Abstract
Mating processes in fungi are governed by mating type genes. Most (if not all) heterothallic filamentous ascomycetes have a dimictic mating system with two alleles (called idiomorphs, as they do not share any significant sequence similarity) located in a single locus (Turgeon & Yoder, 2000). One idiomorph (MAT1-2) contains a single open reading frame (ORF) encoding a regulatory protein with a DNA-binding domain of the high mobility group (HMG) type, while the other (MAT1-1) contains an ORF encoding a protein with a motif called an alpha box, which is also present in the MATα1 protein of the yeast Saccharomyces cerevisiae (Turgeon & Yoder, 2000). Such mating type genes have been identified in a number of filamentous ascomycetes belonging to fungal groups that are widely separated in evolutionary terms (e.g. in species belonging to the pyrenomycetes, discomycetes and loculoascomycetes). However, no mating type genes have been detected so far in any plectomycete genera (e.g. in teleomorphs of Aspergillus or Penicillium species), although some efforts have been made to search for them (Arie et al., 1997; Cisar et al., 1994; Rinyu et al., 2000).
The presence of homologues of mating type idiomorphs and other genes required for mating and meiosis in Aspergillus fumigatus was examined. A. fumigatus is an important airborne pathogen that causes usually fatal invasive infections in immunosuppressed patients. It has no known sexual cycle, although it is closely related to the homothallic Neosartorya fischeri (Geiser et al., 1998; Varga et al., 2000). We searched The Institute for Genomic Research (TIGR; ) unfinished A. fumigatus genomic database (prepared by shotgun sequencing, six times sequence coverage) using mating type gene sequences of Neurospora crassa. Our search resulted in the identification of two ORFs homologous to the mat a-1 gene of Neurospora crassa on contigs 4898 and 4865 (hereafter referred to as ORF4898 and ORF4865, respectively). ORF4898 is 783 bp long, contains an HMG-type DNA-binding domain, carries a putative intron and encodes a putative protein of 247 aa (Fig. 1⇓). The putative protein product of this ORF exhibited extensive similarities to MAT1-2 genes of a number of ascomycetes (Fig. 1⇓). For example, it exhibited 68 % similarity to Neurospora crassa mat a-1 in a region of 87 aa, 72 % similarity to the MAT1-2 sequence of Ceratocystis eucalypti in a region of 83 aa, and 60 % similarity to the MAT1-2 gene of Mycosphaerella graminicola in a region of 126 aa. The other identified sequence, ORF4865, encodes a HMG box protein of 434 aa, and was found to be most closely related to the ste11 gene of Schizosaccharomyces pombe (56 % similarity in a region of 88 aa). This ORF was also found to be similar to the mat1-Mc mating-type-specific polypeptide of Schizosaccharomyces pombe (60 % similarity in a region of 81 aa). The ste11 gene encodes the major transcription factor of sexual development in fission yeast. Expression of ste11 is induced by starvation via a decrease of cAMP-dependent protein kinase activity, and the gene product is involved in the activation of mating-type-specific genes and other genes needed for meiosis (Davey, 1998). However, although starvation does induce fruit-body development in Emericella species (teleomorphic states of Aspergillus nidulans and its relatives), it does not induce ascosporogenesis in Neosartorya species (sexual states of species closely related to A. fumigatus). Instead, Neosartorya species produce fruit bodies and ascospores more abundantly on nutritionally rich media (e.g. on malt extract agar) than during starvation (Raper & Fennell, 1965). These observations indicate that the ste11 homologue has other functions in A. fumigatus. Homology to the MAT1-1 gene sequence of any of the tested fungi (Neurospora crassa, Ceratocystis eucalypti, Mycosphaerella graminicola, Schizosaccharomyces pombe mat1-Pc) was not observed in the TIGR A. fumigatus genomic database. However, homologues of other genes required for mating and meiosis, e.g. those of a cAMP-dependent protein kinase (pka1), a mitogen-activated protein (MAP) kinase kinase kinase (byr2), a multiple drug resistance protein required for secretion of mating factor M (mam1) and the pheromone p-factor receptor of fission yeast, have been identified in the A. fumigatus genome (data not shown). Besides these, a homologue of the steA gene of A. nidulans was also identified in the A. fumigatus genomic database. This gene encodes a transcription factor required for sexual reproduction (Vallim et al., 2000).
