Scooter, a New Active Transposon in Schizophyllum commune, Has Disrupted Two Genes Regulating Signal Transduction
Thomas J. Fowler, Michael F. Mitton

Abstract

Two copies of scooter, a DNA-mediated transposon in the basidiomycetous fungus Schizophyllum commune, were characterized. Scooter is the first transposon isolated from S. commune. Scooter creates 8-bp target site duplications, comparable to members of the hAT superfamily, and has 32-bp terminal inverted repeats. Both copies of scooter are nonautonomous elements capable of movement. Southern blot hybridizations show that scooter-related sequences are present in all S. commune strains tested. Scooter-1 was identified initially as an insertion in the Bβ2 pheromone receptor gene, bbr2, leading to a partial defect in mating. Scooter-2 spontaneously disrupted a gene to produce the frequently occurring morphological mutant phenotype known as thin. The scooter-2 insert permitted cloning of the disrupted gene, thn1, which encodes a putative regulator of G protein signaling (RGS) protein. Spontaneous insertion of scooter into genes with identifiable mutant phenotypes constitutes the first evidence of active transposition of a DNA-mediated transposon in a basidiomycete.

DNA-MEDIATED transposons from the hAT, Fot1/pogo, and Tc1/mariner superfamilies have been identified in the filamentous fungi (for review, see Kempken and Kück 1998). Members of the hAT (hobo/Ac/Tam3) superfamily have characteristic 8-bp target site duplications (TSD) and similarities in amino acid sequence among their putative transposases (Calviet al. 1991). Until now, the sole hAT-like transposon identified in the basidiomycetous fungi was Abr1 from the button mushroom, Agaricus bisporus (Sonnenberget al. 1999). Abr1 does not encode its own transposase and is atypical among the hAT transposons with its 7-bp TSD and the possibility of a preferred sequence motif into which it integrates. Abr1 has not been observed to move and appears stably inherited in cultivated A. bisporus.

Active transposition of hAT family transposons in filamentous fungi has been demonstrated in several ways. Transposon trapping, a technique by which a known gene is disrupted via insertion of an active transposon, was used to identify Folyt1 from Fusarium oxysporum f.sp. lysopersici (Gómez-Gómezet al. 1999). In the hyphomycete Tolypocladium inflatum, the transposon restless was shown to be active by its spontaneous insertion into tnir, a previously uncharacterized gene involved in nitrogen metabolism (Kempken and Kück 2000). Mutant phenotypes that revert to wild type with high frequency can also be indicative of transposon activity. Excision of a transposon from an insertionally inactivated gene can lead to restoration of gene function, resulting in an unstable, or mutable, phenotype (McClintock 1950). The hAT transposon Ascot-1 of Ascobolus immersus was isolated from a mutable allele of b2, a spore color gene (Colotet al. 1998). Folyt1, in addition to being captured by a transposon trap, was shown to cause an unstable mutant phenotype due to occasional excision from the nit1 gene (Gómez-Gómezet al. 1999).

We describe in this article an insertional mutation that was identified through a mating defect in the homobasidiomycetous fungus Schizophyllum commune (Raper and Raper 1973). An early step during mating in S. commune is the regulated exchange of nuclei between partners, a process controlled by two linked mating-type loci, Bα and Bβ (Raper 1966; Koltin and Flexer 1969). Each of these loci has nine versions, or specificities, defined by a combination of genes encoding a single, unique G protein-coupled pheromone receptor and several unique lipopeptide pheromones (Koltinet al. 1967; Stamberg and Koltin 1972; Wendlandet al. 1995; Vaillancourtet al. 1997). Pheromones control donation of fertilizing nuclei from one mate to the other while pheromone receptors control acceptance of nuclei from a mating partner during the process of nuclear migration (Wendlandet al. 1995; Vaillancourtet al. 1997; Fowleret al. 1998). The pheromones and pheromone receptor encoded within a single B specificity are normally incapable of interacting to trigger nuclear migration.

