Abstract
The B mating type locus of the basidiomycete Coprinus cinereus encodes a large family of lipopeptide pheromones and their seven transmembrane domain receptors. Here we show that the B42 locus, like the previously described B6 locus, derives its unique specificity from nine multiallelic genes that are organized into three subgroups each comprising a receptor and two pheromone genes. We show that the three genes within each group are kept together as a functional unit by being embedded in an allele-specific DNA sequence. Using a combination of sequence analysis, Southern blotting, and DNA-mediated transformation with cloned genes, we demonstrate that different B loci may share alleles of one or two groups of genes. This is consistent with the prediction that the three subgroups of genes are functionally redundant and that it is the different combinations of their alleles that generate the multiple B mating specificities found in nature. The B42 locus was found to contain an additional gene, mfs1, that encodes a putative multidrug transporter belonging to the major facilitator family. In strains with other B mating specificities, this gene, whose functional significance was not established, lies in a region of shared homology flanking the B locus.
TWO sets of multiallelic mating type genes, A and B, determine mating compatibility in the homobasidiomycete fungus Coprinus cinereus. A successful mating occurs only between cells that have different alleles of genes at both loci, converting an asexual mycelium known as the monokaryon into a sexually fertile dikaryon. The dikaryon is a prolonged mycelial stage in which the nuclei from each mating partner do not fuse but remain together in each cell. Both sets of mating type genes play a role in the developmental sequence that leads to the formation of the dikaryon and to the subsequent maintenance of the binucleate cells that characterize it (see Casselton and Olesnicky 1998).
The A mating type genes of homobasidiomycetes encode two dissimilar proteins belonging to the homeodomain family (Kües and Casselton 1992; Küeset al. 1994; Stankiset al. 1992). A compatible mating brings together different versions of these proteins that heterodimerize to generate the active transcription factor complex that initiates A-regulated steps in dikaryon development (Banhamet al. 1995). The B genes encode peptide pheromones and their cognate seven-transmembrane domain receptors (Wendlandet al. 1995; Vaillancourtet al. 1997; O'Sheaet al. 1998). Pheromone signaling apparently plays no role in mate attraction in C. cinereus as it does in yeasts (Kurjan 1993) and hemibasidiomycete species (Bölkeret al. 1992; Spelliget al. 1994), but it is evident from studies on B-regulated development that it acts after cell fusion to promote the formation and maintenance of the dikaryon.
Following cell fusion in C. cinereus, a compatible complement of B genes promotes an exchange of nuclei and the migration of donor nuclei through the established cells of each recipient monokaryon. Once the tip cells contain nuclei from both mates, a compatible complement of A genes leads to nuclear pairing, the synchronization of nuclear division, and the initiation of a complex tip cell division that involves the formation of a structure known as the clamp connection. Following nuclear division, one daughter nucleus becomes cut off in the developing clamp cell and the other in a newly formed subterminal cell. Compatible B genes allow the completion of the clamp connection by promoting fusion between the clamp cell and the subterminal cell.
A remarkable feature of the A and the B loci is that they both contain several multiallelic genes. The A locus of C. cinereus contains representatives of three paralogous pairs of genes (Pardoet al. 1996) that encode the two classes of A proteins. Different allele combinations of the paired A genes occur in nature (Lukenset al. 1996; Pardoet al. 1996) and generate an estimated 160 A mating specificities (Raper 1966). Our analysis of a B locus suggests that it also derives its many specificities, an estimated 79, from three sets of paralogous genes. In the B6 locus we identified nine genes arranged into three groups, each encoding a receptor and two pheromones (O'Sheaet al. 1998). The B6 and B3 loci were found to have different alleles of all nine genes, but B6 and B42 were predicted to share the alleles of one group of genes on the basis of DNA homology and failure of this group of B6 genes to activate B-regulated development in B42 hosts. In this report we describe the sequence of the B42 locus, confirm our prediction of shared genes in B6 and B42, and provide more extensive evidence to support our model for B locus organization in C. cinereus. We demonstrate that the cloned genes from each group are found together as a functional unit in different B loci and that compatible B mating specificities require that only one group of genes have different alleles.
