| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Genetics, Vol. 172, 1877-1891, March 2006, Copyright © 2006
doi:10.1534/genetics.105.051128
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
* Department of Biology, Duke University, Durham, North Carolina 27708 and Institute for Forest Botany, Georg-August-University, D-37077 Göttingen, Germany
1 Corresponding author: Department of Biology, Duke University, Science Dr., Box 90338, Durham, NC 27708.
E-mail: tyj2{at}duke.edu
 |
ABSTRACT |
|---|
 TOP ABSTRACT MATERIALS AND METHODSRESULTSDISCUSSIONACKNOWLEDGEMENTSLITERATURE CITED |
|---|
The multiallelic tetrapolar mating system is considered to be a novel innovation that could have only evolved once (RAPER 1966; RAPER and FLEXER 1971). For this reason, the ancestor of the homobasidiomycetes is accepted as having a tetrapolar mating system. Although most (
65%) of the homobasidiomycetes possess a tetrapolar mating system, many species (25%) instead have a bipolar system controlled by a single locus with multiple alleles (RAPER 1966). The distribution of bipolar species is scattered throughout the homobasidiomycete phylogeny, and bipolar species appear to have multiple independent origins from tetrapolar mating systems (HIBBETT and DONOGHUE 2001). The population genetic consequence of the bipolar vs. the tetrapolar mating system is a difference in the amount of interbreeding permitted between the haploid progeny from a single parent (siblings). Specifically, the potential for inbreeding is higher in the bipolar system because 50% of full-sib progeny are mating compatible, whereas only 25% are in the tetrapolar case. This trend in homobasidiomycetes toward the evolution of increased selfing is similar to the situation in Ascomycete fungi (e.g., YUN et al. 1999) and plants (e.g., TAKEBAYASHI and MORRELL 2001) in which selfing is typically derived from a system of greater outcrossing.
Using the model tetrapolar species Schizophyllum commune and Coprinopsis cinerea (Coprinus cinereus), the A mating-type loci of both species and the B mating-type locus of S. commune were discovered to be composed of two tightly linked subloci, the 
- and ß-subunits (DAY 1960; RAPER et al. 1960). Each unique combination of alleles at the subloci specifies a unique mating type, making the subloci redundant in function. Less information exists on the mating-type loci of bipolar species. One puzzling finding was that attempts to dissect the bipolar mating-type locus into component subloci failed, suggesting a different genetic architecture of the bipolar locus (RAPER 1966).
The frequent evolution of bipolar species suggests that the transition from a tetrapolar mating system to bipolar may have a simple genetic basis. One clue to the genetic mechanism is the absence of documented reversals from a bipolar system back to a tetrapolar one. RAPER (1966) put forward three plausible hypotheses concerning the origin of the bipolar mating system. One hypothesis is based on the observation that primary mutations at either of the mating-type loci often display a self-compatible phenotype, resulting in bipolar mating behavior of normally tetrapolar strains possessing such mutated alleles. Such mutants have been recovered many times both by spontaneous origin and by mutagenesis studies (RAPER 1966). If such self-compatible mating-type alleles reach fixation frequency in a tetrapolar population, the population will be rendered effectively bipolar. A second hypothesis concerns the potential translocation of a chromosomal segment containing one of the mating-type loci into close genetic linkage with the other, leading ultimately to fusion of the two mating-type loci into one nonrecombining region. The suggestion that the bipolar mating-type locus is a single, indivisible locus gives credence to this hypothesis. A final hypothesis is that the function of one of the mating-type loci could be gradually assumed by the other locus. This hypothesis relates to the broadly applicable finding that the A and B mating-type loci control distinct but interconnected roles in the process of dikaryotic growth and fruiting in mushrooms (RAPER 1978).
Although Raper's hypotheses were formulated before any fungal mating-type genes had been cloned, they are equally plausible today. Detailed molecular investigation of the mating-type genes of basidiomycetes has demonstrated that the mating-type genes of the smut fungi and Hymenomycetes are homologous (CASSELTON and OLESNICKY 1998; HISCOCK and KÜES 1999). The A mating-type locus encodes for one or more pairs of homeodomain transcription factors. Each pair is composed of two classes of homeodomain transcription factor proteins, the HD1 and HD2 proteins, which share similarity with the mating-type proteins of Saccharomyces cerevisiae (HISCOCK and KÜES 1999). Heteroallelic but not homoallelic HD1 and HD2 proteins can heterodimerize, creating a transcription unit capable of initiating the A mating-type-specific developmental sequence (KÜES and CASSELTON 1992). The B mating-type locus of the basidiomycetes was shown to encode both small peptide pheromones and pheromone receptors that are believed to be coupled to a trimeric G-protein complex (BROWN and CASSELTON 2001). As with the A locus, pheromones can activate only heteroallelic B locus receptors.
Knowledge regarding the molecular sequence, organization, and function of the mating-type genes allows a reassessment of the manner in which a bipolar mating system might evolve from a tetrapolar one. The nature of the self-compatible mutant mating types of C. cinerea and S. commune has been investigated by DNA sequencing (OLESNICKY et al. 1999, 2000; FOWLER et al. 2001). Self-compatible mutants of the B mating type of C. cinerea were created by single amino acid substitutions in the pheromone receptors that caused either illegitimate interactions with homoallelic pheromone or constitutive activation of the B pathway (OLESNICKY et al. 1999, 2000). For the A mating-type locus of C. cinerea, two primary mutations causing self-compatible phenotypes were investigated and found to be the result of a deletion/recombination event that caused the in-frame fusion of HD1 and HD2 genes from the same A haplotype but from different subloci (KÜES et al. 1994; PARDO et al. 1996). Thus, self-compatible mating types may arise through mutation and provide a simple explanation for the origin of bipolar mating behavior through loss of discrimination by one of the two mating-type loci of a tetrapolar species. However, there is no evidence that self-compatible mating-types are involved in the origin of bipolar mating systems in the basidiomycetes, nor is there any evidence of such alleles in natural populations.
