicipitaceous fungi are the potential sources for medicines such as ergot alkaloids from Claviceps purpurea, cordycepin from Cordyceps sinensis, and cyclosporine from Tolypocladium inflatum. C. takaomontana is also a source of bioactive compounds such as 4-acetyl-12,13-epoxyl-9-trichothecene-3,15-diol (12). But the difficulties involved in artificially producing fruiting bodies such as perithecioid ascomata and synnemata hinder their usefulness. The analysis of mating type will shed light on the mating mechanisms and on the anamorph-teleomorph connection.

Recently the mating-type loci have been analyzed, mainly in the phytopathogenic fungi belonging to Discomycetes, Loculoascomycetes, and Pyrenomycetes (3-4, 8, 17-19). The mating types MAT1 and MAT2 are determined by the single mating-type locus MAT1. Alleles for MAT1 and MAT2 are MAT1-1 and MAT1-2, respectively. Although the flanking regions of MAT1-1 and MAT1-2 are homologous, the nucleotide sequences of MAT1-1 and MAT1-2 are highly dissimilar. The term “idiomorph” is usually used instead of “allele” for MAT1-1 and MAT1-2. Heterothallic fungi have one of the idiomorphs MAT1-1 and MAT1-2. On the other hand, most of the homothallic fungi, except for some Neurospora species (6), have both idiomorphs.

The mating system of C. takaomontana is unknown; we determined the nucleotide sequences of the mating-type loci to investigate the molecular background of its mating system.

Fungal strains.

A total of 22 isolates (BCMU IJ01 to IJ09, IJ11, IJ13 to IJ18, and IJ20 to IJ25) of C. takaomontana were obtained from the synnemata formed on the 22 lepidopteran pupae, which were collected at the Prefecture of Aichi in Japan. The nucleotide sequences of the 18S ribosomal DNA of all C. takaomontana isolates were determined. The phylogenetic analysis based on the nucleotide sequences of 18S ribosomal DNA was done using the neighbor-joining method (13). These isolates were closely related to C. takaomontana (accession numbers AB044631 [11] and AB070372 [21]) and formed a monophyly supported by a 97.7% bootstrap value. C. takaomontana IFO 31161, which had been isolated at the Prefecture of Nara in Japan, was purchased from the Institute for Fermentation, Osaka (Japan).

Mating-type locus MAT1-2.

The HMG box of MAT1-2-1 is most conserved among the mating-type gene products of MAT1-1 and MAT1-2. MAT1-2-1 protein belongs to the TCF/SOX family (9) of the HMG protein. Some primer sets for the amplification of DNA encoding an HMG box were reported (2), but no primer sets have been able to amplify the MAT1-2-1 gene of C. takaomontana. We synthesized a degenerate primer set (5′-GAGCCWCAYTTGTCSAAYAA and 5′-TTCTCCGACATTTCCTTGTA) based on the amino acid sequence of the conserved ascomycete HMG box (22). PCR using the degenerate primers was able to amplify a part of the MAT1-2-1 gene of C. takaomontana BCMU IJ13. The 6.2-kb DNA fragment containing the MAT1-2-1 gene (Fig. 1) was obtained using the cassette ligation-mediated PCR (7) and the inverse PCR. The deduced amino acid sequence of MAT1-2-1 of BCMU IJ13 has a 34.2% identity (in 196 amino acid residues) with that of Gibberella zeae (22). MAT1-2-1 of BCMU IJ13 has two introns whose positions were confirmed by sequencing of the product of reverse transcription-PCR (RT-PCR). The mRNA of BCMU IJ13 was prepared from the mycelia formed on the potato dextrose agar (Difco). After the RT, PCR was performed using the primers M2RT-F (5′-ATGGATCTGCTTCTAGATCG) and M2RT-R (5′-TTAAACGACTCGGGGCTCAT) (Fig. 1). The inserted positions of introns were same as those of the pyrenomycete fungi (22). The MAT1-2-1 gene is expressed in vegetative mycelia. This suggests the possibility that MAT1-2-1 can operate on some events besides mating. In Neurospora crassa, the MAT a-1 (a homologue of MAT1-2-1) plays roles in both vegetative incompatibility and fertilization (16).