We applied molecular techniques to the study of several mutations that alter mate recognition during Bβ-dependent matings. These mutations were generated in previous studies and had been mapped in or near specificity 2 of the Bβ locus, Bβ2 (Parag 1962; Raperet al. 1965; Raper and Raudaskoski 1968; Raper and Raper 1973). Among the mutants was one of a group historically called primary B-on mutants in which the B-regulated pathway of sexual development is constitutively functioning. The B-on phenotype is characterized by septal breakdown, continuous intercellular nuclear migration, and a morphological phenotype with few aerial hyphae and short irregular hyphal branches (Parag 1962; Koltin and Flexer 1969). The mutation eliciting this B-on phenotype is in a pheromone gene (Fowleret al. 1998). The encoded mutant pheromone is capable of activating the resident Bβ2 pheromone receptor to transduce a signal for B-regulated development within the unmated homokaryon and also promotes donation of migrant nuclei in a mating with the progenitor Bβ2 strain. In subsequent experiments, this B-on mutant was treated with X rays and a variety of secondary mutants were selected for normal homokaryotic morphology, indicating suppression of the constitutively active B-regulated pathway (Raper and Raper 1973). However, some of these secondary mutants retained the ability to donate migrant nuclei to the grand-progenitor Bβ2 strain and also showed additional mating defects that mapped to the Bβ2 locus. One secondary mutant contained a lesion first called f and later called def (Raper and Raper 1973; Raper 1988). The def mutant mates unilaterally: it can donate fertilizing nuclei to compatible mates, including the grandprogenitor Bβ2 strain, but is incapable of accepting fertilizing nuclei in any mating dependent upon Bβ compatibility. The def phenotype would be logically interpreted to reflect a continued presence of the mutant pheromone but a complete loss of Bβ2 receptor function. In the original analysis, however, this secondary mutation appeared to map to a position just outside the Bβ locus rather than to the receptor gene (Raper and Raper 1973).

In this study we show that the def secondary mutation was caused by disruption of the Bβ2 pheromone receptor gene, bbr2, through insertion of a previously uncharacterized transposon called scooter-1. We propose that the conflict this finding creates with previous mapping data can be resolved by considering that the disrupted receptor gene, bbr2-1, can occasionally be restored to function by excision of the transposon. A second copy of scooter spontaneously tagged thn1, a gene encoding a putative regulator of G protein signaling (RGS) protein. Disruption of thn1 has pleiotropic effects on vegetative growth and sexual development in S. commune. Scooter is the first active transposon identified in a basidiomycete.

MATERIALS AND METHODS

S. commune strains, growth conditions, transfection, and mating: The following S. commune strains were used for these studies and are part of a collection maintained at the University of Vermont. The primary B-on mutant has the Bβ designation Bβ2(1). Gene symbols shown following the designated Bβ mating-type specificity indicate mutations in the strain: 4-8, Bβ2; 4-40, Bβ1; 8-7, Bβ2 arg7; V11-20, Bβ2(1) def trp1; V15-34, Bβ2(1) pab1 bug; V112-3, Bβ7 trp1; V113-9, Bβ3 trp1; V119-19, Bβ9 trp1; V145-2, Bβ2 def trp1; V168-3, Bβ2 trp1 thn1; and V168-5, Bβ2 trp1 thn1. Plate cultures of S. commune were grown at 30° on complete yeast extract medium (CYM)-1.2% agar medium (Raper and Hoffman 1974), supplemented with 2.5 mm tryptophan as necessary. Liquid cultures of S. commune were grown as described in Fowler et al. (1999). Protoplast preparation and transfections were done according to Specht et al. (1988), using Novozyme 234 (InterSpex Products, Foster City, CA) as the wall digesting enzyme, with modifications described by Horton and Raper (1991). Mating of strain 8-7 thn1 with V113-9, sporulation, and isolation of germlings were according to standard methods (Raper and Hoffman 1974).

DNA isolation, digestion, and cloning: S. commune total DNA was isolated from cultures grown in liquid medium by a DNA extraction method using hexadecyltrimethylammonium bromide (CTAB; Saghai-Maroofet al. 1984) with the following change in the protocol. Mycelial pellets were vacuum filtered from the medium, then frozen in liquid nitrogen and powdered with a mortar and pestle, rather than lyophilized. Digestion of DNA with restriction endonucleases was according to the manufacturer’s recommendations for the particular enzymes (Life Technologies, Grand Island, NY). Cloned DNA was prepared from bacterial or phage cultures by standard methods (Maniatiset al. 1982) with the exception that plasmid DNA used for sequencing or transfection of S. commune was prepared using affinity chromatography kits (QIAGEN, Valencia, CA). DNA subcloning was by standard procedures (Maniatiset al. 1982) into phagemid pBluescript KS+ (Stratagene, LaJolla, CA). Plasmids were electroporated into Escherichia coli strain DH5α using the Gene Pulser system according to manufacturer’s instructions (Bio-Rad, Hercules, CA).

Library construction and screening: Two subgenomic libraries were constructed in bacteriophage vector λgt10 (Life Technologies). EcoRI restriction fragments from S. commune genomic DNA of the appropriate sizes, 6 kb for the library of def mutant strain V145-2 and 4 kb for the library of the thn1 mutant of 8-7, were isolated from a 0.8% agarose gel with the GeneClean II kit (Bio101, Vista, CA). Approximately 25 ng of insert DNA was included with 500 ng vector in the ligation reactions. Ligated DNA was packaged using the Gigapack II XL system (Stratagene) and the packaged phage were used to infect E. coli strain C600hfl. For each subgenomic library, ∼5000 plaques were screened by plaque lift assay (Maniatiset al. 1982) using 32P-labeled DNA. The V145-2 library was screened with the insert of plasmid pTF2076-9, which is 2.1 kb and contains a 3′-truncated, functional copy of the wild-type bbr2 gene. A PCR product containing all of scooter-1, as well as 152 bp of bbr2 located adjacent to scooter-1, was amplified from pTF4061-3 with primers 97723-3 (5′-GGCATGGAAC TCGGGCAC) and 971021-1 (5′-CAGCTAAACAAAGGCCAC) and used as a probe for the 8-7 thn1 library. Conditions for the hybridizations and washes during the screening of all libraries were as described for Southern analyses (below), except that final washes were in 0.1× saline sodium citrate (SSC) and 0.1% sodium dodecylsulfate (SDS). Positive plaques were reisolated two times to identify single plaque clones and the EcoRI inserts of the clones were subcloned into plasmid vectors.

