Suppression of fruit-body formation by constitutively active G-protein {alpha}-subunits ScGP-A and ScGP-C in the homobasidiomycete Schizophyllum commune

Heterotrimeric G-protein α-subunits are key regulators of growth and development in eukaryotic cells. For instance, there are two distinct G-protein α-subunit genes in Saccharomyces cerevisiae. One gene, gpa1, has been shown to be involved in the negative regulation of the MAP kinase pathway upon pheromone stimulation (Kurjan, 1993 ; Miyajima et al., 1987 ), whereas the other gene, gpa2, is thought to regulate growth and pseudohyphal development (Isshiki et al., 1992 ; Kubler et al., 1997 ). In Magnaporthe grisea, which significantly limits rice production by causing rice blast disease, three genes encoding α-subunits of the heterotrimeric G-protein have been characterized by gene disruption experiments (Liu & Dean, 1997 ). Whereas the deletion of magA or magC did not affect vegetative growth nor appressorium formation, disruption of magB resulted in a significant reduction in vegetative growth, conidiation and appressorium formation. Four G-protein α-subunit genes (gpa1, gpa2, gpa3 and gpa4) have been found in a heterobasidiomycete, Ustilago maydis. Of these genes, gpa3 is necessary for the transmission of pheromone stimulation and for the development of pathogenesis (Regenfelder et al., 1997 ).

In the homobasidiomycete Schizophyllum commune, genes encoding heterotrimeric G-protein α-subunits (SCGPα1, SCGPα2, ScGP-A, ScGP-B and ScGP-C) have been reported. SCGPα1 and SCGPα2 [GenBank accession nos AF157495 (A. Pardo, M. Gorfer and M. Raudaskoski) and AF306530 (M. Raudaskoski, T. J. Fowler and M. Matrick), respectively] have been predicted to be involved in signal transduction in the mating interaction. The ScGP-A, ScGP-B and ScGP-C genes (GenBank accession nos AB066503, AB051903 and AB051904, respectively) were found in the S. commune monokaryon strain T11. ScGP-A resembles SCGPα1 (92·6% similarity) and ScGP-C shows high similarity with SCGPα2 (98·6%) at the amino-acid level. ScGP-B shows high similarity with the Coprinus congregatus CGPα1 gene (87·8%), which has been reported as a candidate for regulating the blue-light-induced signal transduction photomorphogenesis system found in this species (Kozak & Ross, 1991 ; Kozak et al., 1995 ).

What has been demonstrated in the above reports was based mainly on amino-acid-sequence comparisons, mRNA-expression patterns or biochemical studies. To demonstrate the function of G-protein α-subunits in S. commune more directly, we transformed the S. commune monokaryon strain T11 with constitutively active mutated G-proteins. The relationship between G-protein α-subunits and the thn-1 gene, which has been reported as a putative regulator of the G-protein signalling protein (RGS) (Fowler & Mitton, 2000 ), is also discussed.

Strains and media.
The strains used in this report are listed in Table 1. S. commune T11 (Aα4ß1 Bα1ß6 Trp1 Ura1) was kindly provided by Dr R. C. Ullrich (Vermont University, USA) via Dr Sen (Shinsyu University, Japan). The S. commune monokaryon strain, F1, was separated from a single spore of strain IFO 30496 (Institute for Fermenation, Osaka, Japan) and used as a tester strain for mating. Though the mating type of the F1 strain has not been identified, the T11 strain could be mated normally with the F1 strain to produce the T11/F1 dikaryon strain. These strains were grown on a complex medium containing yeast extract supplemented with 4 mM tryptophan (CYMT) for strains T11 and T11/F1 or without tryptophan (CYM) for strain F1. Monokaryon strains F2-1, F2-2, F2-3, F2-4 and F2-5 were separated from individual spores of the T11/F1 strain. F2-1, F2-2 and F2-3 could be mated with F2-4 and F2-5 to form a dikaryon. F2-1, F2-2 and F2-3 exhibited a barrage phenotype when they were crossed with T11 and exhibited a flat phenotype when they were crossed with strain F1. Strains F2-4 and F2-5 exhibited a flat phenotype when crossed with T11 strains and a barrage phenotype when crossed with F1 strains.

Table 1. S. commune strains used in this study

S. commune dikaryon clones derived from F1 and transformed T11 strains were cultured on ASN medium containing (l^-1) 20 g sucrose, 1·5 g asparagine, 0·46 g KH₂PO₄, 1·0 g K₂HPO₄, 0·5 g MgSO₄, 0·15 g CaCl₂.2H₂O, 0·3 mg ZnSO₄.H₂O, 0·15 mg FeCl₂, 0·1 mg CuSO₄.5H₂O, 0·1 mg MnSO₄.5H₂O, 5 mg thiamin/HCl, 1 mg nicotinamide and 15 g agarose (pH 5·6) for the fruit-body-formation test. Thiabendazole (final concentration, 20 mg l^-1) was added to all culture media to prevent contamination.