MATERIALS AND METHODS
Fungal strains and growth conditions: C. cinereus strains used in this study were: A42B42: JV6 wild type and LN118 trp-1.1,1.6 ade-2; A6B42: PG78 pab-1 trp-1.1,1.6; A6B6: H9 wild type and LT2 trp-1.1,1.6; A5B6: FA2222 trp-1.1,1.6; A2B3: LCO12 trp-3; A6B3: PR94226 ade-5 cho-1; A43B43: AT8 ade-8 trp-3; A42B43: JV56 wild type; A5B5: HT8 trp-1.2; A6B5: TC10 wild type; A3B1: 218 trp-1.1,1.6; A2B1: 68 wild type; A40B40: LCO7 trp-1.1,1.6; A42B40: LCO7R wild type; A41B41: LCO5 trp-1.1,1.6; A42B41: LCO5R wild type; A44B44: NL1 trp-1.1,1.6; A42B44: NL1R wild type. Media and methods for culturing C. cinereus were described by Lewis (1961), with modifications summarized by Mutasa et al. (1990). Hosts for transformation were LT2, LCO12, and LN118. B6 genes were introduced by cotransformation with plasmids pCc1001 or pDB1 containing the C. cinereus trp-1 (Binningeret al. 1987) or trp-3 gene (Burrows 1991), respectively. Transformation was performed as described by Casselton and de la Fuente Herce (1989). Routinely, 50 transformants were tested for the expression of an introduced B gene. B gene expression was assayed by the method of O'Shea et al. (1998), which involved mating transformants to a tester strain having a different A mating specificity from the host but the same B mating specificity. If the introduced gene activated the pheromone response pathway in the host, this led to a compatible mating interaction and formation of a dikaryon. Frequency of cotransformation was variable, but a positive mating reaction was always detected in 12–50% of transformants.
DNA and RNA procedures: For Southern analyses, DNA was isolated by the small-scale method of Zolan and Pukkila (1986) and the analysis was carried out as described by Mellon et al. (1987). DNA sequencing was carried out on an ABI 373 (Perkin-Elmer, Norwalk, CT) automated sequencer. Sequence analysis was performed as described by O'Shea et al. (1998). RNA was isolated by the method of Hoge et al. (1982). Poly (A)+ RNA was purified from total RNA by oligo(dT) cellulose affinity chromatography (Edmondset al. 1971; Aviv and Leder 1972) or by the PolyATtract mRNA Isolation System (Promega, Madison, WI) following the manufacturer's instructions.
Plasmids and PCR strategies: The B42 locus was isolated from a cosmid genomic library constructed from JV6 wild-type DNA cloned into the cosmid vector LORIST 2 (Mutasaet al. 1990). Routine cloning was in pBluescript II (Stratagene, La Jolla, CA) and plasmid amplification was in Escherichia coli strain XL1Blue or DH5α. PCR products were purified by agarose gel electrophoresis and were cloned into pGEM-T and pGEM-T Easy (Promega). Plasmids containing B6 genes were described by O'Shea et al. (1998). Plasmids containing all but one of the B42 genes were obtained by cloning the following restriction fragments: pJH3, on a 1.7-kb HindIII (phb1.142); pJH38, 0.9-kb EcoRI (phb1.242); pJH31, 2.2-kb PstI-SalI (phb2.142); pJH25, 1.8-kb HindIII-EcoRI (phb2.242); pJH30, 4.0-kb PstI-SalI (rcb242); pJH26, 1.0-kb HindIII-PstI (phb3.142); pJH10, 1.4-kb HindIII-EcoRI (phb3.242), and pJH33, 2.1-kb EcoRI (rcb342). rcb142 was amplified by the polymerase chain reaction (PCR) using the primers 5′-CCCCGACGGCCTTGTACTGTAGC-3′ and 5′-GGTCCGGGAGCAGAGCGAG-3′. The group 3 and group 1 genes from B3 were amplified from genomic DNA of strain LCO12 using the GeneAmp XL-PCR Kit (Perkin-Elmer) and the primers 5′-GGCAGGTCTAAAGGTAGCCACG-3′ and 5′-CCACAAGGGCAGTGACTTGCCTTC-3′ for group 3, and 5′-ATCGTGGGAAACGCAGATGGCG-3′ and 5′-CGTTCTGTTCGCCCAGTTGTGAC-3′ for group 1. The initial PCR fragment containing the B3 group 3 genes also contained phb2.23. A fragment containing the group 3 genes only was amplified using the following primers: 5′-GGTTGTCACAGACAGAGAAGG-3′ and 5′-GCCTCAAAGCTTGACGAGTCCG-3′. cDNAs for the rcb2 and rcb3 receptor genes and the mfs1.1 gene were obtained by RT-PCR using the Access RT-PCR Kit (Promega) and the following primers: 5′-TGCGCTTTCCTCGCTGCCATCCTA-3′ and 5′-GACGTTACTCGTGAATGCACTCG-3′ for rcb2, 5′-TCCTCGTCACGAAAGCTGTCCTTC-3′ and 5′-GACTTTTGAGTGTGCGTTAGAGTCG-3′ for rcb3, and 5′-CGTTCTGACGTTCTTTGGTATGTC-3′ and 5′-AGATGAAGAACATGGAGTGATCAGG-3′ for msf1.1. RNA was extracted from strain LN118 in which receptor gene transcription was induced by transformation with a plasmid containing genes from the B6 locus (Halsall 1997).