A complete genomic library of wild-type S. commune (4-8) was constructed by ligating Sau3AI fragments into λDASH II predigested with BamHI (Stratagene). The library was amplified in E. coli strain XL1-Blue (Maniatiset al. 1982). A total of 15,000 plaques from the amplified library were screened with 32P-labeled insert DNA from pTF6099-1. Plasmid pTF-6099-1 contains an insert of 2.7 kb derived from DNA that flanks the scooter-2 insertion in thn1 of S. commune mutant 8-7 thn1. A hybridizing plaque containing a 15-kb insert was identified and the insert was subcloned into plasmid vectors.

A wild-type S. commune cDNA library (4-40, courtesy of M. Raudaskoski) was screened as described above for the complete genomic library, except that only 10,000 plaques were used. The phage vector for this library was λZAP II (Stratagene) and plasmid clones containing the S. commune insert were released from the phage genome according to the manufacturer’s instructions.

Southern analyses and probes: Digested DNAs were size-fractionated on 0.8% agarose gels and transferred to nylon membranes (Hybond-N; Amersham Pharmacia, Piscataway, NJ) by capillary blotting. DNA was crosslinked to the filters with UV light and then prehybridized a minimum of 10 min in standard hybridization buffer (Maniatiset al. 1982) at 65° prior to addition of the probe. Hybridizations were incubated at least 16 hr at 65° and the filters were subsequently washed twice in a solution of 2× SSC and 0.5% SDS for 30 min at 65°. Probes were stripped from filters with a boiling solution of 0.1% SDS and checked by autoradiography.

Probes for all hybridizations were made from plasmid insert DNA isolated with the GeneClean II kit and labeled with 32P by the random primer method (Feinberg and Vogelstein 1983). Southern hybridization data with PstI-EcoRI double-digested S. commune genomic DNA showed that bbr2 probes hybridize to a single band, bbr2, for strains having the Bβ2 mating-type specificity and that the same probes hybridize to either one band or show no hybridization for strains of other Bβ specificities (data not shown). Therefore, at most, one band displayed on the Southern hybridization shown in Figure 3 may be due to hybridization with bbr2-related sequences rather than scooter-1 when probed with DNA containing scooter-1 and 152 bp of bbr2.

DNA sequencing and sequence comparisons: DNA sequencing reactions were done at the Vermont Cancer Center DNA sequencing facility at the University of Vermont using a dideoxynucleotide method and fluorescent labeling system (ABI Prism kit; Perkin Elmer-Cetus, Norwalk, CT). Oligonucleotide primers for sequencing and PCR were purchased from Genosys (The Woodlands, TX). DNA sequences have been deposited as GenBank accessions as follows: scooter-1, AF267871; scooter-2, AF267872; and thn1, AF267870. DNA and protein sequence comparisons were made using BLAST and BLAST 2 Sequences programs located at www.ncbi.nlm.nih.gov (Altschulet al. 1997; Tatusova and Madden 1999) and the Pileup program (Henikoff and Henikoff 1992) from the University of Wisconsin Package (Genetics Computer Group).

Microscopy and photographs: S. commune cultures used for microscopy were grown on thin layers of CYM medium on glass slides. Photomicrographs were taken on a Nikon Eclipse 400 microscope at 40× magnification using pseudo-Nomarski optics (i.e., the phase ring was partially disengaged). Photographs of S. commune colonies were taken at the medical photography studio at the University of Vermont.

RESULTS

The Bβ2 receptor gene bbr2 complements the def mutant phenotype: Inability of the def secondary mutant strain to accept fertilizing nuclei in a Bβ-dependent mating resembles the phenotype predicted for loss of Bβ pheromone receptor function. Absence of Bβ2 receptor function would also explain a lack of self-activation of the B-regulated pathway by the B-on mutant pheromone in this secondary mutant (Fowleret al. 1998). We wondered whether addition of the wild-type Bβ2 receptor gene, bbr2, to a def strain might overcome the missing receptor function to restore the B-on mutant phenotype. A def mutant strain auxotrophic for tryptophan (V11-20) was cotransfected with pTF2076-9 containing the wild-type bbr2 and pRHV1 (Horton and Raper 1995) containing wild-type trp1. In two separate experiments, 5 of 11 and 10 of 21 prototrophic colonies were restored to the B-on phenotype, indicative of constitutive activity of the B-regulated pathway. A control transformation of the def mutant strain with trp1 alone resulted in 22 prototrophs, all with normal homokaryotic morphology. Restoration of the B-on mutant phenotype by bbr2 led us to investigate the Bβ2 receptor gene of the def mutant.