DNA and RNA manipulation.
Generally, PCR reactions were performed by using Platinum Taq DNA polymerase High Fidelity (Barnes, 1994 ) with a Taq PCRxEnhancer System (Gibco-BRL). HotStart Taq DNA Polymerase (Qiagen) was used to perform the degenerate PCR procedure. The primers used in this work are listed in Table 2. PCR products were inserted into a T7Blue T vector (Novogen) by using single 3'-A overhangs (Marchuk et al., 1991 ). The sequencing procedure was performed by using a BigDye Terminator Kit (Perkin Elmer) and an ABI Prism automated DNA sequencer (model 377; Applied Biosystems). 5'-RACE PCR was performed by using a SmaRT-PCR cDNA Synthesis Kit (Clontec) and M-MuLV Reverse Transcriptase (Gibco-BRL). The first stranded cDNA used as a template for RT-PCR was prepared by using a Takara RNA LA PCR Kit (version 1.1; Takara). Genomic DNA was prepared by using a DNeasy Plant Mini and Maxi Kit (Qiagen). Total RNA was prepared by using an RNeasy Plant Mini Kit (Qiagen). mRNA was prepared by using an MPG Guanidine Direct mRNA Purification Kit (CPG). Plasmids were prepared by using a QIAPrep Spin Miniprep Kit (Qiagen). Plasmids for S. commune transformation experiments were prepared by using a Quantum Prep Plasmid Maxi or Midi Kit (Bio-Rad). PCR products and DNA fragments were excised from agarose gels and purified by using a QiaQuick Gel Extraction Kit (Qiagen). Ligation reactions and Escherichia coli transformations (Hanahan, 1983 ) were performed by using a DNA Ligation Kit (version 2; Takara) and DH5α competent cells (Toyobo).

Table 2. Oligonucleotides used as primers in this study

Cloning procedure.
The ScGP-A and ScGP-C genes were cloned as follows. Genomic DNA from strain T11 was amplified using primers GPαF-1, GPαF-2 and GPαR, to obtain the gene corresponding to SCGPα1. The PCR products were cloned and sequenced. As a result of this process, two kinds of gene fragment, corresponding to ScGP-A and ScGP-C, were obtained. 3'-Flanking regions were cloned by 3'-RACE using nested primers GPA-F1 and GPA-F2 for ScGP-A and GPC-F1 and GPC-F2 for ScGP-C. 5'-Flanking regions were cloned by 5'-RACE using nested primers GPA-R2, GPA-R3 and GPA-R4 for ScGP-A and GPC-R2, GPC-R3 and GPC-R4 for ScGP-C. Genomic DNA and cDNA encoding full-length proteins were amplified using primer sets GPA-F3 and GPA-R1, and GPC-F3 and GPC-R1 to confirm the position of introns.

ScGP-B was cloned as follows. Degenerate PCR was performed to obtain the G-protein α-subunits family from S. commune (McPherson et al., 1994 ). mRNA was purified from strain T11 that had been cultured in CYMT liquid medium. cDNA was synthesized using random nanomer as the primer. Degenerate primers GPdegF1, GPdegF2, GPdegR1 and GPdegR2 were designed to amplify the amino-acid sequences MFDVGGQ (GPdegF1 and GPdegF2) and KWIHCFE (GPdegR). The following protocol was used for amplification: denaturation for 15 min at 95 °C, 35 cycles of 30 s at 94 °C, 1 min at 45 °C and 1 min at 72 °C, followed by a final elongation for 7 min at 72 °C. The PCR products were cloned and sequenced. The 3'-flanking region was cloned by 3'-RACE using nested primers GPB-F1 and GPB-F2. TAIL-PCR was performed as described by Liu & Whittier (1995) to obtain the 5'-flanking region; degenerate primers TAIL1, TAIL2, TAIL3, TAIL4, TAIL5 and TAIL6 were used. Nested primers GPB-R2, GPB-R3 and GPB-R4 were used as gene-specific primers. The transcription start point was determined by 5'-RACE using primer sets GPB-R5, GPB-R6 and GPB-R7. Genomic DNA and cDNA encoding full-length proteins were amplified using primers GPB-F3 and GPB-R1 to confirm the position of introns.

The S. commune hydrophobin SC3 gene and glyceraldehyde-3-phosphate dehydrogenase (GPD) gene were previously cloned from the IFO 30496 strain (cloning procedure not shown). SC3 and GPD have been reported previously (Schuren & Wessels, 1990 ; Harmsen et al., 1992 ). We referred to the sequence data to obtain the same genes from the IFO 30496 strain. The promoter and terminator regions of the SC3 and GPD genes were obtained by using the TAIL-PCR procedure.