Sequence data accession numbers: GenBank accession numbers are as follows: AF186386, phb1.142; AF186387, phb1.242; AF186390, phb2.142; AF136522, phb2.242; AF186388, phb3.142; AF186390, phb3.242; AF186383, rcb142; AF186384, rcb242; AF-186385, rcb342; AF186391, mfs1.1; and AF186392, mfs1.2.
RESULTS
Isolation and sequence analysis of the B42 genes: A cosmid genomic library constructed from the B42 wild-type strain JV6 was screened for clones containing sequences from B42 using a homologous flanking sequence from the B6 locus as probe. We used Southern blot analysis to show that several of the clones identified contained DNA sequence that was not present in the genomic DNA of the B6 strain. Since different alleles of the B genes do not cross-hybridize (O'Sheaet al. 1998), B42-specific sequences would be expected to contain B alleles not present in B6. B42-specific sequences adjacent to the common flanking sequence were used to isolate other cosmid clones and to initiate a walk that spanned 60 kb of contiguous DNA sequence, sufficient to contain the entire B42 locus.
Four cosmids, cJH4, cJH5, cJH8, and cJH10, which together covered the entire 60-kb sequence (Figure 1), were tested for B gene function by transformation into two host strains, one having a B6 mating specificity and one a B3 mating specificity. Monokaryons expressing compatible B genes in C. cinereus have no easily recognized phenotype and B gene function was assayed by a mating test described by O'Shea et al. (1998) and detailed in materials and methods. cJH5 and cJH8 contained genes that altered the B mating specificity of both host strains tested, whereas cJH4 contained sequences that only altered B mating specificity in the B3 host. This result would be consistent with cJH4 having genes that we showed previously to be the same in B42 and B6 and cJH5 and cJH8 having additional genes whose alleles are unique to B42. The sequence present in cJH10, though largely B42 specific, failed to alter the B mating activity of either host.
The genomic DNA sequence of the B42 locus is covered by the four overlapping cosmids, cJH4, cJH5, cJH8, and cJH10. The physical map was generated using the restriction enzymes BamHI (B), EcoRI (E), HindIII (H), PstI (P), and SalI (S). The 10 genes identified in the sequence are illustrated as boxes below the map, with arrows indicating the direction of transcription. The three groups of genes determining B specificity are differently shaded, and each encodes two pheromone precursors and a pheromone receptor. The relative positions of these 9 genes within the previously characterized B6 locus and 6 genes in the partially characterized B3 locus are also illustrated. Dashed boxes and lines indicate genes and sequences detected by hybridization only. Solid boxes represent the homologous sequences flanking the B loci. The mfs1 gene encodes a putative multidrug transporter.
To identify the sequence corresponding to the B42 locus we used cloned DNAs to compare the genomic sequences in B42, B6, and B3 strains (data not shown). The homologous borders of the B42 locus were found to flank a 29-kb sequence and a physical map of this is presented in Figure 1. Hybridization analyses identified a 7-kb sequence adjacent to the right border of the B6 locus that was present in B6 and B42 but not B3, a 6-kb sequence that was present in B42 and B3 but not B6, and a 5-kb sequence at the left side of the B6 locus that was different in all three loci. In addition, the B42 locus was found to contain 10 kb of sequence that did not cross-hybridize to B6 or B3 genomic DNA and did not contain any genes that altered B mating specificity in transformation tests.