Figure 1.

—The bbr2 gene has a restriction fragment length polymorphism between wild-type, B-on mutant, and def secondary mutant strains. Genomic DNA was double digested with EcoRI and PstI and analyzed by Southern hybridization. The Bβ2 S. commune strains represented are wild type (strain 4-8, lane 1), B-on mutant (strain V15-34, lane 2), and def secondary mutant (strain V11-20, lane 3). The probe was a 2.1-kb EcoRI-PstI cloned restriction fragment (pTF2076-9) containing only wild-type bbr2 DNA. Approximately 1 μg of DNA was loaded in each lane.

The Bβ2 receptor gene of the def strains has been altered: A comparison was made between Bβ2 wild-type, primary B-on mutant and def secondary mutant strains to determine whether any differences existed in their bbr2 genes. DNA extracted from these S. commune strains was analyzed by Southern blot hybridization. Figure 1 shows the bbr2 probe hybridized strongly to a band in each strain and that the hybridizing band in the def mutant strain was ∼600 bp larger than the hybridizing bands in the wild-type and B-on mutant strains. The probe for bbr2 used in this hybridization was the cloned equivalent of the 2.1-kb hybridizing band from the wild-type strain and contained only bbr2. Therefore, the observed difference in size between the hybridizing restriction fragments of the two mutants indicated that a change in bbr2 occurred at the time of secondary mutagenesis.

The insert in the bbr2 gene of the def mutant has characteristics of a DNA-mediated transposon: We set out to identify the nature of the change in the bbr2 allele of the def mutant, now termed bbr2-1. Additional Southern hybridizations indicated that bbr2-1 resided within an EcoRI restriction fragment of ∼6 kb (data not shown). A subgenomic phage library was constructed from a def mutant strain using EcoRI restriction fragments of ∼6 kb. Plaques that hybridized to the bbr2 probe were shown to contain a 2.7-kb PstI-EcoRI restriction fragment within the insert (data not shown), as expected from the original hybridization results (Figure 1). A comparison of sequences between the 2.7-kb restriction fragment containing bbr2-1 and the wild-type bbr2 allele (Fowleret al. 1999) showed that bbr2-1 contained a single change, a 655-bp insertion interrupting the first exon of bbr2 between codons 62 and 63. The inserted sequence introduces stop codons in all reading frames of bbr2 such that a receptor produced from this allele would include only the N-terminal 10% of wild-type receptor unless alternative splicing of the primary transcript occurred.

Figure 2.

—Nucleotide sequence of S. commune transposon scooter-1. The 647-bp scooter-1 element is shown in uppercase letters and is numbered on the right. The 32-bp terminal inverted repeat sequences are shown in boldface and subterminal repeats are underlined. The 8-bp target site duplication of the bbr2 gene is shown in lowercase and corresponds to base pairs 234-241 of bbr2 (GenBank accession no. AF148501).

The inserted DNA contained hallmarks of a DNA-mediated, or class II, transposable element (Figure 2; Finnegan 1989). One characteristic of DNA-mediated transposition is creation of a short duplication of the interrupted DNA at the site of insertion. This “target site duplication” flanks each end of the inserted transposon. In bbr2-1, 8 bp of the bbr2 coding sequence (GGGGCAAC) located adjacent to the 5′ end of the 655-bp insertion are duplicated directly following the 3′ end of the novel sequence of the insertion. Short terminal inverted repeats and subterminal repeats are also typical of class II transposons. The ends of the novel insert in bbr2-1 (i.e., excluding the 8-bp target site duplication) are inverted 32-bp perfect repeats and several subterminal repeated sequences are also present (Figure 2). Although a transposon may maintain autonomy by encoding its own transposase, this insert does not contain a significant open reading frame or any similarity to sequences in public databases. We conclude that this segment of DNA transposed into bbr2 at the time of secondary mutagenesis and is a previously unknown nonautonomous transposon. We have designated this sequence scooter-1.

The transposon scooter-1 is part of a small group of related sequences: It is common for related versions of a transposon to exist in a single genome and for the different versions to share significant regions of DNA sequence identity (Strecket al. 1986; Fedoroff 1989). To determine whether sequences with strong similarity to scooter-1 exist in the S. commune genome, a Southern blot analysis was done. Total DNA from seven S. commune strains was digested with PstI and EcoRI, restriction enzymes that have no recognition sites within scooter-1. Southern hybridization of these seven strains with the scooter-1 probe showed from 8 to ∼20 discrete hybridizing restriction fragments (Figure 3). Analyses using additional strains and other restriction endonucleases with 6-bp recognition sites produced similar results. The minimum number of hybridizing restriction fragments in any strain was 3 (data not shown). Scooter-1 or scooter-related sequences may be ubiquitous in S. commune, inasmuch as we have analyzed 14 strains collected from five continents and have failed to identify any strain that is devoid of scooter-related sequences.