FIG. 1.

Structures of the mating-type loci of C. takaomontana BCMU IJ25 and BCMU IJ13. Numbered arrows represent positions and directions of the primers (arrow 1, Lyase-F; arrow 2, M112-R; arrow 3, Alpha-F; arrow 4, Alpha-R; arrow 5, M1R-F; arrow 6, M1R-R; arrow 7, M1RT-F; arrow 8, M1RT-R; arrow 9, HMG-R; arrow 10, HMG-F; arrow 11, M2RT-R; arrow 12, M2RT-F). *, the boundary between MAT1-1 and the right flanking region was not able to be determined.

A latter part of the MAT1-1-1 gene, lacking the region encoding a conserved alpha box, was located upstream from the MAT1-2-1 gene (Fig. 1). This pseudogene (542 bp) does not have an initiation codon and is interrupted by several termination codons. A small gap (54 bp) is present between the nucleotide sequences of MAT1-1-1 of BCMU IJ13 and BCMU IJ25 (see below). The nucleotide sequence of the partial MAT1-1-1 of BCMU IJ13 has 91.7% identity (in 531 bp) with that of BCMU IJ25 but has little identity with those of MAT1-1-1 of other fungi reported so far. The terminal runoff of MAT1-1-1 was also reported for the flanking region of MAT1-2 of Pyrenopeziza brassicae (14).

A putative gene encoding a DNA lyase was found downstream from the MAT1-2-1 gene. The deduced amino acid sequence of the DNA lyase of C. takaomontana has 46.3% identity (in 607 amino acid residues) with that of Mycosphaerella graminicola (20). In M. graminicola, the DNA lyase gene is present downstream from the MAT1-2-1 (20). Database searching by BLAST (1) revealed a DNA lyase gene next to the mat a-1 of N. crassa (accession number M54787) (16).

Mating-type locus MAT1-1.

The alpha box of MAT1-1-1 is most conserved among the mating-type gene products of MAT1-1, but the homologies of MAT1-1-1 among the filamentous ascomycetes are relatively low. In G. zeae (22) and M. graminicola (20), a DNA lyase (described as Exo1 in reference 22) gene is present next to the MAT1-1. Database searching by BLAST (1) revealed a DNA lyase gene next to MAT1-1-3 of Cryphonectria parasitica (accession number AF380365) (10). The DNA lyase gene might be present next to both idiomorphs. The DNA lyase gene of C. takaomontana BCMU IJ25 was amplified by PCR using the primers based on the nucleotide sequence of the DNA lyase gene of C. takaomontana BCMU IJ13. The 10.2-kb DNA fragment containing the DNA lyase gene, MAT1-1-2, and MAT1-1-1 (Fig. 1) was obtained using the cassette ligation-mediated PCR (7) and the inverse PCR. MAT1-1 of the pyrenomycete fungi reported so far contains three genes, MAT1-1-3, MAT1-1-2, and MAT1-1-1 (10, 19). However, the MAT1-1-3 gene could not be found in C. takaomontana BCMU IJ25.

The deduced amino acid sequence of MAT1-1-2 of C. takaomontana has 26.8% identity (in 336 amino acid residues) with that of G. fujikuroi (22). The deduced amino acid sequence of MAT1-1-1 of C. takaomontana has 38.4% identity (in 203 amino acid residues) with that of G. zeae (22). MAT1-1-1 has one intron whose position was confirmed by sequencing of the product of RT-PCR. The mRNA of C. takaomontana BCMU IJ25 was prepared from the mycelia formed on the potato dextrose agar (Difco). After the RT, PCR was performed using the primers M1RT-F (5′-CGCTTTCAGAAGTTATTATGTG) and M1RT-R (5′-TGCTGGGGACAAGAAAGACTAG) (Fig. 1). The inserted position of intron was same as those of the hypocrealean fungi (22). MAT1-1-1 of C. takaomontana BCMU IJ25 is expressed in vegetative mycelia. This suggests the possibility that MAT1-1-1 could operate in some events besides mating. In N. crassa, the MAT A-1 (a homologue of MAT1-1-1) plays roles in both vegetative incompatibility and fertilization (5).