We predicted previously that the nine genes identified in the B6 locus belong to three functionally independent or paralogous groups, each comprising a receptor and two pheromone genes (O'Sheaet al. 1998). The cross-hybridization observed in our Southern analyses is consistent with this interpretation and indicates that B6 and B42 share alleles of the three group 1 genes, B3 and B42 share alleles of the three group 2 genes, and all three specificities have different alleles of the three group 3 genes. According to this interpretation, we started our sequence analysis of B42 at the right of the physical map illustrated (Figure 1) where the common B6 and B42 genes would be found.
We showed that the B42 specificity is determined by nine genes, six encoding pheromone precursors, and three encoding pheromone receptors within an 18-kb nucleotide sequence. The genes were organized, like those in the B6 locus, into three groups each comprising a receptor and two pheromone genes. Significantly, the genes predicted to constitute group 1 were located in homologous sequence shared with B6, the genes predicted to constitute group 2 were located in the region of homology with B3, and the third group of genes was located in the adjacent 5-kb sequence, which is nonhomologous in all three loci. A comparison of the B6, B42, and B3 loci and the relative positions of the genes within each are provided in Figure 1.
The DNA sequences of the three group 1 genes from the B6 and B42 loci were found to be >97% identical, indicating that these genes are homoallelic and confer the same mating specificity. All three genes were found to be embedded in a highly conserved sequence present in B6 and B42 extending over the 7 kb of homology detected by hybridization analyses. The three genes in group 2 were located in the 6-kb sequence shown to be homologous in B3 and B42 but not in B6. Comparison of the group 2 genes of B42 and B6 showed less conservation in sequence than the group 1 genes, thus indicating allelic differences between the group 2 genes of these specificities. The DNA sequence surrounding the group 2 genes was even less conserved between B6 and B42 than that of the genes themselves. This lack of conservation in sequence would act to prevent recombination and would ensure that the three group 2 genes remain together as a functional unit. The group 3 genes and the sequences surrounding them were similarly unconserved between B6 and B42, indicating that these two loci also have different alleles of these genes.
Comparison, presented as percentage identity/similarity, of predicted C. cinereus receptor proteins and those of another hymenomycete, S. commune
The C. cinereus pheromone receptors are predicted to have seven-transmembrane domains and to show homology to the Ste3p a-factor receptor of Saccharomyces cerevisiae and the Pra1 and Pra2 receptors encoded in the a mating type locus of the hemibasidiomycete Ustilago maydis (Bölkeret al. 1992; O'Sheaet al. 1998). A comparison of the proteins encoded by the B42 and B6 genes is summarized in Table 1. Overall, the three B42 receptors have only 35–53% sequence identity and show as little similarity to each other as they do to corresponding receptor proteins encoded by the B mating type genes of another homobasidiomycete species, Schizophyllum commune (Table 1). Comparisons with the C. cinereus B6 proteins showed that the homoallelic Rcb142 and Rcb16 receptors were 99% identical. Heteroallelic receptors showed less conservation in sequence, but this is clearly variable with Rcb26 and Rcb242 displaying only 37% identity but Rcb342 and Rcb36 showing as much as 78% identity.
The pheromones of C. cinereus belong to the family characterized by a-factor of S. cerevisiae (reviewed by Bölker and Kahmann 1993; Vaillancourt and Raper 1996). The predicted sequences of the B42 pheromone precursors are presented in Table 2. Mature peptide pheromones from other basidiomycete species and from yeasts range in size from 9 to 15 amino acids. All the pheromone precursors encoded by the B42 genes have the C-terminal CaaX motif (CIVA, CVVS, CIIA) that signals C-terminal truncation and carboxymethylation and farnesylation of the cysteine residue. In five of the sequences, there is a conserved amino acid pair, ER, 12–14 amino acids N-terminal to the cysteine residue of the CaaX motif which we predicted to be an internal processing site (Casselton and Olesnicky 1998). Recently we have provided direct evidence to support this prediction by demonstrating that a synthetic pheromone based on the Phb2.242 sequence can activate a compatible C. cinereus receptor expressed heterologously in S. cerevisiae (Olesnickyet al. 1999). A possible alternative motif is a DQ found in the Phb3.1 precursor. Based on this prediction, the group 1 and group 2 genes from B42 encode mature peptides of 12 amino acids, and the members of each pair differ in only 3 amino acids. In contrast, the two group 3 pheromones show less sequence similarity and one is predicted to have 12 amino acids and the other 13.