Cosegregation of a scooter-related sequence and the thin mutant phenotype: A spontaneously occurring mutant phenotype called thin, commonly seen in some S. commune strains, is caused by mutation of the thn1 gene (Raper and Miles 1958; Raperet al. 1958). The thin phenotype appears as a fast-growing sector during colonial growth. While the thin sector may look fluffier than the wild-type growth from which it arose (Figure 4A), it grows in a relatively submerged fashion on semisolid medium once subcultured away from wild type (Figure 4, B and C). A similar observation had been previously made by Schuren (1999). Thin hyphae are distinguished microscopically by their characteristic corkscrew morphology (Schwalb and Miles 1967) as contrasted to the relatively smooth morphology of wild-type hyphae (Figure 4, D and E). The high frequency of sectoring to the thin phenotype in certain strains led us to hypothesize that mutagenesis by a transposon might be responsible for the change. To test whether a scooter-related sequence is involved, a spontaneous sector of thin mutant hyphae was identified and subcultured from a colony with otherwise normal homokaryotic hyphal morphology. The original strain (8-7) and its thin mutant derivative were compared by Southern blot analysis. Hybridization of a scooter-1 probe to genomic DNA digested with EcoRI showed that the mutant had three distinct hybridizing restriction fragments not present in the wild type (Figure 5A).

Figure 3.

—The S. commune genome has a family of sequences related to scooter-1. Southern analysis of seven S. commune strains shows hybridization to a small family of sequences. The probe consisted of the entire scooter-1 element as well as 152 bp of bbr2. Under the hybridization conditions used here, probes containing only bbr2 either would hybridize to a single band or would not hybridize at all. Therefore, at most, one band of hybridization in each lane might be due to bbr2 and not scooter-1. Lane 7 contains the def secondary mutant strain from which scooter-1 was originally isolated; the arrow indicates the restriction fragment corresponding to the scooter-1 insertion into bbr2. Strains in lanes 3-7 are derived from crosses that have one common parental strain. Loading order (lane, strain): 1, 4-40; 2, 4-8; 3, V113-9; 4, V119-19; 5, V112-3; 6, V11-20; 7, V145-2. Approximately 1 μg of genomic DNA, double-digested with EcoRI and PstI, was loaded for each lane.

This mutant was mated with a compatible wild-type strain (V113-9) and the progeny of this mating were tested for segregation of the thin mutant phenotype with one or more of the restriction fragment polymorphisms. The progeny showed a 1:1 segregation ratio for wild type vs. thin (13:13) as would be expected for a trait controlled by a single gene. Southern hybridization analysis of nine offspring, four wild type and five thin, using a scooter-1 probe showed a perfect correlation between a hybridizing EcoRI restriction fragment of ∼4 kb and the thin phenotype (Figure 5B). None of the wild-type progeny exhibited this particular band. A comparable analysis of six more individuals (four thin and two wild type, data not shown) exhibited the same correlation.

Figure 4.

—Phenotype of a S. commune thn1 mutant. (A) Wild-type colony with a wedge-shaped spontaneous thn1 mutant sector on the left; (B) thn1 mutant colony grown in isolation; and (C) wild-type colony, progenitor of colony shown in B. Colonies shown in B and C were inoculated and grown simultaneously under identical conditions and are shown at the same scale. (D) thn1 mutant hyphae; and (E) wild-type hyphae. Strains shown in D and E were grown under identical conditions.

Isolation of a second scooter element and flanking DNA: Cosegregation of the thin phenotype and a restriction fragment identified by a scooter-1 probe indicated that a scooter-1-like sequence was either in the thn1 gene or closely linked to it. We proceeded to construct a subgenomic phage library of 4-kb EcoRI restriction fragments from one of the thin mutant offspring of the mating. Screening of the library with a scooter-1 probe yielded several positive clones. EcoRI inserts of two slightly different sizes were represented among the clones, corresponding to two bands of ∼4 kb seen in genomic Southern analysis (Figure 5B, last lane). A clone with the larger insert size, which corresponded to the cosegregating band, was selected for restriction analysis, subcloning, and sequencing. The clone contained a scooter transposon that is 91% identical in nucleotide sequence to scooter-1. This second copy of scooter, named scooter-2, created an 8-bp target site duplication (AGGCGATG) at the point of insertion and has 32-bp imperfect terminal inverted repeats that are identical at 31 positions. The 3′ terminal inverted repeat of scooter-2 differs from the 3′ end of scooter-1 by a single base pair, a T to A difference (Figure 2, base pair 625). Subterminal repeats in scooter-2 are also nearly identical to those identified in scooter-1, having a single difference in one of the repeat units (Figure 2, base pair 131, G to A).