The origin of a bipolar mating system from a tetrapolar one has been addressed only in the smut fungi. In a landmark study, BAKKEREN and KRONSTAD (1994) demonstrated that the bipolar mating-type locus of Ustilago hordei was formed from the fusion of the a and b mating-type loci observed in tetrapolar smut fungi into one nonrecombining mating-type region with two alleles. Thus, genetic linkage through translocation is the best explanation for the origin of bipolar mating in U. hordei, conforming to one of Raper's postulated mechanisms.
Recently, the bipolar mating system of the mushroom Pholiota nameko was characterized using linkage mapping and DNA sequencing (AIMI et al. 2005). These data demonstrated linkage between the mating-type locus and an A mating-type homolog but not a B mating-type homolog. Although intriguing, these results fail to differentiate among the possible genetic mechanisms of evolving a bipolar mating system, particularly because multiple B mating-type homologs exist in most mushroom genomes (this study), and only one B homolog was investigated in P. nameko.
The mushrooms in the genus Coprinus (sensu lato) provide an excellent group with which to study the evolution of mating systems and mating-type genes, because every known mating system is represented by multiple species. Furthermore, the molecular control of mating type in mushrooms has been intensely studied in C. cinerea, simplifying cloning and comparative analyses. In this article, we investigated the genetic architecture of the bipolar mating system of the common wood-decaying fungus Coprinellus (Coprinus) disseminatus. Using knowledge of the mating-type loci in other homobasidiomycetes, we cloned and sequenced DNA regions containing homologs of the mating-type genes. We then used a population genetic approach to determine what changes in the mating-type genes of the hypothetical tetrapolar ancestor might have occurred during its transition to a bipolar system.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONACKNOWLEDGEMENTSLITERATURE CITED |
|---|
Culture isolation and growth:
Homokaryotic strains were derived from wild-collected fruiting bodies, collected in 2000–2001 from within an 15-km radius of the Duke University Campus in Durham, North Carolina (Table 1). One fruiting body was collected per stump or mound. After fruiting bodies were collected in the field or greenhouse, they were placed over aluminum foil for at least 1 hr to collect the dark brown spores. Spores were scraped from the foil into H2O, and a series of dilutions were plated on half-strength Emerson's YpSS (Y/2) nutrient agar plates with 1.5% agar (STEVENS 1974). Following 1–2 days of growth at room temperature (RT), hyphal colonies were inspected under the microscope at 100x to verify that they derived from a single spore. Single-spore isolates were subcultured at least twice on Y/2. Long-term storage was on 1.5% malt extract agar slants at 4°. Also included were three homokaryotic European strains and a Japanese dikaryotic strain (IFO 30972). The Japanese strain was fruited in the greenhouse on a substrate composed of sawdust, rye berries, and soil in a 4:1:1 ratio. Two mating-compatible homokaryons from this dikaryotic strain were isolated for further study.
|
1 cm from each other. Plates were incubated in the dark at RT for 1–2 weeks. The formation of the dikaryon in a genetic cross typically results in vigorous growth from the margins of the paired homokaryons. Large clamp connections become apparent, albeit sparse, at cell junctions at 100x magnification. Finally, after 2 weeks, brownish rhizomorphic strands clearly distinguish pairs that have mated. We verified dikaryotization on the basis of the production of clamp connections in all crosses.
DNA amplification and sequencing:
Mycelium for DNA extraction was prepared by stationary growth of isolates in 2 ml of 1.5% malt extract broth until stationary phase. The mycelium was removed from the broth, rinsed in H2O, and lyophilized. Approximately 50 mg were used for DNA extraction, using a CTAB buffer following the protocol of ZOLAN and PUKKILA (1986).
To study homologs of the putative B mating-type genes, we used degenerate PCR primers to amplify STE3-like pheromone receptors from homokaryotic strains following standard PCR protocols (JAMES et al. 2004a). Primers br1-F and br1-1R were used to amplify two receptors each from C345.1 and TJ01/19.2; in addition, we used primers br2-F and br2-2R to amplify a small STE3-like fragment from strain TJ01/19.2 (see JAMES et al. 2004b for primer sequences). All degenerate PCR amplicons were gel purified using a QIAquick gel extraction kit (QIAGEN, Valencia, CA), ligated into pCR2.1 (Invitrogen, San Diego), and transformed into Escherichia coli strain TOP10 (Invitrogen). Plasmid templates for DNA sequencing were prepared using the QIAprep spin miniprep kit (QIAGEN) and sequenced on both strands using universal forward and reverse M13 primers. Sequencing reactions utilized the BigDye sequencing kit (Applied Biosystems, Foster City, CA) and were analyzed on an ABI3700 DNA sequencer.