Mating-type loci of other C. takaomontana isolates.

We analyzed the structures of the mating-type loci of 20 other isolates from Aichi in Japan. Three sets of degenerate primers were synthesized {Alpha-F [5′-CG(A/G)GC(A/T)AA(A/G)CG(A/G)CCATTGAA(C/T)GC] and Alpha-R [5′-CCCATCTC(A/G)TC(A/T)CGGAC(A/G)AA(G/C)GA] for the alpha box-encoding part of MAT1-1-1, M1R-F [5′-(C/T)TGA(A/G)ATCGAAAGATCTCCC] and M1R-R [5′-GACAAGAAAGACTAGAAAAC] for the latter part of MAT1-1-1, and HMG-F [5′-AAGATTCC(A/G)AG(A/G)CC(A/G)CC(G/C)AA] and HMG-R [5′-CGAGGTTGATA(C/T)TGATA(C/T)TG] for the HMG box-encoding part of MAT1-2-1} (Fig. 1). PCRs were performed using an ExTaq DNA polymerase (Takara, Japan) according to the manufacturer's recommendations. After denaturation at 95°C for 1 min, amplification with 35 cycles of denaturation (95°C for 30 s), annealing (60°C for 30 s), and polymerization (72°C for 30 s) was done.

As the result of degenerate PCRs (Fig. 2), two isolates, BCMU IJ21 and BCMU IJ23, were found to the same organization as BCMU IJ25, with a complete MAT1-1-1. MAT1-1-3 was not found in isolate BCMU IJ25, although other pyrenomycete fungi usually have MAT1-1-3 upstream from MAT1-1-2 (10, 19). The lengths of the PCR products using the primers Lyase-F (5′-ACTGGCTGTGATGACAGGAC) and M112-R (5′-CTCGAGTTGCAACAGGCACG) (Fig. 1) were identical among the isolates BCMU IJ21, BCMU IJ23, and BCMU IJ25 (data not shown), so the isolates BCMU IJ21 and BCMU IJ23 might lack MAT1-1-3. But the possibility that the arrangement of MAT1-1-3 of C. takaomontana was different from those of other fungi could not be excluded. On the other hand, the rest of the BCMU isolates had MAT1-2-1 and the latter part of MAT1-1-1 (Fig. 2).

FIG. 2.

Results of degenerate PCRs for the alpha box-encoding part of MAT1-1-1 (A), the latter part of MAT1-1-1 (B), and the HMG box-encoding part of MAT1-2-1 (C). PCR products were electrophoresed on a 1.5% agarose gel. Lane M is a 100-bp DNA ladder (New England BioLabs). Lanes 1 to 25 show the PCR products from C. takaomontana BCMU IJ01 to IJ09, IJ11, IJ13 to IJ18, and IJ20 to IJ25, respectively.

The 9.7-kb DNA fragment of the mating-type locus of C. takaomontana IFO 31161 from Nara in Japan was determined. Strain IFO 31161 has also MAT1-2-1 and the latter part of MAT1-1-1. It can be speculated that C. takaomontana or its ancestor originally retained MAT1-1 and MAT1-2 and was homothallic (Fig. 1). The isolates BCMU IJ21, BCMU IJ23, and BCMU IJ25 lost MAT1-2 (and MAT1-1-3). The rest of the BCMU isolates and IFO 31161 lost MAT1-1 except for the latter part of MAT1-1-1, which was retained. The 22 BCMU isolates and IFO 31161 might have lost their homothallic nature by the partial deletions of mating-type loci.

Nucleotide sequence accession numbers.

The nucleotide sequences of the 18S ribosomal DNA of C. takaomontana isolates (BCMU IJ01 to IJ09, IJ11, IJ13 to IJ18, and IJ20 to IJ25) were deposited in the DDBJ/EMBL/GenBank databases with the accession numbers AB086203 to AB086224, respectively. The nucleotide sequences of the mating-type loci of C. takaomontana were also deposited, with accession numbers AB084921 (BCMU IJ13), AB096216 (BCMU IJ25), and AB103335 (IFO 31161).