A genomic Southern blot analysis was used to confirm that the DNA flanking the cloned scooter-2 also segregated with the thin mutant phenotype among the progeny of a cross of an unrelated wild-type strain (V113-9) and the thin mutant, suggesting this adjacent DNA was also linked to the thn1 gene. The probe was 2.7 kb of DNA flanking the scooter-2 insertion derived from the original 4-kb subclone of thin mutant DNA. EcoRI restriction fragments of 6.6 and 2.2 kb were identified in the wild-type parent and in the wild-type progeny from the cross by this probe made from “flanking DNA” (Figure 5C). The same probe hybridized to DNA isolated from thin mutant progeny only at the 4-kb EcoRI restriction fragment (Figure 5C) originally identified by the scooter-1 probe (Figure 5B). There are no EcoRI sites in the 653-bp scooter-2 sequence; we therefore predicted that the probe made from DNA flanking scooter-2 would hybridize to a 3.4-kb EcoRI band in the wild-type strain (8-7) from which the original thin sector arose. This expectation was confirmed by Southern blot analysis (Figure 5D).

Figure 5.

—Cosegregation analysis of scooter-2 and the thin mutant phenotype. (A) Genomic Southern analysis of a wild-type strain (8-7, lane 1) and its thin mutant derivative (lane 2) with a scooter-1 probe. Asterisks indicate polymorphisms. (B) Southern hybridization with a scooter-1 probe was used to compare nine progeny from a cross between the thin mutant (A, lane 2) and an unrelated wild-type strain (V113-9). The phenotype of each individual is indicated above the lane, and the arrow indicates a band that segregates with the thin mutant phenotype. (C) Autoradiograms of two filters probed with 2.7 kb of DNA that flanks the scooter-2 insertion in the thin mutant (pTF6099-1). The single lane on the left shows the hybridization pattern of the wild-type outcross parent (op). A filter with samples from the segregating progeny shows two distinct patterns of hybridization for wild-type vs. thin mutants (right, identically loaded as in B). The arrow indicates the same band as identified in B. (D) An autoradiogram of the filter used in A, reprobed with 2.7 kb of DNA that flanks scooter-2. As in A, the wild-type progenitor (lane 1) is compared to its thin mutant derivative (lane 2). All DNA samples were digested with EcoRI and loaded at ∼1-2 μg per lane.

Wild-type genomic and cDNA clones of thn1 are identified using DNA flanking the scooter-2 insert: Preliminary sequence analysis of the thin mutant clone suggested the scooter-2 element had interrupted the coding region of a gene. To characterize this putative gene, genomic and cDNA libraries constructed from wild-type S. commune strains were screened using the probe made from DNA flanking scooter-2. A phage clone isolated from the genomic library yielded an ∼15-kb insert that was subsequently subcloned into plasmid vectors. Comparison of sequences derived from the wild-type and mutant strains confirmed the target site duplication created by scooter-2.

Screening of the cDNA library yielded six positive clones. All six cDNA clones, including the one containing the largest insert of 2.5 kb, were partially or completely sequenced and were found to represent the same gene. The cDNA sequence was identical to the regions of the wild-type genomic clone that had been sequenced, except for the presence of one intron in the genomic clone. This comparison of sequences, in conjunction with a genomic Southern blot that showed only a single band of hybridization when the full-length cDNA was used as a probe (data not shown), confirmed that these clones represent a single gene, thn1. This gene was interrupted by scooter-2 between codons 104 and 105 in the thn1 mutant (Figure 6).

Thin mutants can be complemented by addition of the wild-type thn1 gene: To be assured that we had identified the gene responsible for the thin phenotype, we attempted to complement a thin mutant with a genomic copy of the putative wild-type thn1 gene. A plasmid containing this 3.1 kb of wild-type genomic DNA was cotransfected with a trp1 plasmid into protoplasts isolated from an auxotrophic thn1 mutant strain. The plasmid insert containing thn1 included a genomic region that corresponded with the longest cDNA, as well as ∼600 bp of DNA upstream of the transcribed region. In two separate transfection experiments, 5 of 16 and 6 of 37 prototrophic transformants were restored to wild-type hyphal and colonial morphology. A control transfection with only the trp1 plasmid yielded only thin mutant colonies. These experiments confirm that we have isolated thn1, a gene responsible for maintaining normal hyphal morphology and colonial growth.

Figure 6.

—Comparison of deduced amino acid sequences of RGS domain proteins Thn1 (S. commune), FlbA (A. nidulans), and Sst2p (S. cerevisiae). Identical amino acids at each position are shown in a black background, conservative substitutions are shown in a gray background, and gaps are indicated by dots. Regions of Thn1 that fit the RGS domain consensus are underlined. Numbering is according to SwissProt Accessions P38093 for FlbA and P11972 for Sst2p. Alignments were done using the Pileup program (Genetics Computer Group).