The SC3 promoter (1·9 kbp), SC3 terminator (0·4 kbp), GPD promoter (0·6 kbp) and GPD terminator regions (0·6 kbp) were amplified from the genomic DNA of IFO 30496, to insert restriction enzyme sites, using primers SC3-Pro-F and SC3-Pro-R, SC3-Ter-F and SC3-Ter-R, GPD-Pro-F and GPD-Pro-R, and GPD-Ter-F and GPD-Ter-R, respectively. NcoI sites were generated at the first Met codon. Each PCR product was cloned into the pT7Blue T vector and sequenced. The vector was digested with restriction enzymes, and the promoter (terminator) fragments were excised by using KpnI and NcoI (XbaI and BamHI). Though the primer SC3-Pro-F did not contain a KpnI site, there was a native KpnI site in the SC3 promoter region. Each fragment was purified and inserted into the pGVB2luc plasmid (Toyo B-Net), to construct the SC3 and GPD gene-expression vectors. Following this procedure, the luciferase gene (luc) was exchanged for the mutated ScGP-A, ScGP-B and ScGP-C genes (see below).

Site-directed mutagenesis of ScGP-A, ScGP-B and ScGP-C and expression vector construction
. We used the plasmids containing ScGP-A, ScGP-B and ScGP-C genomic DNA fragments (containing introns) as templates for the site-directed PCR mutagenesis procedure (Clackson et al., 1994 ). The mutant genes ScGP-A (clone Q207R), ScGP-B (clone Q204R) and ScGP-C (clone Q204R), in which the glutamine at positions 207, 204 and 204, respectively, was exchanged for arginine, were generated by PCR using primers GPA-Nco and GPA-QRR, GPA-Xba and GPA-QRF, GPB-Nco and GPB-QRR, GPB-Xba and GPB-QRF, GPC-Nco and GPC-QRR, and GPC-Xba and GPC-QRF. The following protocol was used for amplification: denaturation at 94 °C for 2 min, 15 cycles at 94 °C for 30 s, 55 °C for 2 min and 68 °C for 2 min, followed by a final elongation at 68 °C for 7 min. The PCR products were excised from agarose gels and then purified. In the second round of PCR, the PCR products were mixed with each other and used as templates. GPA-Nco and GPA-Xba, GPB-Nco and GPB-Xba, and GPC-Nco and GPC-Xba were used as primer sets to amplify the fragments harbouring the point mutations. The PCR conditions used were the same as above. The PCR products were then cloned and sequenced. As a result of this process, the glutamine codons of ScGP-A (CAG), ScGP-B (CAA) and ScGP-C (CAG) were replaced by arginine codons (CGG, CGA and CGG, respectively). The mutant genes generated, namely ScGP-A (clone Q207R), ScGP-B (clone Q204R) and ScGP-C (clone Q204R), were excised from pT7Blue T by using NcoI and XbaI. The purified gene fragments were ligated into the SC3 and GPD expression vectors. The ligation products were then cloned, sequenced and used for S. commune transformation experiments (Fig. 1).

(31K): Fig. 1. Gene-expression vectors used in this study. The construction procedures for all of the plasmids used in this study are described in Methods. SC3Pro, SC3Ter, GPDPro and GPDTer indicate the promoter (1·5 kbp) and terminator (0·4 kbp) regions of the S. commune hydrophobin SC3 gene and the promoter (0·6 kbp) and terminator (0·6 kbp) regions of the S. commune GPD gene, respectively. luc indicates the luciferase gene. The solid lines indicate the pGVB2luc vector sequence. ScGP-A (clone Q207R), ScGP-B (clone Q204R) and ScGP-C (clone Q204R) were mutated G-proteins containing point mutations. Since these fragments were amplified from genomic DNA by PCR, they contained intact introns (not shown).

Transformation procedure.
Transformation experiments were conducted with pGEM7⁺/TRP1, which contains the Schizophyllum TRP1 sequence (Munoz-Rivas et al., 1986a ). pGEM7⁺/TRP1 was kindly provided by Dr R. C. Ullrich, via Dr Sen. The transformation procedure was performed following the protocol of Munoz-Rivas et al. (1986b ) with some modifications. Strain T11 was inoculated into liquid CYMT. The mycelium was cultivated at 30 °C with reciprocal shaking (80 r.p.m.) in a conical flask (500 ml) containing 100 ml of liquid medium. To provide fresh mycelia for preparing protoplasts, the mycelium was homogenized with a homogenizer (Cell Master, Iuchi) at 10000 r.p.m. for 5 min and then subcultured for a day in 10 conical flasks with identical shaking. The mycelia were harvested by filtration and washed several times with sterile water and 0·5 M MgOsm (0·5 M MgSO₄, 10 mM MES, pH 6·3). About 20 ml of the damp mycelium was incubated in 50 ml of 1·0 M MgOsm (1 M MgSO₄, 20 mM MES, pH 6·3) containing 20 mg cellulaseonozuka RS ml^-1 (Yakult), 5 mg lysing enzymes L2773 ml^-1 (Sigma) and 0·25 mg chitinaseGODO ml^-1 (Seikagaku Kogyo) at 30 °C for 6 h with continuous shaking (60 r.p.m.). The protoplasts were separated from the crude suspension by density gradient and suspended in SorbOsm (0·5 M sorbitol, 10 mM MES, pH 6·3). The protoplasts (5x10⁷ cells, 400 µl) were then mixed with 240 µl of DNA solution (10 µg pGEM7⁺/TRP1, 30 µg gene-expression vectors, 0·5 M SorbOsm, 50 mM CaCl₂) and incubated on ice for 30 min. Polyethylene glycol (PEG) (50%, w/v; PEG4000 in 0·5 M SorbOsm, 640 µl) was then added to the solution, which was incubated on ice for 5 min. Regeneration medium (10 ml; CYMT, 0·5 M sucrose) was then added. After incubation for a further 20 h, the regenerating protoplasts were separated from the tryptophan-supplemented medium by pelleting them by centrifugation (1000 g, 10 min), aspirating the supernatant and resuspending the pellets in 0·5 M sucrose. The protoplasts were washed four times with 0·5 M sucrose in this manner. The final pellets were suspended in 0·5 M sucrose, mixed with 1% melting agar (SeaPlaque Agarose; BMA) in regeneration medium without tryptophan at 37 °C and poured into 16 Petri dishes. Two or three days later, each of the 48 independent regenerating clones was recovered and used for further experiments.