Different B loci derive their specificities from multiple alleles of three groups of genes: Using a combination of Southern blot analyses and transformation with cloned genes, we set out to identify whether other B loci in our collection shared alleles of genes that we had identified in B42 and B6. To increase the scope of this analysis, we exploited the fact that B42 and B3 contain the same alleles of the group 2 genes to isolate fragments containing the group 1 and group 3 genes from B3 by PCR. Two pairs of PCR primers were designed to amplify the B3 genes. One primer was designed to anneal to the shared sequence to the left of the B6 and B3 loci and another to the phb2.242 allele shared by B42 and B3 (Figure 1). These primers amplified an 8.5-kb sequence predicted to contain the B3 group 3 genes and preliminary sequence data have confirmed that these include a receptor gene and two pheromone genes. Another primer was designed to anneal to the homologous sequence flanking the border to the right of the B locus in all strains and another to the rcb242 allele shared by B42 and B3. These primers amplified a 9-kb sequence predicted to contain the B3 group 1 genes. Sequence analysis has identified only two genes, one encoding a receptor and one a pheromone (Figure 1).
Pheromone precursors encoded by the B42 genes
Genomic DNAs from strains having B1, B3, B5, B6, B40, B41, B42, B43, and B44 mating specificities were digested with HindIII and following Southern blotting, were probed with each of the nine individual genes cloned from B42 and B6. Probes to the group 1 and group 3 B3 genes were excised from the sequences generated by PCR. The results of these hybridization analyses together with transformation tests using the same cloned genes are summarized in Tables 3 and 4. Where we detected hybridization in Southern blots, all three genes within a particular group cross-hybridized and none was able to activate B-regulated development in the host to which hybridization occurred. We conclude that hybridization identifies homoalleles. Where we detected no cross-hybridization, generally all three genes from that group activated B-regulated development in the host tested. The only exceptions to this rule involved heteroallelic pheromone genes that could activate receptors found in some specificities but not others (Table 4).
Our previous transformation analysis of B6 showed that a single compatible pheromone or receptor gene is sufficient to activate the B pathway in a compatible host cell. In this study we found that, in some cases, only one of the two pheromone genes within a group was able to activate B-regulated development in a particular host. Table 3 summarizes the overall specificity of a particular group and data in Table 4 record the specificities of individual pheromone genes. These data allow us to deduce which B specificities have homoalleles of genes we have cloned and this is summarized schematically in Figure 2. We conclude that there are only two alleles of each of the group 1 genes in the nine B loci tested, four had the alleles present in B3, and five had those present in B42 and B6. The group 3 alleles present in B6 were found in B40 and B41, and those from B3 were found in B1 and B44. The only loci that shared group 2 alleles were B42 and B3. We conclude that, for compatibility, B loci need to have different alleles of genes in just one of the three groups; B1 and B44 share the same alleles of the group 1 and group 3 genes, yet they confer different B mating specificities and must, therefore, have different alleles of their uncharacterized group 2 genes. A similar argument applies to B40 and B41. These data are consistent with our prediction that each B mating specificity is derived from a unique combination of alleles of all three sets of genes.