Deduced amino acid sequence indicates that Thn1 is an RGS protein related to FlbA and Sst2p: A search of protein databases using the amino acid sequence predicted from the longest thn1 cDNA revealed a strong similarity to a large number of protein sequences featuring the RGS domain. Proteins of the RGS domain family function as signal regulators by acting as GTPase-activating proteins (GAPs) for Gα subunits of heterotrimeric G proteins (Berman et al. 1996a,b) or by directly interfering with activation of effector molecules by the Gα subunits (Hepleret al. 1997). Two RGS proteins from the fungi are well known, FlbA of Aspergillus nidulans (Lee and Adams 1994a) and Sst2p of Saccharomyces cerevisiae (Dietzel and Kurjan 1987). Their sequences are compared to the predicted amino acid sequence of the thn1 protein, Thn1, in Figure 6. Pairwise comparison of the deduced amino acid sequences of Thn1 and FlbA revealed 36% identity and 54% similarity over the length of Thn1, while a similar comparison of Thn1 with Sst2p showed 22% identity and 39% similarity.

DISCUSSION

Scooter-1 and scooter-2 of S. commune are closely related copies of the first transposon shown to be active in a basidiomycete. Scooter appears to belong to the hAT (hobo/Ac/Tam3) superfamily of class II transposons, having an 8-bp target site duplication, short terminal inverted repeats, and subterminal repetitive sequences like other hAT superfamily members (Calviet al. 1991). Two other class II transposon-like sequences have been identified in basidiomycetes: Abr1 in A. bisporus, which has been tentatively placed in the hAT superfamily based on a 7-bp target site duplication and subterminal repeats (Sonnenberget al. 1999), and Pce1 of Phanerochaete chrysosporium (Gaskellet al. 1995). There is no DNA sequence identity between the currently identified members of these three transposon groups.

Scooter-1 and scooter-2 appear to be nonautonomous elements that were mobilized by a transposase encoded elsewhere in the S. commune genome. Each of these scooter elements has terminal inverted repeats and subterminal repeats that may constitute cis-acting sequences for transposase binding as is the case in the Ac/Ds transposons (Kunze and Starlinger 1989). Each also has a relatively short length and no long open reading frame capable of encoding a transposase. Identification of a transposase gene for the scooter transposon would clarify whether scooter is truly part of the hAT superfamily. One S. commune strain that must produce transposase, indicated by the recent movement of scooter-2 into thn1, was shown to have several scooter-related sequences in its genome (Figure 5A). One of the restriction fragments to which scooter-1 hybridizes may contain an autonomous element, since DNA sequence identities are commonly found between autonomous and nonautonomous members of transposon groups (Döringet al. 1984; Pohlmanet al. 1984; Strecket al. 1986).

Our current discovery that the def secondary mutation in Bβ2 was caused by insertion of scooter-1 into the Bβ2 receptor gene provides evidence of recent mobility of the scooter element. Previous mapping data indicated recombination between the B-on mutation in the Bβ2 locus and the def mutation, now called bbr2-1 (Raper and Raper 1973). Reinterpretation of these data following discovery of the transposon strongly supports the idea that scooter-1 occasionally excised from its position in bbr2-1. Two outcrosses between the def mutant and wild-type strains appeared to yield 13 recombinants among the progeny analyzed, suggesting that the def mutation was located 0.8 cM outside the Bβ locus. The number of recombinants in the two expected classes were extremely skewed, however, with 12 of the apparent recombinants fitting into one class and only 1 into the second class. Members of the majority class had the phenotype of the B-on mutant. In retrospect, the preponderance of individuals with the B-on phenotype can best be explained as the result of excision of scooter-1 from bbr2-1 in a manner that restored Bβ2 receptor function while in the presence of the mutant pheromone that activates this receptor. The one member of the minority recombinant class, and perhaps a small number of the majority class as well, may represent true recombinations between the pheromone gene and the receptor gene, which are now known to lie ∼1.2 kb apart within the Bβ2 locus. Direct molecular evidence to support this interpretation is not possible to obtain since these strains no longer exist.