We screened our population sample for DNA polymorphisms at gene regions both linked and unlinked to the mating-type locus, using the nondegenerate PCR primers shown in Table 2. All reactions (except A and Aß) were conducted in a similar manner to that used for degenerate PCR with annealing temperatures fixed at 50°. Amplicons were purified using the QIAquick PCR purification kit (QIAGEN) and sequenced as above.
|
Long-distance PCR and amplicon sequencing:
We used long-distance PCR to amplify A and Aß subloci from homokaryons, using the enzyme LA Taq (Takara, Berkeley, CA) following the manufacturer's instructions. The primers for amplification are given in Table 2. The thermocycling parameters used were: initial denaturation at 94° for 1 min; followed by 35 cycles of 94° for 30 sec, 60° for 30 sec, and 72° for 4 min; and finally a 10-min extension at 72°. Amplicons were digested with the enzyme MspI (Promega) to determine allelism, and two subsets (eight of A and three of Aß) were chosen for DNA sequencing. Amplicons were purified using the QIAquick gel extraction kit and ligated into the vector pCR2.1-TOPO (Invitrogen). The resulting plasmids were sequenced using a combination of standard subcloning procedures, using plasmid pUC119 (SAMBROOK et al. 1989) and the GeneJumper kit containing kanamycin or chloramphenicol resistance transposons (Invitrogen). Plasmid sequencing from both ends of the transposon used the primers GJSeq-A3 and GJSeq-B2 for kanamycin and GJSeq-A3 and GJSeq-B4 for chloramphenicol (JAMES et al. 2004b).
Cosmid library construction, screening, and sequencing:
A cosmid library was prepared in the vector SuperCos-Pab1 (BOTTOLI et al. 1999), using the DNA of strain C345.1. This strain has been deposited into the Belgian Coordinated Collections of Microorganisms as MUCL 43037. DNA of C345.1 was prepared from mycelium grown in 1 liter of Y/2 broth under rotary shaking (125 rpm) at RT. Preparation and screening of the library followed the protocol of JAMES et al. (2004b). The library was screened by PCR of bacterial cells for two genes, MIP and CDSTE3.1 (see Table 2 for primer sequences). DNA of all cosmid clones was isolated using the QIAprep spin miniprep kit (QIAGEN). Six overlapping cosmids were sequenced for the MIP/mating-type locus region, using a random shotgun subcloning method involving partial digestion with restriction enzymes (ZHOU et al. 1988). A single CDSTE3.1 positive cosmid, C25.B2.5, was also sequenced using the GeneJumper primer insertion kit with the kanamycin resistance transposon (Invitrogen). Sequencing reactions were accomplished using the BigDye kit and primers GJSeq-A3 and GJSeq-B2. Assembly of sequence traces into contigs was performed with Sequencher v. 4.1 (Gene Codes, Ann Arbor, MI). Gaps in the contigs were filled by primer walking with synthesized oligonucleotides (Operon, Alameda, CA). Approximately 4.0-fold coverage of the 75.5-kb MIP chromosomal region and 3.1-fold coverage of the 41.9-kb CDSTE3.1 cosmid were achieved.
Expression of C. disseminatus genes in C. cinerea:
C. cinerea monokaryon 218 (A3, B1, trp1.1,16, bad; BINNINGER et al. 1987; KÜES et al. 2002) was transformed according to the protocol given by GRANADO et al. (1997). The trp1+ vector pCc1001 (BINNINGER et al. 1987) was used in cotransformations with plasmid pCR2.1-TOPO containing the entire A sublocus from strains TJ00/99.1 and TJ01/16.3 (plasmids 99.1A and 16.3A) and the entire Aß sublocus from strains TJ00/99.1 and TJ00/89.2 (99.1Aß and 89.2Aß). Also used in cotransformation with pCc1001 were C25.B2.5 containing the two putative pheromone receptor genes CDSTE3.1 and CDSTE3.3 and three pheromone genes CDPHB1, CDPHB2.1, and CDPHB2.2 and C25_e1.10, a 10-kb subclone of C25.B2.5 containing CDSTE3.1 and CDPHB1 in vector pZERO (Invitrogen). Cotransformations used 1 µg DNA for each plasmid. Transformants were picked onto minimal medium (GRANADO et al. 1997) and checked for A-regulated clamp cell production (KÜES et al. 1992) and B-regulated subapical peg formation and clamp cell fusion (BADALYAN et al. 2004) under a microscope. Functional expression of pheromone receptor and/or pheromone genes was further analyzed in mating reactions (O'SHEA et al. 1998) on YMG/T medium (GRANADO et al. 1997) with monokaryon PS004-2 (A42, B1; P. SRIVILAI, unpublished data). Fruiting abilities of dikaryons were tested under C. cinerea standard fruiting conditions (GRANADO et al. 1997). A minimum of 20 transformants were analyzed per cotransformation experiment. Control transformations utilized solely pCc1001.
Data analyses:
Identification of genes on the sequenced cosmid clones used homology searching of the GenBank database with the BLASTX algorithm (ALTSCHUL et al. 1997). CDD searches were also used to determine putative conserved domains that helped in the determination of gene function (MARCHLER-BAUER and BRYANT 2004). Searches for the small peptide pheromone genes were attempted using NCBI's ORFfinder (http://www.ncbi.nlm.nih.gov/gorf/gorf.html) and the software GeneMark v. 2.5, using both Schizosaccarhomyces pombe and Saccharomyces cerevisiae models (BORODOVSKY and MCININCH 1993).