The mature pheromones predicted to be encoded by each group of B42 or B6 genes in some cases have very similar sequences, differing in just four amino acids (Table 2; O'Sheaet al. 1998). In other cases the two pheromones have apparently unrelated sequences, and showed a different spectrum of activity in the host strains tested. This is illustrated by the data presented in Table 4. The B42 group 3 pheromones encoded by the phb3.1 and phb3.2 genes are highly diverged in sequence (Table 2). One pheromone (Phb3.142) activated B-regulated development in B1, B3, B5, B43, and B44 hosts whereas the other (Phb3.242) activated development in B6, B40, and B41 hosts. The corresponding B6 group 3 pheromones are also dissimilar in sequence (O'Sheaet al. 1998), and although it is not possible to quantify the activity of a gene using our mating test, it was apparent that the mating responses they induced in different hosts varied in strength. When tested in transformation, phb3.16 induced a much stronger mating response in B1, B3, B5, and B44 hosts than phb.3.26, and most notably phb3.26 failed to activate development in the B43 host. No differences in host specificity were observed with either the B42 group 2 or B42 group 1 pheromones that have similar sequences (Table 2) but a little variation was displayed by the corresponding B6 pheromones that have unrelated sequences (O'Sheaet al. 1998; Table 4). Perhaps more surprising is the fact that, overall, the group 2 pheromones encoded by both the B6 and the B42 genes have the same activity spectrum yet they differ markedly in sequence.
Southern blot and DNA-mediated transformation analyses identify shared alleles in nine different B specificities of C. cinereus
The B42 locus contains a gene encoding a membrane transporter belonging to the major facilitator family: A 10-kb B42-specific DNA sequence adjacent to the group 3 genes did not activate B-regulated development in either a B6 or a B3 host. Northern analysis (not shown) identified a single gene in this region with a transcript of 2.5 kb. The sequence from genomic DNA and an RT-PCR-derived cDNA identified the gene product as a putative membrane transporter belonging to the major facilitator family. The presence of this gene (mfs1.1) within the B42-specific sequence prompted us to look for a corresponding gene in B6 genomic DNA (Figure 1). This gene, mfs1.2, was detected in a region adjacent to the group 3 genes of B6 in a sequence that lacks homology to B42 DNA but shows homology to a corresponding sequence in B3 and seven other genomic DNAs derived from strains with the different B mating specificities listed in Table 3. The DNA sequences of mfs1.1 and mfs1.2 are only 60% identical, which accounts for the lack of cross-hybridization in Southern blots, but the predicted proteins show 72.5% identity (83% similarity).
Activity spectra of pheromone gene pairs in different hosts
Identification of shared alleles in nine different B loci of C. cinereus derived from hybridization and transformation analysis with cloned genes from B42, B6, and B3. Homoalleles are represented by similarly patterned boxes. Uncharacterized alleles are shown as open boxes.
DISCUSSION
Paralogous sets of genes confer multiple B mating specificities: We have shown that the genes within the B42 locus of C. cinereus, like those of the previously described B6 locus, are arranged into three discrete sets each comprising two pheromone genes and one receptor gene. The three genes comprising each group are kept together as an indivisible unit by being embedded in allele-specific DNA sequences. Using a combination of Southern blot analysis and transformation, we have identified shared alleles of genes in nine B loci and demonstrated that variation in only one set of alleles is the minimum requirement to generate different B mating specificities. We have shown previously that the A locus of C. cinereus derives its many specificities from three pairs of paralogous genes (Pardoet al. 1996), in this case genes encoding the two components of a heterodimeric transcription factor. By making use of natural deletions, we were able to introduce different combinations of genes into a partially A-null background and to demonstrate unambiguously that only different alleles from the same pair of genes were able to generate compatible protein combinations. No comparable B-null background exists for us to test individual combinations of receptor and pheromone genes but the genetic data we have presented are consistent with the B genes belonging to three functionally redundant and paralogous groups, which are closely linked.
The 79 versions of this locus predicted to occur in nature (Raper 1966) can be generated with just a few alleles of each of the three paralogous groups of genes. We have identified 2 alleles of the group 1 genes and at least 4 alleles of the group 3 genes in the nine B loci we examined. With just these few alleles of the group 1 and the group 3 genes it would require only 8–10 alleles of the group 2 genes to generate 79 (2 × 4 × 10) unique combinations of the three paralogous sets of genes. It would thus require a total family of some 16 different receptors and 30 different pheromones to generate all the predicted B mating specificities found in nature.