Scooter-2 acted as a molecular tag for identification of thn1. Mutations in thn1 give rise to a frequently encountered morphological mutant of S. commune. The predicted gene product of thn1 is a RGS protein, one of a few RGS proteins currently identified in the fungi. More than 40 genes encoding putative RGS proteins have been identified in animals as well. The common feature of all RGS proteins is a consensus domain of ∼120 amino acids (De Vrieset al. 1995) that is thought to interact directly with the Gα subunit of heterotrimeric G protein. This interaction modulates the rate of conversion of GTP-bound Gα protein to GDP-bound Gα protein to regulate signals passed from an activated seven-transmembrane-domain receptor to downstream effector molecules via the heterotrimeric G protein. The founding member of the RGS protein family, Sst2p of S. cerevisiae, contributes to the process of desensitization and recovery from pheromone stimulation during mating (Chan and Otte 1982; Dietzel and Kurjan 1987). The other well-characterized fungal RGS protein is FlbA of A. nidulans (Lee and Adams 1994a). FlbA belongs to a signaling pathway that leads to inhibition of vegetative hyphal proliferation and to promotion of asexual sporulation. It is thought to act as a GTPase-activating protein for the Gα protein, FadA (Yuet al. 1996; Adamset al. 1998). A knock-out mutation of flbA results in A. nidulans colonies with increased hyphal proliferation, reduced sporulation, and eventual autolysis of the older hyphae (Lee and Adams 1994a). Like FlbA, disruption of thn1 results in pronounced changes in vegetative hyphal morphology. The effects on asexual sporulation in A. nidulans have no parallel in S. commune because S. commune does not produce asexual spores. One characteristic of thn1 mutants, however, is the inability to produce fruiting bodies from a compatible mating homoallelic for the thn1 mutation (Schwalb and Miles 1967).

The three deduced fungal RGS protein sequences contain conserved regions outside of the RGS domain, suggesting that these proteins may contain regions important for other common functions. There is strong sequence identity (36%) between Thn1 and FlbA throughout the entire Thn1 sequence. Thn1, like Sst2p, does not have a region similar to the N-terminal 204-amino-acids region of FlbA, which has no defined function as yet. Thn1 and Sst2p share identical start positions relative to FlbA, suggesting that the N-terminal extension of FlbA is an acquisition more recent than the divergence of ascomycetes and basidiomycetes. In all other respects, the primary sequences of Thn1 and FlbA are much more similar to each other than either of these proteins is to Sst2p. Of great interest is a region of Thn1 between amino acid residues 51 and 110 that shares a very high level of similarity with FlbA, but not with Sst2p. Particularly striking is a string of seven consecutive threonine residues (amino acids 87-93).

The strong sequence similarity of Thn1 to other RGS proteins, such as FlbA and Sst2p, suggests a similarity in biochemical functions. One consequence of the loss of Thn1 function is a reduction or loss of mRNAs encoding hydrophobins (Wesselset al. 1991). These hydrophobins include one normally expressed in both homokaryons and dikaryons (Sc3) and two that are dikaryon specific (Sc1 and Sc4). We hypothesize that Thn1 is regulating a heterotrimeric G protein-signaling pathway that, in turn, regulates hydrophobin expression. Our observation that the thn1 mutant has an altered macroscopic phenotype when isolated from its sector within the wild-type colony corroborates earlier observations of Schuren (1999). He noted a similar phenotypic switch of thn1 mutants from the production of abundant aerial hyphae in the presence of wild-type hyphae to a relatively submerged habit of growth in isolation. He also produced some evidence that a diffusible molecule <8 kD may be responsible for this effect. This finding has similarity to the response of A. nidulans fluG mutants to a small molecule produced by A. nidulans wild-type hyphae, resulting in a shift from fluffy vegetative growth to conidiation (Lee and Adams 1994b). The small molecule proposed by Schuren is obviously not Thn1, but it could be part of a pathway in which Thn1 operates. Since some, but not all, of the pleiotropic effects of thn1 mutations were overcome by incubation with wild-type hyphae (Schuren 1999), Thn1 may be part of a complex network of signaling, as is true for FlbA in A. nidulans (Hickset al. 1997; Adamset al. 1998; Yuet al. 1999).

Identification of thn1 provides a starting point for elucidating the components of what is likely to be a G protein-mediated signaling system that regulates pathways for vegetative and fruit body development in S. commune. It remains to be determined whether Thn1 exerts any effect on the G protein-regulated pathway of pheromone response that controls sexual development, a pathway in which there is a longstanding interest. As more is learned about scooter, we hope this transposon will become useful as a tool for gene tagging in the quest to understand these signaling pathways.

Acknowledgments

The authors are grateful for the following contributions to this work: Dr. Carlene A. Raper for providing laboratory space and supplies for this project, as well as fruitful discussions and excellent advice; Professor M. Raudaskoski for providing a S. commune cDNA library; Dr. L. Vaillancourt for constructing the S. commune genomic library; A. Merla for the microscopic images in Figure 1; J. Dion and Dr. J. Bond for alignments shown in Figure 6; M. Shepard for technical assistance; Dr. F. Schuren for providing a manuscript prior to publication; Drs. E. Froeliger, J. S. Horton, J. Kurjan, and M. Tierney for helpful comments during preparation of the manuscript. This work was supported by grant MCB9513513 from the National Science Foundation to Carlene A. Raper.

Footnotes

  • Communicating editor: M. E. Zolan

  • Received February 4, 2000.
  • Accepted September 1, 2000.

LITERATURE CITED

View Abstract