Selection of co-transformed clones.
All regenerated clones were cultured on a cellophane membrane overlying CYM for 3 days in Petri dishes. Following this incubation, pieces of the mycelium were detached and placed in 96-well plates. The mycelium was frozen at -80 °C until screening. After thawing at room temperature, Triton-X (0·05%) in TE was added to each well; the contents of the wells were again frozen and thawed. The supernatants were added to the PCR solution for direct screening. The genomic DNA of clones, which contained both the full region of the SC3 or GPD promoter and the 5' regions of mutated G-protein or luciferase genes, was amplified using primer sets PGVB2-F and GPA-R4 for ScGP-A, PGVB2-F and GPB-R7 for ScGP-B, PGVB2-F and GPC-R4 for ScGP-C, and PGVB2-F and Luci-R for pSC3luc. Primer PGVB2-F (PGVB2-R) was located adjacent to the KpnI (BamHI) site of the pGVB2luc plasmid. We amplified 2040% of the genomic DNA derived from each clone and assessed it to see if the clones were positive transformants. The genomic DNA of positive transformants was amplified using primer sets PGVB2-R and GPA-F2 for ScGP-A, PGVB2-R and GPB-F4 for ScGP-B, PGVB2-R and GPC-F1 for ScGP-C, and PGVB2-R and Luci-F for pSC3luc, to screen for the co-transformed clones that contained both the full region of the SC3 or GPD terminator and the 3' regions of mutated G-protein or luciferase genes. For each construct, between nine and 17 positive transformants were obtained which were used for further analysis.

Evaluation of the mutated-gene-expression level.
All co-transformed clones (see above) were grown on cellophane membranes for 3 days to evaluate the mRNA-expression level of mutated G-protein α-subunits. Both the monokaryon and dikaryon clones were cultured for the preparation of RNA, since the mRNA-expression level of mutated genes is affected by the cell condition (i.e. whether it is monokaryon or dikaryon). cDNA was synthesized by using oligo-dT as the primer. The cDNA sequences of endogenous ScGP-A, ScGP-B and ScGP-C and exogenous mutated genes were amplified simultaneously by RT-PCR using primer sets GPA-F2 and GPA-R5 for ScGP-A, GPB-F4 and GPB-R8 for ScGP-B, and GPC-F4 and GPC-R5 for ScGP-C. The 3' regions of primers GPA-R5, GPB-R8 and GPC-F4 were constructed to overhang introns for prohibiting the contaminated genomic DNA from amplification. The RT-PCR products were excised from agarose gels, purified and sequenced directly. On the resulting sequence electropherograms, two peaks, one derived from the endogenous gene (coding glutamine) and one derived from the exogenous mutated gene (coding arginine), were observed. The heights of the two peaks were measured and the ratios of endogenous gene (abbreviated as Q) to mutated gene (abbreviated as R) were calculated by using the formula [height of R/height of (R+Q)]. The quotient of this calculation was termed the mutated-gene-expression level.

Fruit-body-formation test in the dikaryon.
To confirm the influence of the mutated genes on fruiting, all clones were mated with the F1 tester strain on CYM agar. After 7 days incubation, the presence of clamp-connections was observed microscopically. The resultant dikaryons were cultured on ASN agar plates for 7 days in continuous darkness at 25 °C, followed by 4 days cultivation under light (1000 lx) or cultivation in continuous darkness for an additional 4 days at 25 °C. The degree of fruit-body formation was monitored and sorted into five grades: no primordia observed (0); a few primordia observed (0·25); many primordia observed (0·5); a few fruit-bodies observed (0·75); many fruit-bodies observed (1). The tests for fruit-body formation were performed in duplicate, and the mean grades of two tests were used as the Fruit-body-formation index.