Aspergillus fumigatus HMG box domain homologues. (a) Alignment of partial amino acid sequences of HMG box domains of filamentous ascomycetes and Schizosaccharomyces pombe. (b) Neighbour-joining tree of amino acid sequences of HMG box proteins of filamentous ascomycetes and S. pombe. A human mitochondrial transcription factor sequence (mtTF1) was used as the outgroup. Names of mating type idiomorphs are displayed as they appear in the GenBank database. Bootstrap values are shown. Both the alignment and the tree were constructed using clustal_x (Thompson et al., 1997).
The high levels of similarity observed between MAT1-2 sequences of various fungi and the ORF4898 sequence of A. fumigatus isolate Af293 (Fig. 1⇑) indicate that this ORF is part of a(n earlier) functional MAT1-2 gene, so A. fumigatus was able to reproduce sexually at least some time in the past. Regarding the evolutionary origin of A. fumigatus, the ancestor of this fungus could have been either a homothallic or a heterothallic Neosartorya species. Among homothallic fungi, Cochliobolus species and Gibberella zeae carry both mating type genes in one genome (Yun et al., 1999, 2000), while homothallic species in the order Sordariales carry either both mating type genes or MAT1-1 only (Glass et al., 1990). Since none of the homothallic filamentous ascomycetes examined so far carry only the MAT1-2 gene, we presume that A. fumigatus Af293 represents one mating type of a heterothallic species. The other putative mating type could have become extinct, as observed in Ophiostoma novo-ulmi and Cryptococcus neoformans populations (Mitchell & Brasier, 1994; Yan et al., 2002). Alternatively, the other mating type gene, MAT1-1, might be present in other isolates of A. fumigatus, as observed in the asexual Fusarium culmorum species (Kerényi & Hornok, 2002). The heterothallic nature of A. fumigatus is further supported by the observation that a number of A. fumigatus isolates produced abortive cleistothecia when paired with A mating type testers of the heterothallic Neosartorya fennelliae, but never produced such structures when a mating type testers of Neosartorya fennelliae were used (Udagawa & Takada, 1985). This hypothesis is in contrast with that proposed by Geiser et al. (1996), who suggested that asexual aspergilli are derived from homothallic species. In Aspergillus section Fumigati, in particular, homothallism was suggested to be an ancient character and heterothallism was suggested to be derived, based on phylogenetic analysis of β-tubulin and hydrophobin sequences (Geiser et al., 1998). However, heterothallism was proposed to be ancient in the genus Cochliobolus (Yun et al., 1999), and the putative heterothallic nature of asexual Fusarium species has also been suggested based on the detection of either MAT1-1 or MAT1-2 homologues in different Fusarium culmorum isolates (Kerényi & Hornok, 2002).
Identification of a mating type gene homologue in A. fumigatus has several implications in clinical microbiology, including epidemiology, strain typing, pathogenicity testing and population genetics. Strain-typing studies often assume that a fungus is asexual with clonal reproduction. However, the application of statistical tests developed for distinguishing recombination from clonality on isoenzyme- and PCR-based datasets of a number of A. fumigatus populations indicated recombining population structures in a recent study (Varga & Tóth, 2003). Currently available techniques do not enable us to distinguish between the alternative hypotheses of whether past or present meiotic exchanges or parasexuality were responsible for the recombining population structure of A. fumigatus. Identification of a mating type gene homologue in A. fumigatus raises the possibility that either a cryptic sexual stage or past meiotic processes are responsible for the recombining population structures observed.
In conclusion, homologues of the MAT1-2 gene and other genes needed for successful mating and meiosis of A. fumigatus have been identified in an A. fumigatus genomic database. This is the first report on the identification of a mating type gene homologue in any Aspergillus or Penicillium species. Homology to the other mating type idiomorph, MAT1-1, was not observed. Based on these observations, we propose that A. fumigatus is derived from a heterothallic ancestor either through loss of the other mating type or by loss of sexual compatibility through mutation(s) in other gene(s) needed for mating or meiosis. Further studies are needed to clarify whether the HMG box proteins of A. fumigatus are functional, to search the A. fumigatus genome for other sequences shown to be necessary for meiotic recombination in other fungi, and to examine their function. Other A. fumigatus isolates should also be examined for the presence of mating type gene homologues. This approach has been used successfully in Candida albicans (Magee et al., 2002) and Fusarium species (Kerényi & Hornok, 2002), and could thus provide evidence for past or ongoing meiotic recombination in A. fumigatus as well.