Multiple sequence alignments were performed using the alignment editor GeneDoc (NICHOLAS and NICHOLAS 1997). A phylogeny of the STE3-like pheromone receptors was conducted by analyzing the sequences from C. disseminatus with other pheromone receptor sequences retrieved from GenBank. Following exclusion of ambiguously aligned regions, 158 aligned amino acids were used for phylogeny reconstruction. A maximum-likelihood phylogeny was reconstructed using the software PROTML from the MOLPHY package, using 1000 heuristic searches and the JTT-F substitution matrix (ADACHI and HASEGAWA 1996). Support for nodes was assessed using approximate bootstrap probabilities by the RELL method (HASEGAWA and KISHINO 1994). Basic statistics of DNA diversity were calculated using the software MEGA v. 3.0 (KUMAR et al. 2004) for 
and DnaSP v. 3.53 (ROZAS and ROZAS 1999) for S.
|
RESULTS |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONACKNOWLEDGEMENTSLITERATURE CITED |
|---|
When all homokaryotic testers were intercrossed, nearly every mating was successful. We found that the Japanese homokaryons (IFO 30972.16-17) were unable to mate with any of the European or North American isolates, suggesting that the Japanese strains represent a different species or intersterility group. The complete population crossing design allowed us to assign a series of mating types (Table 1) and to estimate the number of total alleles in the species through the observation of mating-type repeats. We use the nomenclature of RAPER (1966) in referring to the sole mating type of the bipolar C. disseminatus as the A mating-type locus. Six mating types were each shared by two strains; one mating type (A5) was recovered from three strains. Using the formula of O'DONNELL and LAWRENCE (1984), 123 mating types were estimated to be in the global population with 95% confidence intervals of 73–254. These data suggest that C. disseminatus has a much larger number of mating types than most of the bipolar species studied to date, the majority of which have allele estimates of 
50 (MURPHY and MILLER 1997).
The pheromone receptors of C. disseminatus are not part of the mating-type locus:
We developed two sets of degenerate PCR primers for amplifying the STE3-like pheromone receptors from homobasidiomycetes. Amplification using primers br1-F and br1-1R on genomic DNA of C. disseminatus strain C345.1 yielded two fragments homologous to STE3-like pheromone receptors in the same amplification reaction. One of them, termed CDSTE3.1, possesses the same four introns as do most homobasidiomycete STE3 receptor genes. The other fragment (CDSTE3.2) was unique among the known STE3 sequences in lacking three of the four introns. Amplification of genomic DNA of strain TJ01/19.2 with primers br1-F and br1-1R also produced two STE3-like amplicons. One of these was very similar to CDSTE3.2, and the other fragment represented a third receptor paralog, CDSTE3.3. Amplification of DNA from strain TJ01/19.2 using PCR primers br2-F and br2-2R also yielded a fourth paralog, CDSTE3.4. Using specific primers or cosmid sequencing, we were ultimately able to recover homologs of CDSTE3.1–CDSTE3.4 from strain TJ01/19.2 and homologs of CDSTE3.1–CDSTE3.3 from strain C345.1, demonstrating at least three receptor paralogs in the two C. disseminatus homokaryons tested.
We have previously used PCR to amplify and genotype a small fragment of MIP from C. disseminatus, and we demonstrated that the MIP gene fragment cosegregates with the mating-type locus of C. disseminatus, using a small progeny array (JAMES et al. 2004a). PCR primers specific to the four STE3-like sequences were designed to test whether these putative pheromone receptors also cosegregated with the mating-type locus. The segregation of three putative pheromone receptors (CDSTE3.1–CDSTE3.3) was examined among the progeny array of parental dikaryon TJ00/38, and all three of these receptors displayed no apparent linkage with the mating-type locus (Table 3). Two of the loci (CDSTE3.1 and CDSTE3.3) displayed complete cosegregation with each other, suggesting that they are closely linked in the genome. Although segregation data for CDSTE3.4 are lacking, one piece of evidence suggests that this gene is not mating-type specific. The PCR primers specific to this gene amplified only DNA of two sibling strains (TJ01/19.2 and TJ01/19.4). These two sibling strains also possess different mating-type alleles but have identical sequences at CDSTE3.4.
|
|
and Aß subunits of the A mating-type locus of C. disseminatus. As seen with other homobasidiomycete mating-type loci, 32 kb upstream from the mating-type genes is the PAB1 gene, encoding for para-amino benzoic acid synthase (KÜES et al. 2001; JAMES et al. 2002). The A mating-type locus region from C. disseminatus was compared with the genome sequence available for the related species C. cinerea (Figure 2). These data show a very conserved gene order between the two species for nearly the entire 75-kb region with the exception of a duplicated pair of putative drug-binding proteins inserted into the C. disseminatus region (BLAC1 and BLAC2). Other genes of known function displaying conserved synteny between C. disseminatus and other homobasidiomycetes (JAMES et al. 2004b; T. Y. JAMES, unpublished data) are RNA polymerase II (RPB2), glycine dehydrogenase (GLYDH), and a putative kinase with similarity to yeast protein YPL109 (YPL109). In summary, we sequenced a large region of the chromosome surrounding the MIP gene, revealing conserved synteny of this region in comparison with a model mushroom species. Importantly, we discovered putative A mating-type genes clustered into an 10-kb region encoding four homeodomain transcription factor genes.
|
|
and Aß subloci in two separate reactions. Restriction digests of the A PCR products revealed extensive DNA diversity in MspI cut sites (Figure 3). Homokaryons with the same mating type as that determined by genetic crosses shared identical or similar restriction digestion patterns, whereas for different alleles it is difficult even to assess the homology of restriction fragments due to excess polymorphism. A similar result was found for the digests of Aß amplicons (data not shown).
|
and three Aß amplicons and sequenced them to determine the pattern of DNA sequence diversity in relation to the functional domains of the proteins they encode. As observed with other mating-type genes, the level of DNA sequence and amino acid diversity between alleles was tremendous. The nine CDA1 and CDA2 sequences could be divided into five sequence types (presumably alleles). Similarity among the five heteroalleles at the N-terminal region of the protein before the homeodomain motif ranged from 53 to 69% for CDA1 and from 45 to 65% for CDA2. In contrast, similarity in the C-terminal region after the homeodomain motif ranged from 64 to 79% for CDA1 and from 89 to 97% for CDA2.