Northern analysis of the ScGP-A, ScGP-B, ScGP-C and SC3 genes.
To evaluate the expression of ScGP-A, ScGP-B and ScGP-C and the hydrophobin SC3 gene, Northern analysis was carried out. The monokaryon clones and dikaryon strains T11/F1-4 or T11/F1-5, which exhibited a flat phenotype, were cultured on cellophane membranes for 3 days prior to the isolation of total RNA. After its isolation, 3 µg total RNA was loaded into each well of the gel and an RNA ladder (Promega) was used as a size marker. The hybridization probes were prepared as follows. The cDNA sequences of endogenous ScGP-A, ScGP-B and ScGP-C were amplified by RT-PCR using primer sets GPA-F3 and GPA-R6, GPB-F5 and GPB-R1, and GPC-F5 and GPC-R1, respectively. The SC3 cDNA was amplified by RT-PCR using primer set SC3-F and SC3-R, which covered the SC3 ORF. PCR products were cloned using the T7Blue T vector. The cDNA fragments were excised by digestion of the DNA with SphI for ScGP-A, BamHI for ScGP-B, HindIII and EcoRV for ScGP-C, and EcoRI and HindIII for SC3; these fragments were used for hybridization probes. Labelling and detection of the probes was performed using Gene Images random prime labelling and a CDP-star detection module (Pharmacia). Pre-hybridization and hybridization were performed at 62 °C in a hybridization buffer containing 5x SSC, 0·1% (w/v) SDS, 5% (w/v) dextran sulfate and a 20-fold dilution of liquid block (Pharmacia). Northern blots were washed at 62 °C in a wash buffer containing 1x SSC and 0·1% (w/v) SDS for 15 min. A second wash was carried out at the same temperature as the first, but this time a pre-heated solution containing 0·1x SSC and 0·1% (w/v) SDS was used for 15 min. Chemiluminescence was detected using an ECL-Minicamera (Pharmacia) and Polaroid film type T667 (Kodak).

Cloning of the ScGP-A, ScGP-B and ScGP-C genes
As a first step toward investigating the function of G-protein α-subunits in S. commune, we amplified genomic DNA from the T11 strain to obtain the gene corresponding to SCGPα1, which had already been submitted to GenBank by Pardo and colleagues (GenBank accession no. AF157495). Two genes, called ScGP-A and ScGP-C, were obtained. During this work, the sequence of SCGPα2 was submitted to GenBank by Raudaskoski and colleagues (GenBank accession no. AF306530). ScGP-C has high similarity with SCGPα2 (98·6%) at the amino-acid level, and thus may be a homologue of SCGPα2 (Fig. 2b). Similarly, the high similarity between ScGP-A and SCGPα1 (92·6%) indicated that these genes are homologues (Fig. 2a).

(85K): Fig. 2. Alignment of (a) SCGPα1 and ScGP-A, (b) SCGPα2 and ScGP-C and (c) C. congregatus CGPα1 and ScGP-B. The sequence alignments were performed by using CLUSTAL W. Asterisks indicate positions which have a single, fully conserved residue. Double dots indicate that one of the following strong groups is fully conserved (STA, NEQK, NHQK, NDEQ, QHRK, MILV, MILF, HY, FYW). Single dots indicate that one of the following weaker groups is fully conserved (CSA, ATV, SAG, STNK, STPA, SGND, SNDEQK, NDEQHK, NEQHRK, FVLIM, HFY).

Degenerate PCR was performed to find other G-protein α-subunit families in S. commune. Only one kind of PCR fragment was amplified, though the reason for this was unclear. This PCR fragment corresponded to ScGP-B. The sequence of ScGP-B exhibited significant similarity with C. congregatus CGPα1 (87·8%) at the amino-acid level (Fig. 2c). In addition, both CGPα1 and ScGP-B seem to possess both a cholera-toxin-sensitive site (R178 of CGPα1 and ScGP-B) and a pertussis-toxin-sensitive site (C350 of CGPα1, C351 of ScGP-B), unlike many other G-proteins. These features suggest that ScGP-B in S. commune is a homologue of CGPα1 in C. congregatus.

Transformation of constitutively active mutated ScGP-A, ScGP-B and ScGP-C genes into S. commune
To investigate the function of the ScGP-A, ScGP-B and ScGP-C genes, we constructed dominant-positive mutated genes whose products were locked in the active GTP-bound state. To produce these mutants, a glutamine was substituted for an arginine (clone Q207R, ScGP-A; clone Q204R, ScGP-B; clone Q204R, ScGP-C) by using site-directed mutagenesis (see Methods). The glutamine residue is thought to be necessary for the hydrolysis of GTP (Coleman et al., 1994 ). Mutations analogous to the ones we have used have been shown to lower the GTPase activity in many G-protein α-subunit proteins and, thus, prevent turnover to the inactive GDP-bound state (Landis et al., 1989 ; Murakami et al., 1999 ). The TRP1-deficient T11 monokaryon strain was co-transformed with the mutated gene-expression vectors (Fig. 1) and the marker gene pGEM7⁺/TRP1. Many clones (1017; see Methods) that had been co-transformed with the mutated genes were obtained and used for various phenotypic analyses.