In the only other mushroom that has been studied, S. commune, multiple B mating specificities are derived from just two sets of paralogous genes. The genes are separated into two discrete loci known as Bα and Bβ by as much as 4.5 map units. Genetic recombination studies established the functional redundancy of these two loci and the fact that there were nine specificities of each (Raper 1966). Molecular analyses have revealed that the Bα1 and Bβ1 loci each encode a single receptor and at least three pheromones (Wendlandet al. 1995; Vaillancourtet al. 1997). There is no evidence of shared alleles of receptor or pheromone genes in any of the different specificities.
The evolution of multiple alleles of pheromone and receptor genes in hymenomycetes has necessitated a remarkable degree of specificity in pheromone-receptor recognition. A single receptor may be activated by several pheromones, and each pheromone may activate several receptors. Perhaps it is not surprising to find that multiple pheromones are required to activate the many different receptors. Where receptor sequences are similar, only slight differences in pheromone sequence may be sufficient to distinguish a compatible from an incompatible target. For example, the C. cinereus receptors encoded by the rcb3 genes present in B42 and B6 (Rcb36 and Rcb342) are 78% identical in sequence. They are distinguished by the pheromones encoded by the B42 and B6 alleles of phb3.2 (Phb3.242 and Phb3.26), which in their predicted mature version differ in just four amino acids (ERRTHGGNGLTFWC and ERRTQGGGGLTWFC, respectively, this study and O'Sheaet al. 1998). The two group 3 pheromones encoded in B42 (Phb3.142 and Phb3.242) are completely different in sequence, and do not activate the same receptors (Table 4). Phb3.142 activates Rcb33 and preliminary sequence data for the B3 genes indicate that this receptor has only 38% identity to Rcb36. However, pheromones with quite different sequences may activate the same receptor, as occurred when the B42 and B6 group 2 pheromones were tested in several different backgrounds. They do not always activate development to the same extent but this is impossible to quantify in the C. cinereus mating assay. In S. commune, where transformants with an activated B-pathway have a recognizable phenotype, attempts were made to assess the response to different pheromones and this was seen to be variable (Wendlandet al. 1995; Vaillancourtet al. 1997). In this highly complex system, multiple genes for pheromones would seem to be necessary to extend the range of receptors that can be activated by an allelic set of genes and, where only weak receptor activation is possible, to allow two or more pheromones to act together to maximize the level of activation.
A role for a membrane transporter in pheromone activity: Because gene disruptions are difficult to effect in C. cinereus, we cannot at present say whether Mfs1, the predicted multidrug transporter, has a role in mating in C. cinereus. The mating pheromones of the basidiomycetes belong to the CaaX-modified class typified by a-factor of S. cerevisiae. Processing of the 36- and 38-amino acid yeast a-factor precursor molecule has been well documented and leads to a mature pheromone of 12 residues with the C-terminal cysteine carboxymethylated and farnesylated (Caldwellet al. 1994; Tamet al. 1998). Then, this peptide is actively transported across the cell membrane by the Ste6p protein, a member of the ATP-binding cassette (ABC) family of transporters. Msf1 belongs to a different family of transporters, the major facilitator superfamily (MFS). The association of mfs1 with the B locus may, therefore, be accidental rather than being required for pheromone secretion.
Genes of unrelated function can become trapped accidentally in the long stretches of nonhomologous DNA that one finds at a locus such as B. There are, for example, two genes of unknown function in the corresponding a2 locus of U. maydis that apparently have no relevance to mating (Urbanet al. 1996). Like other alleles of the C. cinereus gene, mfs1-1 may have resided originally in a homologous sequence flanking the B42 locus and may have become translocated into it by a DNA rearrangement. Flanked by B42-specific DNA, its chances of being recombined into other genomic backgrounds would be reduced, there would be no selection against nucleotide substitutions, and the gene would become fixed in its current position.
Acknowledgments
We thank Professor Ian Connerton for providing DNA sequencing facilities and Suzanne O'Shea, Pushpalata Chaure, and Heike Koch for help with the DNA sequence analysis. J.R.H. thanks the Biotechnology and Biological Sciences Research Council (BBSRC) for a research studentship, M.J.M. was supported by the Gatsby Technical Education Project by the award of a Sainsbury Studentship, and L.A.C. is a BBSRC Senior Research Fellow.
Footnotes
-
Communicating editor: R. H. Davis
- Received September 20, 1999.
- Accepted November 15, 1999.
- Copyright © 2000 by the Genetics Society of America