If the genomic region surrounding the homeodomain genes in C. disseminatus was all part of one nonrecombining region, it would be expected that the same forces of balancing selection that promote sequence divergence of mating-type alleles (MAY et al. 1999) would also have a strong effect on increasing polymorphism of the entire region. Thus, we investigated the genetic variation in the genes surrounding the C. disseminatus mating-type locus by sequencing seven non-mating-type gene regions (PAB1, GLYDH, MIP, CDHH, CDRF, YPL109, and RPB2) spaced over an 70-kb region centered around the mating-type locus for our sample of 49 homokaryons (non-Japanese isolates). We used these data to estimate DNA polymorphism () or the average number of pairwise differences per site between two sequences (NEI 1987). The amount of DNA polymorphism () at the mating-type genes was approximately an order of magnitude higher than that of the genes that flank the A locus (Table 4). Variation was also significantly higher (P < 0.05) at the gene regions directly adjacent to the mating-type locus (i.e., MIP, CDHH, CDRF). However, the substitutions in Table 4 include both synonymous and nonsynonymous changes for protein-coding genes such as MIP. We also looked at polymorphism only at silent positions where substitutions have no effect on the encoded proteins (S), such that differences in standing genetic variation should reflect only differences in coalescence time due to balancing rather than to positive selection. The DNA diversity at silent positions was similarly low for all genes outside of the mating-type locus (Figure 4); however, the genes upstream of the A locus generally had a higher level of silent polymorphism than the genes downstream of the A locus. These data suggest that balancing selection on the mating-type loci may have the effect of elevating polymorphism of the neighboring genes, but this effect is greatly reduced over short physical distances, presumably through recombination.
|
|
Searches for possible pheromone genes in the CDSTE3.1 region revealed three genes encoding putative peptide pheromones (CDPHB1, CDPHB2.1, and CDPHB2.2; Figure 5). One pheromone gene is in close proximity to the receptor CDSTE3.1, and two are found in the genomic region surrounding CDSTE3.3. Only two of the three pheromone genes show a significant match with any published pheromone sequences (supplemental Table 2 at http://www.genetics.org/supplemental/), but all proteins were predicted as probable ORFs or exons using the GeneMark algorithm (BORODOVSKY and MCININCH 1993). OLESNICKY et al. (1999) suggested that the conserved ER motif located 11 amino acids N-terminal to the modified cysteine of pheromone protein phb2.2 of C. cinereus was likely to have a functional role in peptide processing. All three putative C. disseminatus pheromones display this pair of amino acids in the positions homologous to the ER motif in C. cinereus phb2.2 and other homobasidiomycete pheromones (FOWLER et al. 2001; RIQUELME et al. 2005).
The region of the C. disseminatus genome containing the two pheromone receptors demonstrates some conserved gene order with the B mating-type locus of C. cinerea (Figure 2). The gene regions from the two species share four genes (PERO, RAB7, CDUP14, and CDUP15) and the pheromone and pheromone receptor genes. However, the amount of gene rearrangement at the B mating-type locus is very high compared to the synteny of the A mating-type loci of C. cinerea and C. disseminatus (Figure 2). Genes with known function from the CDSTE3.1 region include ERG26, encoding a putative dehydrogenase involved in ergosterol biosynthesis, PERO, encoding a protein with strong similarity to lignin degrading peroxidases, and two RAB7 genes, encoding putative GTPases involved in vesicle trafficking.
Evolution of the STE3-like pheromone receptors:
Our analyses of the segregation of the STE3-like pheromone receptors demonstrate that these genes are not part of the mating-type locus. A phylogenetic analysis of the basidiomycete pheromone receptor homologs was used to determine if the STE3-like genes from C. disseminatus actually derive from B mating-type receptors of tetrapolar mushrooms. The amino acid sequences of the pheromone receptors of homobasidiomycetes, heterobasidiomycetes, and two Ascomycete outgroups were aligned together with the four C. disseminatus putative receptors (see Figure 6 legend for GenBank numbers). Only the STE3 domain (pfam02076) sequence region, containing the seven transmembrane helices, was alignable without ambiguity. The maximum-likelihood phylogeny estimated using PROTML is shown in Figure 6. The homobasidiomycete receptors comprised two clades, "groups 1 and 2." CDSTE3.1 groups very closely with the three group 1 receptors S. commune BBR2 and C. cinerea RCB2.6 and RCB1.3. CDSTE3.2 and CDSTE3.3 are also part of group 1 in a clade with C. cinerea RCB3.6 and RCB3.42. Finally, CDSTE3.4 is nested within the receptors of the group 2 clade. These results demonstrate that the pheromone receptors from C. disseminatus have specifically diverged from within the family of mating-type-specific pheromone receptors found in other homobasidiomycetes.