Evaluation of the level of gene expression of the mutated exogenous genes
Since we anticipated that the levels of gene expression of the mutated exogenous genes would vary among the co-transformed clones for various reasons, for each clone we calculated the ratios of expression of endogenous genes to mutated exogenous genes and defined the quotient of this calculation as the mutated-gene-expression level (see Methods). Since this index was not the absolute but the relative level of expression, some clones with a high index value (over 0·8) were used for Northern analysis to determine the expression more quantitatively. The results clearly showed that the mutated genes were strongly expressed in many clones (Fig. 3ad, upper columns). The RNAs of the mutated G-proteins ScGP-A, ScGP-B and ScGP-C were of the same size (approx. 1·6 kbp). Since the mutated genes that were inserted into the gene-expression vectors had intact introns and different sizes (ScGP-A, 1·4 kbp; ScGP-B, 1·5 kbp; ScGP-C, 1·9 kbp; see Fig. 1), this result suggests that the introns of the mutated G-proteins had been removed. In contrast, the endogenous G-proteins could not be detected. This could have been due to the low expression levels compared to the transformed mutated genes driven by the SC3 or the GPD promoter.

Table 1). (b) Monokaryon clones co-transformed with expression vector pSC3GPB (lanes 110) or pGPDGPB (lanes 11 and 12). Lanes 13 and 14 contained the same as their respective lanes in (a). (c) Monokaryon clones co-transformed with the expression vector pSC3GPC (lanes 19). Lanes 10 and 11 contained monokaryon clones transformed with pGEM7+/TRP1 only. Lanes 12 and 13 contained dikaryon strains T11/F1-4 and T11/F1-5. (d) Monokaryon clones co-transformed with expression vector pGPDGPC (lanes 18). Lanes 9 and 10 contained monokaryon clones co-transformed with pSC3luc. Lanes 11 and 12 contained monokaryon clones transformed with pGEM7+/TRP1 only. Lanes 13 and 14 contained dikaryon strain T11/F1-4 and T11/F1-5. (e) Monokaryon clones transformed with pGEM7+/TRP1 only (lanes 16). Lanes 7 and 8 contained dikaryon strains T11/F1-4 and T11/F1-5. Lanes 914 contained monokaryon clones co-transformed with pSC3luc. Note that the RNA samples loaded into lanes 1 and 7 were also used in (ad) for comparing the expression level of the SC3 gene across multiple blotting membranes (see Results and Discussion).

Analysis of the influence of the mutated ScGP-A, ScGP-B and ScGP-C genes on aerial-hyphae formation in the monokaryon
All co-transformed and control (pGEM7⁺/TRP1 only) clones were grown on CYM agar plates for 7 days to observe the degree of aerial-hyphae formation. When compared with the control clones transformed with only pGEM7⁺/TRP1, aerial-hyphae formation in some co-transformed clones expressing mutated ScGP-A or ScGP-C genes was reduced (Fig. 4). Over half of the clones that strongly expressed the mutated genes exhibited the suppression of aerial-hyphae formation, though there were some exceptions (see notation s and ss in Fig. 3). Upon microscopic examination, the mycelia of some clones exhibiting the suppression of aerial-hyphae formation appeared as short irregular branches, though the degree of these branches was not prominent. In S. commune, a similar phenotype to this has been described as flat (reviewed by Kothe, 1999 ), though the degrees of suppression of aerial-hyphae formation and the short irregular mycelium branches found in strains F2-4/T11 and F2-5/T11 (see Table 1) were more prominent. Currently, it is unclear whether the suppression of aerial-hyphae formation and the deformity of submerged mycelium observed in clones expressing ScGP-A or ScGP-C are the same as that of the flat phenotype, because the phenotype of the clones expressing these two genes is not so clearly observed.

(23K): Fig. 4. Correlation between the mutated-gene-expression level and aerial-hyphae formation in the monokaryon. Plots represent independent clones harbouring mutated genes. Circles represent the clones co-transformed with pSC3GPA, pSC3GPB and pSC3GPC. Triangles represent the clones co-transformed with pGPDGPA, pGPDGPB and pGPDGPC. The half-shaded circles and triangles represent the clones whose aerial-hyphae formation was slightly suppressed. The fully shaded circles represent the clones whose aerial-hyphae formation was suppressed. All control clones co-transformed with pSC3luc or pGEM7+/TRP1 only appeared normal; no clone had suppressed aerial-hyphae formation (data not shown).

Northern analysis of the hydrophobin SC3 gene expressed in transformed clones
Since it has been reported that the hydrophobin SC3 gene is a major component involved in the formation of aerial hyphae and the disruption of SC3 affects the properties of the hyphal wall in S. commune (Wösten et al., 1994 ; van Wetter et al., 2000 ), Northern analysis was performed to examine if the expression of SC3 was reduced in the clones expressing mutated ScGP-A or ScGP-C genes. The clones that were considered to strongly express the mutated genes were grown on a cellophane membrane, and their total RNA was extracted. To compare the level of SC3 expression between multiple blotting membranes, two RNA samples obtained from control clones (pGEM7⁺/TRP1 only) and strain T11/F2-4 (a flat phenotype strain) were loaded onto all blotting membranes. The exposure time for the photographic film was adjusted so that the two bands were seen at the same intensity on each membrane, thus enabling a comparison of the signal intensities of all clones across multiple membranes. Unexpectedly, the level of SC3 expression was so variable that it was difficult to relate the suppression of aerial-hyphae formation with this gene (Fig. 3ae; lower panels). Though the expression of SC3 was suppressed in clones harbouring the mutated ScGP-C gene (Fig. 3c, d), in general it was difficult to conclude that the mutated gene suppressed SC3 expression, because some control clones transformed with pSC3luc (Fig. 3e; lanes 11 and 13) or the mutated ScGP-B gene (Fig. 3b; lanes 3, 5, 6 and 10) also exhibited suppression of SC3 expression. Since the loaded RNA samples were intact, these results were not due to artefacts from contamination with trace RNAase (data not shown). It was unclear why such a variation in the levels of SC3 mRNA expression was observed. Since the level of SC3 expression is affected by the growth stage of the organism (Wessels, 1997 ), the subtle difference of the growth condition on each clone might influence the levels of expression seen. In addition to this effect, titration of transcription factors caused by the introduction of the SC3 promoter region might occur. This is a phenomenon that has been reported previously (Schuurs et al., 1997 ).

Mating test of clones transformed with the mutated G-proteins
A mating test was done to confirm the influence of the mutated G-proteins on clamp-connection formation. Some of the clones that strongly expressed the mutated genes were mated with strains F2-1, F2-2 and F2-3, to ascertain whether the constitutively active G-protein α-subunits entirely stimulated the mating pheromone signalling pathway. The strains resulting from these matings were expected to contain mating-type locus A (not present in T11 strains) and mating-type locus B (present in T11 strains), because these clones exhibited the barrage phenotype when mated with T11. Mating-type locus B has been shown to contain genes encoding pheromones and their G-protein-coupled receptor (Specht, 1995 ; Wendland et al., 1995 ; Vaillancourt et al., 1997 ; Gola et al., 2000 ), hence strains containing this mating type were expected to form clamp-connections by directly stimulating the pheromone signalling pathway. However, none of the clones that strongly expressed the mutated genes could form clamp-connections, which was contrary to our expectations (data not shown).

Effect of mutated G-proteins on fruit-body formation
To examine the influence of the mutated G-proteins on fruit-body formation, clones harbouring the mutated genes were mated with the F1 strain. The resulting dikaryons were cultured on ASN agar plates for 7 days in continuous darkness, followed by 4 days under light. Contrary to the slight suppression of aerial-hyphae formation, the mutated ScGP-A and ScGP-C genes markedly blocked fruiting in the dikaryon strains (Figs 5a, b and 6a, c). That is, most of the clones expressing the mutated ScGP-A or ScGP-C genes (whose mutated-gene-expression level was nearly 1) could not form fruit-bodies. This phenomenon was unlikely to have been an artefact, because all of the clones harbouring the control plasmids (pSC3luc or pGEM7⁺/TRP1 alone; see Figs 5c and 6d) and the mutated ScGP-B gene (Fig. 6b) formed fruit-bodies normally. As all of the clones harbouring the mutated genes could form clamp-connections when mated with tester strain F1 (data not shown), it seemed that the mutated ScGP-A and ScGP-C genes interfered with the process of fruit-body formation. It is thought that light, especially blue light, is needed to induce primordia formation in S. commune (Perkins, 1969 ; Perkins & Gordon, 1969 ), hence we also cultured the dikaryons on ASN agar plates for 11 days in continuous darkness to see if some clones could form primordia. All of the clones tested in this way failed to form primordia (data not shown).

(69K): Fig. 5. Suppression of fruit-body formation caused by mutated ScGP-A and ScGP-C genes in the dikaryon. (a) Clone 34 co-transformed with pSC3GPA. The mutated-gene-expression level was 1. (b) Clone 32 co-transformed with pSC3GPC. The mutated-gene-expression level was 1. (c) Clone 1 transformed with pGEM7+/TRP1 only, as a negative control. White arrows in (c) indicate the fruit-bodies formed in the peripheral area of the colony. All photographs were taken 4 days after the cultures were illuminated.

(10K): Fig. 6. Correlation between the mutated-gene-expression level and the fruit-body-formation index in the dikaryon. (ac) Show the correlation between the level of expression of the mutated genes and the fruit-body-formation index defined in Methods. Every plot represents the independent clones harbouring mutated genes, screened as described in Methods. , Clones co-transformed with pSC3GPA, pSC3GPB and pSC3GPC. , Clones co-transformed with pGPDGPA, pGPDGPB and pGPDGPC. (d) Shows the degree of fruit-body formation of control clones co-transformed with pSC3luc () or pGEM7+/TRP1 () only. Note that the x-axis does not indicate the mutated-gene-expression level but simply the clone number, since these clones harbour no exogenous mutated genes.