|
The observed DNA diversity () at the two tightly linked receptors CDSTE3.1 and CDSTE3.3 and at the putative pheromone gene CDPHB1 was similar in comparison with that at the genes near, but not part of, the A mating-type locus (Table 4). CDSTE3.2 was identified as a unique receptor-like gene because it lacks three of the four introns observed in all other homobasidiomycete receptors. Variation at CDSTE3.2 suggests that it may be a pseudogene because 3 of 11 sequences contained transcripts predicted to be interrupted by stop codons. Furthermore, the amount of DNA diversity () at CDSTE3.2 was higher than average (0.086) for non-mating-type gene regions (Table 4). For two homokaryotic strains, PCR amplification using specific primers for CDSTE3.2 produced two distinct copies. Separation of the two copies using plasmid subcloning produced two different but closely related sequences. Taken together, the CDSTE3.2 locus appears to be composed of one or two fast-evolving pheromone receptor genes or pseudogenes.
Function of mating-type gene homologs in C. cinerea:
When transformed with A mating-type genes from its native species, C. cinerea monokaryon 218 produces a fluffy mycelium with unfused clamp cells at the hyphal septa. Transformants receiving B mating-type genes show retarded colony growth and reduced production of aerial mycelium on complete medium. Transformants receiving both A and B mating type genes grow fluffy with some septa having fused clamp connections. At other septa, the apical cell displays an unfused clamp while the subapical side produces an enlarged peg (KÜES et al. 1998, 2002; BADALYAN et al. 2004). Phenotypes of strain 218 transformants receiving C. disseminatus constructs 99.1A, 16.3A, 99.1Aß, or 98.2Aß were indistinguishable from phenotypes of transformants with native A mating-type genes (Figure 7A). Functional expression of plasmids carrying C. disseminatus homeodomain transcription factor genes in host monokaryon 218 was efficient, with transformation rates between 40 and 51%, irrespective of coming from the A or the Aß sublocus. Transformants of 218 receiving DNA from clones C25 B2.5 and C25_e.10 containing C. disseminatus pheromone and pheromone receptor genes had normal septa (Figure 7B) and no special growth phenotype on YMG/T complete medium. When C. cinerea monokaryon 218 was cotransformed with the C. disseminatus A and B constructs, between 28 and 42% of transformants had fused clamp cells at some septa and apical unfused clamps and subapical pegs at other hyphal septa (Figure 7, C and D).
|
|
DISCUSSION |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONACKNOWLEDGEMENTSLITERATURE CITED |
|---|
The switch to a bipolar mating system:
Of the three hypotheses put forward by RAPER (1966) for the origin of the bipolar mating system in homobasidiomycetes from the tetrapolar system, the hypothesis concerning a chromosomal translocation placing the A and B mating-type genes in close physical association can be ruled out for C. disseminatus. This evidence comes from our analyses that show that homeodomain transcription factor genes cosegregate with mating type and display a level of DNA polymorphism characteristic of other mushroom mating-type genes. In contrast, no such patterns were observed among the pheromone receptor genes found in the C. disseminatus genome. These results contrast with the findings in the heterobasidiomycete yeast U. hordei (BAKKEREN and KRONSTAD 1994) in which the single mating-type locus is composed of a pheromone gene, a pheromone receptor gene, and a pair of homeodomain genes embedded into a nonrecombining chromosomal segment. Thus, a translocation of the A mating-type locus into the B mating-type region (or vice versa) was suggested by these data. A similar event must have occurred in the ancestor of the heterobasidiomycete yeast Cryptococcus neoformans in which the mating-type locus is a nonrecombining gene-dense region containing many genes important in mating and pathogenesis, including a few pheromone genes, a pheromone receptor gene, and a homeodomain transcription factor gene (HULL et al. 2002, 2005; LENGELER et al. 2002).
Another hypothesis put forward by RAPER (1966) for the origin of bipolar mating systems was that one of the two mating-type loci of a tetrapolar ancestor could mutate to become self-compatible, thus rendering its allelic state meaningless in crosses. The data observed for C. disseminatus are generally consistent with this hypothesis. We found the C. disseminatus homologs of the B mating-type genes of C. cinerea through degenerate PCR and cosmid sequencing. Although the identified pheromone receptors appear to be fully functional on the basis of in silico predictions, they do not show the characteristic hyperpolymorphism associated with mushroom mating-type genes that are under very strong balancing selection (Table 4).
Raper's final hypothesis that the function of one of the mating types could be gradually assumed by the other mating type is not consistent with our data. Using heterologous expression in C. cinerea, we were able to demonstrate that the A and B homologs of C. disseminatus have very similar cellular phenotypes when transformed into C. cinerea as do the respective native C. cinerea genes (Figure 7), suggesting that genetic control of the A and B pathways has been maintained separately in C. disseminatus. At least one of the pheromone receptors must be involved in the same G-protein-coupled B locus pathway as in other tetrapolar homobasidiomycetes because they are able to drive clamp-cell fusion and even fruit body development and sporulation (Figure 7). It remains to be tested whether the receptors or the pheromone genes are constitutively activating or self-compatible mutants or whether they can interact with the native C. cinerea B locus proteins. One line of evidence does suggest that CDSTE3.1 may actually be a self-compatible receptor allele. This protein contains a substitution in the third cytoplasmic loop at position 196 of a phenylalanine for a residue that is invariantly aliphatic (valine, isoleucine, or leucine) in all described C. cinerea, S. commune, and Ustilago spp. STE3 pheromone receptor homologs (data not shown). Moreover, the substitution of glutamine for leucine in precisely the homologous position in S. cerevisiae STE3 causes a partially constitutive and hypersensitive a-factor receptor (BOONE et al. 1993).
The mating-type loci of C. disseminatus and those of the bipolar mushroom P. nameko AIMI et al. (2005) are similar in that both species appear to utilize homeodomain proteins rather than pheromone receptors to determine mating type. The fact that these evolutionary independent lineages may have taken the same course in evolving a bipolar mating system from a tetrapolar one could suggest that loss of B mating-type function is easier than loss of A mating-type function. Such a process could occur through differences in rates of mutation to self-fertility of homeodomain proteins vs. pheromone receptors.
The structure and evolution of the mating-type locus:
C. disseminatus is the first bipolar mushroom species reported to have a mating-type locus composed of more than a single subunit. We have termed the two separate subloci A and Aß to facilitate comparison with the traditionally defined mating-type subloci of C. cinerea that they closely resemble. Both the A and Aß subloci encode a pair of divergently transcribed homeodomain genes (Figure 1); the combination of alleles at the two subloci presumably determines the mating-type specificity of an individual. The total number of mating types in the species was estimated to be 123, using a complete crossing experiment of 49 homokaryotic isolates. This value is rather similar to that observed at the A mating type of C. cinerea, where 120–164 mating types are estimated (RAPER 1966; MAY and MATZKE 1995). We have observed seven A and eight Aß alleles in C. disseminatus on the basis of RFLP patterns with MspI, in small samples of 13 and 16 isolates, respectively (see Figure 3 for a representative gel). These molecular phenotypes suggest a symmetric allele number between subloci. Transformation data using C. cinerea as a host show that genes from both subloci initiate clamp cell formation, consistent with the hypothesis that the subloci are functionally redundant and independently contribute to mating-type specificity.
Although we isolated four pheromone receptors from one haploid strain of C. disseminatus and have demonstrated that they are not part of the mating-type locus, additional copies of the STE3-like pheromone receptors may still exist that were not detected by our PCR-based methods. If such additional receptors exist, then they are not likely to be part of the mating-type locus. Balancing selection can elevate the polymorphism of neighboring genomic regions, but the increase in diversity is a negative function of the recombination distance between the region and the actual target of selection (HUDSON and KAPLAN 1988). For the genomic region surrounding the C. disseminatus A mating-type locus (i.e., MIP, CDHH, and CDRF loci), DNA polymorphism () is >0.02 (Table 4), but at distances >10 kb from the mating-type locus, it appears that sites experience little, if any, elevated DNA polymorphism by linkage to the mating-type locus (Figure 4). Such a contrast in nucleotide diversity between the mating-type locus and the regions bordering it likely reflects recombination that separates those sites subject to strong balancing selection from sites that evolve in a more neutral manner.
An indispensable role for pheromone receptors in homobasidiomycetes:
Four lines of evidence point to a clear origin of CDSTE3.1 and CDSTE3.3 receptors from other homobasidiomycete mating-type-specific pheromone receptors. One line of evidence is that the chromosomal region containing CDSTE3.1 and CDSTE3.3 displays some conserved gene order when compared to the B mating-type locus chromosomal region of C. cinerea (Figure 2). Second, phylogenetic analyses place these proteins among other mating-type proteins of the model mushroom species (Figure 6). Third, the genes exert B mating-type-typical function in a heterologous species (Figure 7). Finally, the receptors appear to be each in close physical proximity with one or two putative peptide pheromones (Figure 5).
If the genome of C. disseminatus contains both active pheromone and receptor genes that are not polymorphic, what cellular function do they perform? As mentioned previously, it is possible that they encode proteins locked into a self-compatible complex, turning on the B-specific developmental pathway through a MAPK cascade, much as their ancestors did for millions of years of fungal evolution. However, it seems quite unnecessary to maintain a functional pheromone/receptor system to turn on a signaling cascade that could be readily turned on by constitutive activation of its downstream partner; e.g., mutations in GPA1 can constitutively activate the pheromone response system in yeast (BANUETT 1998) and mutations in pcc1 can constitutively activate false-clamp cell production and fruit body development in C. cinerea (MURATA et al. 1998). That C. disseminatus has maintained an apparently functional pheromone receptor system suggests that the role of the B mating-type receptors is more complex than activating only the MAPK pathway through G-proteins, that the G-proteins interact with more than the pheromone receptors, or that the receptors of C. disseminatus have taken on a new functional role.
The homobasidiomycete fungi are unique in that they are the only fungal clade to have evolved a multiallelic pheromone/receptor mechanism to perform incompatibility discrimination between individuals—all other fungi have only a biallelic system. The homobasidiomycete pheromone/receptor systems are also possibly unique because these proteins have not been detected extracellularly (BROWN and CASSELTON 2001). It has been suggested that the pheromone/receptor system of mushroom fungi functions in the recognition between the two compatible nuclei of the dikaryotic cell (SCHUURS et al. 1998; DEBUCHY 1999). This function may be indispensable and an intact pheromone/receptor system might be required for proper dikaryon maintenance in all species, bipolar or tetrapolar.
|
ACKNOWLEDGEMENTS |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONACKNOWLEDGEMENTSLITERATURE CITED |
|---|
|
FOOTNOTES |
|---|
|
LITERATURE CITED |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONACKNOWLEDGEMENTSLITERATURE CITED |
|---|
ADACHI, J., and M. HASEGAWA, 1996 MOLPHY version 2.3: programs for molecular phylogenetics based on maximum likelihood. Comput. Sci. Monogr. 28: 1–150.