We started this research to obtain some clues about the role of G-protein α-subunits in S. commune. In particular, we expected that constitutively active G-protein α-subunits would stimulate the mating pheromone signalling pathway directly without the need for the compatible mating pheromone. Our results showed that none of the mutated G-proteins could fully mimic the compatible mating pheromone. That is, none of the clones expressing mutated genes showed an apparent loss of aerial mycelium a characteristic feature of a flat phenotype. From these observations it appears that the ScGP-A, ScGP-B and ScGP-C genes are not involved in the mating pheromone signalling pathway. However, there is the possibility that the pheromone receptor does not need G-protein α-subunits but instead needs the free ßγ dimer to transduce the signal to cells. Alternatively, the pheromone receptor may require both the α-subunits and free ßγ dimer to transduce the signal. In Saccharomyces cerevisiae, the mating pheromone signal is transduced by the free ßγ dimer; the α-subunit, GPA1, in this organism plays only a negative role in pheromone signalling by repressing the ßγ dimer (Clapham & Neer, 1993 ; Leberer et al., 2000 ). As Kothe pointed out in her review (Kothe, 1996 ), knock-out mutants should be created to help us to understand the role of the G-proteins that are involved in mating pheromone signalling in detail.

Though it is difficult to speculate about the mechanism by which mutated ScGP-A and ScGP-C genes block fruiting, the suppression of fruit-body and aerial-hyphae formation seen in this study is similar to that caused by a thn mutation in Schizophyllum commune. This thn mutation has a distinct phenotype: the suppression of aerial-hyphae formation, the appearance of corkscrew-like or wavy hyphae in the substrate and the suppression of both aerial-hyphae and fruit-body formation in a dikaryon homozygous for thn (Raper & Miles, 1958 ; Schwalb & Miles, 1967 ; Raper, 1988 ; Wessels et al., 1991 ). Though the change in appearance of mycelium in the growth substrate was not obvious upon microscopic examination of clones expressing the mutated ScGP-A or ScGP-C genes described above, the other characteristics were similar to those seen in the thn mutant. Recently, the thn-1 gene of S. commune was identified (Fowler & Mitton, 2000 ). It was reported that the thn-1 cDNA of this organism showed strong similarity to a large number of protein sequences (including FlbA from Aspergillus nidulans) featuring the RGS domain. The RGS domain family acts on heterotrimeric G-protein α-subunits to increase the ratio of inactive GDP-bound forms (Berman et al., 1996a , b ; Hepler et al., 1997 ). We anticipate that both the loss-of-function mutation in thn-1 and the gain-of-function mutations in G-protein α-subunits generated in this study culminate in an increase in the number of particular G-protein α-subunit(s) locked in the active GTP-bound state. Hence, thn-1 might function as a regulator of the ScGP-A and ScGP-C genes. Interestingly, the constitutively active mutated FadA (G42R) gene, which is under the control of FlbA and belongs to the heterotrimeric G-protein α-subunit family, exhibits a similar phenomenon in A. nidulans (Yu et al., 1996 ). A mutation in FadA causes a fluffy phenotype and blocks sporulation. Given that ScGP-A and/or ScGP-C are homologues of Fad, one of the roles of these genes might be to suppress the differentiation of the mycelium, by affecting processes such as sporulation and fruiting.

Since the mutated ScGP-B gene did not generate any characteristic phenotypic features, no clue as to the function of this gene was obtained. Because the amino-acid sequence of ScGP-B is similar to that of CGPα1, which is supposed to be a candidate for regulating the blue-light-induced signal transduction photomorphogenesis system found in C. congregatus, we anticipated that the constitutively active ScGP-B gene might stimulate the formation of primordia without light. However, all of the clones tested failed to form primordia in darkness. As has been suggested for the ScGP-A and ScGP-C genes, the construction of knock-out mutants of ScGP-B might help us to see if ScGP-B is needed to transmit signals induced by blue light, which stimulates the formation of primordia in S. commune.

cAMP levels might be up- or down-regulated in clones expressing mutated genes, because G-protein α-subunits regulate adenylate cyclase in many species (Hamm, 1998 ; Morris & Malbon, 1999 ). However, the levels of cAMP in the clones used in this study were not measured. Since cAMP has been shown to be a signalling factor for fruit-body formation in the basidiomycete Coprinus macrorhizus (Uno & Ishikawa, 1973 ), the measurement of cAMP levels in S. commune clones expressing mutated genes will help us to examine the relationship between adenylate cyclase and G-protein α-subunits in this organism.

We wish to express our sincere thanks to Dr H. Kinoshita and Dr K. Sen, Shinsyu University, for their valuable technical advice concerning S. commune transformation experiments. We would also like to thank Dr T. Yoshida for his technical advice concerning DNA manipulation, and Y. Shibuya and I. Sato for their technical assistance. This work was supported in part by a Grant-in-Aid (Bio Design Program) from the Ministry of Agriculture, Forestry and Fisheries, BDP02-VI-1-1.