AIMI, T., R. YOSHIDA, M. ISHIKAWA, D. BAO and Y. KITAMOTO, 2005 Identification and linkage mapping of the genes for the putative homeodomain protein (hox1) and the putative pheromone receptor protein homologue (rcb1) in a bipolar basidiomycete, Pholiota nameko. Curr. Genet. 48: 184–194.[CrossRef][Medline]
ALTSCHUL, S. F., T. L. MADDEN, A. A. SHAFFER, J. ZHANG, Z. ZHANG et al., 1997 Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Res. 25: 3389–3402.
BADALYAN, S. M., E. POLAK, R. HERMANN, M. AEBI and U. KÜES, 2004 Role of peg formation in clamp cell fusion of homobasidiomycete fungi. J. Basic Microbiol. 44: 167–177.[CrossRef][Medline]
BADRANE, H., and G. MAY, 1999 The divergence-homogenization duality in the evolution of the b1 mating type gene of Coprinus cinereus. Mol. Biol. Evol. 16: 975–986.[Abstract]
BAKKEREN, G., and J. W. KRONSTAD, 1994 Linkage of mating-type loci distinguishes bipolar from tetrapolar mating in basidiomycetous smut fungi. Proc. Natl. Acad. Sci. USA 91: 7085–7089.
BANHAM, A. H., R. N. ASANTE-OWUSU, B. GÖTTGENS, S. A. J. THOMPSON, C. S. KINGSNORTH et al., 1995 An N-terminal dimerization domain permits homeodomain proteins to choose compatible partners and initiate sexual development in the mushroom Coprinus cinereus. Plant Cell 7: 773–783.[Abstract]
BANUETT, F., 1998 Signaling in the yeasts: an informational cascade with links to the filamentous fungi. Microbiol. Mol. Biol. Rev. 62: 249–274.
BINNINGER, D. M., C. SKRZYNIA, P. J. PUKKILA and L. A. CASSELTON, 1987 DNA-mediated transformation of the basidiomycete Coprinus cinereus. EMBO J. 6: 835–840.[Medline]
BOONE, C., N. G. DAVIS and G. F. SPRAGUE, JR., 1993 Mutations that alter the third cytoplasmic loop of the a-factor receptor lead to a constitutive and hypersensitive phenotype. Proc. Natl. Acad. Sci. USA 90: 9921–9925.
BOTTOLI, A. P. F., K. KERTESZ-CHALOUPKOVÁ, R. P. BOULIANNE, J. D. GRANADO, M. AEBI et al., 1999 Rapid isolation of genes from an indexed genomic library of C. cinereus in a novel pab1+ cosmid. J. Microbiol. Methods 35: 129–141.[CrossRef][Medline]
BORODOVSKY, M., and J. MCININCH, 1993 GenMark: parallel gene recognition for both DNA strands. Comput. Chem. 17: 123–133.[CrossRef]
BROWN, A. J., and L. A. CASSELTON, 2001 Mating in mushrooms: increasing the chances but prolonging the affair. Trends Genet. 17: 393–400.[CrossRef][Medline]
BULLER, A. H. R., 1924 Researches on Fungi, Vol. III. Longmans Green & Co., London.
BURNETT, J. H., 1975 Mycogenetics. John Wiley & Sons, London.
BUTLER, G. M., 1981 Effects of growth retarding environmental factors on growth kinetics and clamp connection occurrence in the dikaryon of Coprinus disseminatus. Trans. Brit. Mycol. Soc. 77: 593–603.
CASSELTON, L. A., and N. S. OLESNICKY, 1998 Molecular genetics of mating recognition in basidiomycete fungi. Microb. Mol. Biol. Rev. 62: 55–70.
DAY, P. R., 1960 The structure of the A mating-type locus in Coprinus lagopus. Genetics 45: 641–651.
DEBUCHY, R., 1999 Internuclear recognition: a possible connection between euascomycetes and homobasidiomycetes. Fungal Genet. Biol. 27: 218–223.[CrossRef][Medline]
FOWLER, T. J., M. F. MITTON, L. J. VAILLANCOURT and C. A. RAPER, 2001 Changes in mate recognition through alterations of pheromones and receptors in the multisexual mushroom fungus Schizophyllum commune. Genetics 158: 1491–1503.
GRANADO, J. D., K. KERTESZ-CHALOUPKOVÁ, M. AEBI and U. KÜES, 1997 Restriction enzyme-mediated DNA integration in Coprinus cinereus. Mol. Gen. Genet. 256: 28–36.[CrossRef][Medline]
HALSALL, J. R., M. J. MILNER and L. A. CASSELTON, 2000 Three subfamilies of pheromone and receptor genes generate multiple B mating specificities in the mushroom Coprinus cinereus. Genetics 154: 1115–1123.
HASEGAWA, M., and H. KISHINO, 1994 Accuracies of the simple methods for estimating the bootstrap probability of a maximum likelihood tree. Mol. Biol. Evol. 11: 142–145.
HIBBETT, D. S., and M. J. DONOGHUE, 2001 Analysis of character correlations among wood decay mechanisms, mating systems, and substrate ranges in homobasidiomycetes. Syst. Biol. 50: 215–242.[Medline]
